PHD Thesis Nshiers

ACOUSTIC DESIGN FOR INPATIENT FACILITIES IN HOSPITALS
Thesis submitted in partial fulfilment of the requirements of London South Bank University for the degree of Doctor of Philosophy By Nicola Jane Shiers
Supervisor: Professor B.M. Shield, London South Bank University Second Supervisor: Rosemary Glanville, London South Bank University
December 2011
Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________
Table of Contents
LIST OF FIGURES . viii LISTOF TABLES . xiii Acknowledgements . xv Acoustic glossary and definitions xvi Chapter 1 Chapter 2 2.1 2.2 2.3 2.4 2.5 Introduction .. 2 Acoustic standards and guidance . 5 Introduction .. 5 UK design guidance .. 5 2.2.1 HTM 08-01 ....... 6 European healthcare design guidance 10 World Health Organisation guidelines . 11 Control of infection . 2.5.1 HTM 60 .. 2.5.2 National Standards of Cleanliness for the NHS.......... 2.5.3 HFN 30 Infection Control in the built environment .. 12 12 13 14
2.6 2.7 Chapter 3 3.1 3.2
Discussion ... 14 Conclusions . 15 Previous research on hospital noise 16 Introduction .. 16 Noise measurement studies . 16 3.2.1 Limitations of noise measurement studies ....... 18 3.2.2 Understanding the overall hospital noise climate ... 21 3.2.3 Identifying noise sources .... 22 3.2.4 Discussion .... 23 Sleep studies .. 23 3.3.1 Modification of room acoustics and its effects on sleep . 24 3.3.2 Discussion 25 The effects of behaviour modification on hospital noise .. 25 3.4.1 Discussion . 26 The effects of room acoustic design modifications 3.5.1 Control of infection and room acoustics .. 3.5.2 Physiological response to acoustic modification ........... 3.5.3 Discussion .... 27 27 28 29
3.3
3.4 3.5
3.6 Chapter 4
Conclusions ..... 29 The effects of noise on staff and patients . 30 i
4.1 4.2
Introduction ... 30 Effects of noise on staff .. 30 4.2.1 Stress levels and burnout 30 4.2.2 Cognitive function / memory 30 4.2.3 Effects of acoustic design on the work environment ...... 31 4.2.4 Discussion . 31 Effects of noise on patients .. 4.3.1 Recovery rates 4.3.2 Subjective response to noise . 4.3.3 Speech privacy . 4.3.4 Single bed patient rooms 4.3.5 Discussion . 32 32 32 33 33 33
4.3
4.4 Chapter 5 5.1 5.2
Conclusions .. 34 Study design ... 35 Introduction . 35 Study outline aims and objectives 5.2.1 Acoustic survey 5.2.2 Questionnaire surveys 5.2.3 Comparison studies 35 36 36 37
5.3
Acoustic survey methodology . 38 5.3.1 Equipment . 38 5.3.2 Control of Infection .. 38 5.3.3 Acoustic parameters 39 5.3.4 Presentation of sound levels .. 39 5.3.5 Measurement interval .. 39 5.3.6 Measurement locations .. 39 5.3.7 Identifying sources of high level noise without an observer .. 40 5.3.8 Reverberation times ..40 Questionnaire survey design .. 40 5.4.1 Staff questionnaires . 41 5.4.2 Patient questionnaires . 41 Preliminary work 42 5.5.1 Building relationships with hospitals and Healthcare Trusts .. 42 5.5.2 Ethics and Trust approval 43 Conclusions .. 44
5.4
5.5
5.6
Chapter 6 6.1 6.2 6.3 6.4
Pilot study ..... 45 Introduction ... 45 Background .. 45 Sky Ward .. 46 Building acoustic design considerations .. 47 6.4.1 Nurse stations and common areas 47 6.4.2 Patient accommodation 47 ii
6.5
Ward routines 48 6.5.1 Staffing and patient levels 48 6.5.2 Staff shift patterns and ward rounds .. 49 6.5.3 Cleaning ..49 6.5.4 Meal times .. 49 6.5.5 Medical equipment with alarms .. 49 6.5.6 Access to patient accommodation . 49 Measurement locations .. 50 6.6.1 Nurse stations 50 6.6.2 Four bed bays 53 6.6.3 Single patient rooms . 55 Equipment and microphone positioning ... 56 6.7.1 Nurse stations 57 6.7.2 Four bed bays 57 6.7.3 Single patient rooms . 58 Other considerations 58 6.8.1 Identifying the optimal level above setting for trigger files 58 6.8.2 Publicising the study . 59 6.8.3 Reverberation times ..59 Questionnaire survey considerations 60 Overall acoustic survey results .. 60 Nurse stations .. 62 6.11.1 Sources of high level noise . 63 Four bed bays .. 66 6.12.1 Sources of high level noise . 67 Single patient rooms 69 6.13.1 Sources of high level noise . 70 Establishing a representative measurement interval 71 Other measured acoustic parameters . 73 6.15.1 Reverberation times . 73 6.15.2 Ambient noise levels 74 Results of the staff questionnaire surveys .. 75 6.16.1 Staff profile . 75 6.16.2 Noise annoyance and interference 75 6.16.3 Important sounds .. 78 Patient questionnaires . 79 6.17.1 Parent / patient profile .. 79 6.17.2 Noise annoyance .. 80 6.17.3 Positive sounds . 82 6.17.4 Privacy and ease of hearing 83 6.17.5 Patients questionnaire comments . 83 Summary of results .. 83
6.6
6.7
6.8
6.9 6.10 6.11 6.12 6.13 6.14 6.15
6.16
6.17
6.18 6.19 6.20
Follow up discussions 84 Conclusions . 87 iii
Chapter 7 7.1 7.2 7.3 7.4
Bedford Hospital .. 91 Introduction ... 91 Background .. 91 Building acoustic design considerations . 92 Hospital policies and equipment common to both wards 92 7.4.1 Meal times 92 7.4.2 Ward design . 93 7.4.3 Occupancy levels 94 7.4.4 Shift patterns 94 7.4.5 Visiting hours 94 7.4.6 Ward access . 94 7.4.7 Access to patient accommodation 94 7.4.8 Cleaning staff 94 7.4.9 Mobile phone policy . 94 7.4.10 Entertainment systems 95 7.4.11 Rubbish bins .. 95 7.4.12 Staff call .. 95 7.4.13 Medical equipment alarms .. 95 7.4.14 Trolleys ... 95 7.4.15 Internal telephones 95 7.4.16 Hand gels 95 Medical ward . 96 7.5.1 Ward specific information .98 Medical ward overall noise survey results 100 7.6.1 Nurse station and ward entrance 101 7.6.2 Multi-bed bays 103 7.6.3 Single patient rooms . 105 7.6.4 Further analysis of high level noise sources . 106 Surgical ward 109 7.7.1 Ward specific information .109 Surgical ward overall noise survey results .. 112 7.8.1 Nurse station .. 113 7.8.2 Multi-bed bays 115 7.8.3 Single patient rooms . 116 7.8.4 Further analysis of high level noise sources . 118 Results of the staff questionnaire surveys ............ 120 7.9.1 Staff profile . 120 7.9.2 Noise annoyance .. 121 7.9.3 Interference with work . 124 7.9.4 Important sounds .. 125 Results of the patient questionnaire surveys .. 126 7.10.1 Patient profiles.... 126 7.10.2 Noise annoyance and disturbance . 127 7.10.3 Positive sounds ..... 131 7.10.4 Ease of hearing and privacy ... 132 Questionnaire comments .. 132 iv
7.5 7.6
7.7 7.8
7.9
7.10
7.11
7.12 7.13 Chapter 8 8.1 8.2 8.3 8.4
Summary . 132 Conclusions . 133 Ceiling intervention study, Bedford Hospital . 135 Introduction .. 135 Bay information 136 Effect of ceiling tile change on noise levels . 137 Effect of ceiling tile change on reverberation time .. 140 8.4.1 Unoccupied reverberation times . 140 8.4.2 Occupied reverberation times . 141 Comparison of unoccupied and occupied RTs .. 142 Conclusions .. 143 Addenbrookes Hospital ................ 144 Introduction .. 144 Background .. 144 Ward D8 (surgical) .. 145 9.3.1 Building design . 145 9.3.2 Ward layout 146 9.3.3 Ward specific information 146 9.3.4 Managing the study .. 148 Overall noise survey results Ward D8 9.4.1 Nurse station . 9.4.2 Multi-bed bays .. 9.4.3 Further analysis of high level noise sources 9.4.4 Representative measurement interval .. 151 152 154 155 158
8.5 8.6 Chapter 9 9.1 9.2 9.3
9.4
9.5
Ward N3 (medical) . 160 9.5.1 Building design . 160 9.5.2 Ward layout 160 9.5.3 Ward specific information 161 9.5.4 Managing the study .. 163 Overall noise survey results Ward N3 .. 166 9.6.1 Nurse station . 167 9.6.2 Multi-bed bays .. 170 9.6.3 Single patient rooms 171 9.6.4 Further analysis of high level noise sources 172 Ward M4 (surgical) .. 174 9.7.1 Building construction .... 174 9.7.2 Ward layout 174 9.7.3 Ward specific information 175 9.7.4 Managing the study .. 177 Overall noise survey results Ward M4 .. 181 9.8.1 Nurse station . 182 9.8.2 Multi-bed bays .. 184 9.8.3 Single patient rooms 185 v
9.6
9.7
9.8
9.8.4 9.9
Further analysis of high level noise sources 186
Results of the staff questionnaire surveys ............ 188 9.9.1 Staff profile . 188 9.9.2 Noise annoyance .. 189 9.9.3 Interference with work .. 190 9.9.4 Important sounds .. 192 Results of the patient questionnaire surveys .. 193 9.10.1 Patient profiles... 193 9.10.2 Noise annoyance and disturbance 194 9.10.3 Positive sounds .... 198 9.10.4 Ease of hearing and privacy 199 Questionnaire comments .. 199 Summary . 199 Conclusions . 201 Blind estimation of reverberation time . 202 Introduction .. 202 Initial validation 202 Validation using real and simulated measurements . 203 10.3.1 Validation 1 204 10.3.2 Validation 2 205 10.3.3 Validation study conclusions 207 Estimation of RT in occupied hospital wards .. 208 10.4.1 Methodology .. 208 10.4.2 MLE-RT20 estimates from day time data ... 209 Comparison of day and night time MLE-RT20 estimates 212 Summary . 214 Conclusions 215 Analysis of objective and subjective data 216 Introduction ... 216 Factors affecting noise levels 216 11.2.1 Effect of bay size 216 11.2.2 Surgical and medical wards 219 11.2.3 Impact of high level noise events on overall noise levels 220 11.2.4 Noise levels and reverberation times . 221 Factors affecting patient perceptions of noise . 11.3.1 Overall 11.3.2 Patient gender .. 11.3.3 Age . 11.3.4 Hearing impairment . 11.3.5 Length of stay .. 11.3.6 Bed position .. 11.3.7 Speech intelligibility and privacy vi 222 222 224 225 226 227 229 230
9.10
9.11 9.12 9.13 Chapter 10 10.1 10.2 10.3
10.4
10.5 10.6 10.7 Chapter 11 11.1 11.2
11.3
11.4
Factors affecting staff perceptions of noise 231 11.4.1 Overall 231 11.4.2 Staff gender 231 11.4.3 Age .. 232 11.4.4 Time worked on the ward 233 11.4.5 Time worked at the hospital 233 11.4.6 Relationship between noise annoyance and noise interference 234 Discussion . 234 Conclusions 236 Noise control in inpatient care .. 237 Introduction .. 237 Optimising the acoustic design of the ward .... 12.2.1 Design for infection control 12.2.2 The effects of adding acoustic absorbency.. 12.2.3 Ward design . 12.2.4 Building construction .. 12.2.5 Building age and overall noise levels ... 237 237 238 238 239 240
11.5 11.6 Chapter 12 12.1 12.2
12.3
Ward equipment ... 242 12.3.1 Nurse call systems 242 12.3.2 Internal telephones 242 12.3.3 Medical equipment alarms .. 243 12.3.4 Doorbell .. 244 12.3.5 Rubbish bins .. 244 12.3.6 Ward furniture 244 12.3.7 Wheeled equipment .. 244 12.3.8 Ring binders ... 245 12.3.9 Doors ... 245 Human behaviour 245 WHO guidelines . 247 Acoustic parameters . 247 Conclusions 248 Conclusions . 249 Introduction .. 249 Overall conclusions .. 249 13.2.1 Building design .. 249 13.2.2 Patient accommodation 250 13.2.3 Staff and patient perceptions .. 250 13.2.4 Ward equipment 250 13.2.5 Human behaviour .. 251 13.2.6 Guidelines .. 251 Recommendations .... 251 Further work 252
12.4 12.5 12.6 12.7 Chapter 13 13.1 13.2
13.3 13.4
References ... 253 vii
Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________
LIST OF FIGURES Figure 2.1 Figure 3.1 Figure 3.2 Figure 5.1 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Figure 6.6 Figure 6.7 Figure 6.8 Figure 6.9 Figure 6.10 Figure 6.11 Figure 6.12 Figure 6.13 Figure 6.14 Figure 6.15 Figure 6.16 Figure 6.17 Figure 6.18 Figure 6.19 Figure 6.20 Figure 6.21 Figure 6.22 Figure 6.23 Figure 6.24 Figure 6.25 Figure 6.26 Figure 6.27 Figure 6.28 Figure 6.29 Figure 6.30 Figure 6.31 Figure 6.32 Figure 6.33 Figure 6.34 Figure 6.35 Figure 6.36 Recommended RTs for different room functions and volumes, HTM 2045 (NHS Estates, 1996) LAeq values measured in hospitals during day time hours as a function of the year of study publication. (Busch-Vishinac et al, 2005) LAeq values measured in hospitals during night time hours as a function of the year of study publication. (Busch-Vishinac et al, 2005) Sound level meter, environmental case and associated equipment The Octav Botnar Wing Main entrance to the Octav Botnar Wing Sky Ward Reception Typical four bed bay Ultima ceiling tile sound absorption coefficients () over a range of frequencies Layout of Sky Ward with microphone positions Nurse Station 1 Internal corridor Internal telephone & security monitor Wall mounted speaker grill Nurse station 2 Nurse station 2 desk 4-bed bay B Hand washing sink, door to shower room and lockers Patient bed and fold down chair Ward entrance with rubbish bins Patient bed showing bed head services Locked door onto balcony Door to ensuite, pull down bed, sink and rubbish bins Patient bed and opening windows Rubbish bins and hand washing sink Pull down bed Flat screen television Microphone position at nurse station 1 Microphone position at nurse station 2 4-bed bay B with microphone placed on top of lockers Single patient room 1 with microphone position shown Single patient room 2 with microphone position shown Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse stations Nurse call console The number and levels (LAmax) of occurrences of the nurse call system at nurse station 1, measured at 3 m over 5 days The number and levels (LAmax) of occurrences of the ward doorbell at nurse station 2, measured at 3 m over 19 hours Average LAmax of the nurse call system, internal telephone and ward doorbell Average LAeq,1hr and LA90,1hr levels over 24 hours for 4-bed bays A and B Percentages of high level noise events by type measured in 4-bed bays A and B viii
9 17 17 38 46 46 47 47 48 51 50 50 52 52 53 53 54 54 54 54 55 55 55 56 56 56 56 57 57 58 58 58 62 63 64 64 65 66 67 68
Figure 6.37 Figure 6.38 Figure 6.39 Figure 6.40 Figure 6.41 Figure 6.42 Figure 6.43 Figure 6.44 Figure 6.45 Figure 6.46 Figure 6.47 Figure 6.48 Figure 6.49 Figure 6.50 Figure 6.51 Figure 7.1 Figure 7.2 Figure 7.3 Figure 7.4 Figure 7.5 Figure 7.6 Figure 7.7 Figure 7.8 Figure 7.9 Figure 7.10 Figure 7.11 Figure 7.12 Figure 7.13 Figure 7.14 Figure 7.15 Figure 7.16 Figure 7.17 Figure 7.18 Figure 7.19 Figure 7.20 Figure 7.21 Figure 7.22 Figure 7.23 Figure 7.24 Figure 7.25 Figure 7.26
Average LAeq,1hr and LA90,1hr levels over 24 hours for single patient rooms A and B Percentages of high level noise events by type for single patient rooms A&B LAeq,1hr levels measured over five consecutive days at nurse station 1 Average LAeq,1hr levels over 24 hours for week 1 and week 2 at nurse station 1 Distribution of the extent of staff annoyance Percentage of staff rating an annoyance noise event with a 2, 3 or 4 Distribution of the extent of noise interference with work Percentage of staff rating an interference noise event with a 2, 3 or 4 Mean importance rating of certain noise events Parents by age bracket Patients by age bracket Distribution of the extent of parent / patient annoyance during the day time Percentage of parents / patients rating an annoyance noise event with a 2, 3 or 4 Distribution of the extent of parent / patient disturbance during the night time Percentage of parents / patients rating a disturbance noise event with a 2, 3 or 4 Original building, Bedford Hospital (1803) Five storey ward block Main hospital entrance Medical ward kitchen Hospicom entertainment console Automatic hand gel dispenser Single patient room Microphone suspended from ceiling Microphone on tripod Detailed plan of the medical ward showing microphone positions Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station and ward entrance LAmax,2s and LAeq,2s fluctuating over a ten minute interval at the nurse station Average LAeq,1hr levels over 24 hours for the multi-bed bays Average LA90,1hr levels over 24 hours for the multi-bed bays Average LAeq,1hr levels over 24 hours for the single rooms LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a three hour period in single room A Average number of high level noise events recorded at each location per day Average number of high level noise events recorded at each location per night Detailed plan of the surgical ward showing microphone positions Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station LAmax,2s (green trace) and LAeq,2s (red trace) measured over a thirty minute interval during the night (2.40am onwards) at the nurse station LAmax,2s and LAeq,2s fluctuating over a 11 minute interval at the nurse station Average LAeq,1hr for each multi-bed bay and combined average LA90,1hr level for all bays over 24 hours Average LAeq and LA90 levels for single rooms A and B and multi-bed bays ix
69 70 72 72 76 76 77 78 79 79 79 80 81 81 82 91 92 92 93 96 96 96 97 97 99 101 102 102 104 104 105 106 106 108 111 113 113 114 115 116 117
Figure 7.27 Figure 7.28 Figure 7.29 Figure 7.30 Figure 7.31 Figure 7.32 Figure 7.33 Figure 7.34 Figure 7.35 Figure 7.36 Figure 7.37 Figure 7.38 Figure 7.39 Figure 7.40 Figure 7.41 Figure 7.42 Figure 7.43 Figure 7.44 Figure 8.1 Figure 8.2 Figure 8.3 Figure 8.4 Figure 8.5 Figure 8.6 Figure 8.7 Figure 8.8 Figure 8.9 Figure 9.1 Figure 9.2 Figure 9.3 Figure 9.4 Figure 9.5 Figure 9.6 Figure 9.7 Figure 9.8 Figure 9.9 Figure 9.10 Figure 9.11 Figure 9.12 Figure 9.13
LAmax,2s (green trace) and LAeq,2s (red trace) showing the noise levels due to a medical equipment alarm over a period of 13 minutes Average number of high level noise events captured at each location per day Average number of high level noise events captured at each location per night Age of respondents by band Time worked on the ward Time worked at the hospital Staff perception of noise in terms of annoyance The percentage of staff rating an annoyance noise event with a 2, 3 or 4 Staff perception of the extent to which noise interferes with work The percentages of staff rating an interference noise event with a 2, 3 or 4 Mean importance rating of certain noise events Gender split by ward type Patients age by band Length of patient stay when completing the questionnaire Patient perception of the day time ward noise environment The percentage of patients rating an annoyance noise event with a 2, 3 or 4 Patient perception of the night time ward noise environment The percentage of patients rating a disturbance noise event with a 2, 3 or 4 Photographs of the bay during refurbishment Absorption coefficients of Armstrong Bioguard Plain ceiling tiles Source: Manufacturers product specification sheet Absorption coefficients of Armstrong Bioguard Acoustic ceiling tiles Source: Manufacturers product specification sheet Average LAeq,1hr levels over 24 hours pre and post ceiling change Average number of trigger files recorded over 24 hours by event type The frequency content of noise of metal cutlery Source (S) and receiver (R) positions used to measure reverberation time in the unoccupied bay before and after the ceiling change Average unoccupied RT20 measurements with 95% confidence limits (Impulse Response Method) Occupied MLE-RT20 estimates pre and post the ceiling replacement Original building, Addenbrookes Hospital Detailed plan of ward D8 showing microphone positions Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a ten minute interval at the nurse station during the night Average LAeq,1hr levels over 24 hours for the multi-bed bays Average number of high level noise events recorded at each location per day Average number of high level noise events recorded at each location per night LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 19 minute interval at in the elderly trauma unit Average LAeq,1hr levels over 24 hours for two non-consecutive weeks in the 12-bed bay Detailed plan of the ward N3 showing microphone positions Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station x
118 119 120 121 121 121 122 123 124 124 125 126 126 127 128 129 130 131 135 136 136 138 139 139 140 141 142 144 150 152 153 153 155 156 157 158 159 165 167 168
Figure 9.14 Figure 9.15 Figure 9.16 Figure 9.17 Figure 9.18 Figure 9.19 Figure 9.20 Figure 9.21 Figure 9.22 Figure 9.23 Figure 9.24 Figure 9.25 Figure 9.26 Figure 9.27 Figure 9.28 Figure 9.29 Figure 9.30 Figure 9.31 Figure 9.32 Figure 9.33 Figure 9.34 Figure 9.35 Figure 9.36 Figure 9.37 Figure 9.38 Figure 9.39 Figure 9.40 Figure 9.41 Figure 9.42 Figure 9.43 Figure 9.44 Figure 10.1 Figure 10.2 Figure 10.3 Figure 10.4 Figure 10.5 Figure 10.6 Figure 10.7
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 13 minute interval at the nurse station during the afternoon Frequency content of door bang at the nurse station Average LAeq,1hr for each multi-bed bay and combined average LA90,1hr level for all bays over 24 hours Average LAeq,1hr levels over 24 hours for the single rooms Average number of high level noise events recorded at each location per day Average number of high level noise events recorded at each location per night Percentage break down of high level noise events by type in 4-bed bay B Plan of Ward M4 detailing shared areas and microphone positions Detailed plan of Ward M4 showing study locations and microphone positions Average day and night LAeq levels measured at each location Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse stations LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 15 minute interval at the nurse station at 05.30 Average LAeq,1hr and LA90,1hr level for each multi-bed bay over 24 hours Average LAeq,1hr and LA90,1hr levels over 24 hours for the single rooms Average number of high level noise events recorded at each location per day Average number of high level noise events recorded at each location per night Age of respondents by band Time worked on the ward Time worked at the hospital Staff perception of noise in terms of annoyance The percentage of staff rating an annoyance noise event with a 2, 3 or 4 Staff perception of the extent to which noise interferes with work The percentages of staff rating an interference noise event with a 2, 3 or 4 Mean importance rating of certain noise events Gender split by ward type Patients age by band Length of patient stay when completing the questionnaire Patient perception of the day time ward noise environment The percentages of patients on Ward D8 rating an annoyance noise event with a 2, 3 or 4 Patient perception of the night time ward noise environment The percentages of patients rating a disturbance noise event with a 2, 3 or 4 Clinical skills laboratory used for validation 1 Average RT20 measurements with 95% confidence limits (Impulse Response Method) Average RT20 measurements with 95% confidence limits (Impulse Response Method) Accuracy of RT20 estimations in relation to actual measured values MLE-RT20 estimates for five multi-bed bays in Ward D8, Addenbrookes Hospital (day time data) with 95% confidence limits MLE-RT20 estimates for six locations in Ward N3, Addenbrookes Hospital (day time data) with 95% confidence limits MLE-RT20 estimates for seven locations in the surgical ward, Bedford Hospital (day time data) with 95% confidence limits xi
168 169 170 171 172 173 173 179 180 182 183 183 184 185 187 187 188 189 189 189 190 191 191 192 193 193 194 195 196 197 198 204 204 207
210 210 211
Figure 10.8 Figure 10.9 Figure 10.10 Figure 11.1 Figure 11.2 Figure 11.3 Figure 11.4 Figure 11.5 Figure 11.6 Figure 11.7 Figure 11.8 Figure 11.9 Figure 11.10 Figure 11.11 Figure 11.12 Figure 11.13 Figure 11.14 Figure 11.15 Figure 11.16 Figure 11.17 Figure 11.18 Figure 11.19 Figure 11.20 Figure 11.21 Figure 11.22 Figure 12.1 Figure 12.2
Comparison of day and night time estimates, 7-bed bay, Ward D8, Addenbrookes Hospital Comparison of day and night time estimates, 12-bed bay, Ward D8, Addenbrookes Hospital Comparison of day and night time estimates, 4-bed bay, medical ward, Bedford Hospital Average day time levels by bay size for all main study wards Average night time levels by bay size for all main study wards Average day time noise levels and average number of day time high level noise events for each bay Average night time noise levels and average number of night time high level noise events for each bay Average day time noise levels and estimated reverberation times in each bay Overall patient perception of the noise climate Overall percentages of patient annoyed / disturbed by noise Mean patient perception rating of noise by gender Percentages of patients annoyed / disturbed by noise by gender Mean rating of patient perceptions of day and night noise and age Percentages of patients annoyed / disturbed and age Percentage of hearing impaired by age group Percentage of patients annoyed / disturbed with hearing impairment Mean rating of patient perceptions of day and night noise and length of stay Percentages of patients annoyed / disturbed and length of stay Percentages of patients annoyed / disturbed and bed position Patient privacy and bay size Staff levels of annoyance and interference Level of noise annoyance / interference by staff gender Level of noise annoyance / interference by staff age Level of noise annoyance / interference by time worked on the ward Level of noise annoyance / interference by time worked at the hospital Average day time levels by building age for all patient accommodation Average night time levels by building age for all patient accommodation
212 213 213 218 218 220 221 222 223 223 224 224 225 226 226 227 228 228 229 230 231 232 232 232 234 241 241
xii
Acoustic Design for Inpatient Facilities in Hospitals List of Tables ___________________________________________________________________________________________________________________
LIST OF TABLES Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 2.7 Table 2.8 Table 2.9 Table 3.1 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 7.1 Table 7.2 Table 7.3 Table 7.4 Table 7.5 Table 7.6 Table 8.1 Table 8.2 Table 8.3 Table 9.1 Table 9.2 Table 9.3 Table 9.4 Table 9.5 Table 9.6 Table 9.7 Example of criteria for intrusive noise from external sources, HTM 08-01 (The Stationary Office, 2008) Example of criteria for internal noise from mechanical and electrical services, HTM 08-01 (The Stationary Office, 2008) Example of sound insulation parameters of rooms, HTM 08-01 (The Stationary Office, 2008) Example of sound insulation ratings (dB, DnT,w) to be achieved on site, HTM 08-01 (The Stationary Office, 2008) Extract from Chapter 8 Checklists, HTM 08-01 (The Stationary Office, 2008) Standards and guidelines for healthcare design in Europe (from Bergman and Janssen, 2008) Acoustic parameters (from Bergman and Janssen, 2008) World Health Organisation guidelines for hospital wards and treatment rooms Recommended ceiling characteristics for hospital room types, HTM 60 (NHS Estates, 2005) Measurement data from studies cited in Section 3.2 Measurement location and time interval Average LAeq measured for 24 hour, day and night time periods at each location Average and maximum noise levels of identified events in single room A Reverberation times measured in different ward accommodation Ambient noise levels measured in unoccupied patient accommodation Medical ward - measurement locations, time periods and patient gender Average LAeq measured for 24 hour, day and night time periods at each location Examples of noise events at the nurse station Examples of noise sources and levels on the medical ward Measurement location, time periods and patient gender type Average LAeq for 24 hour, day and night time periods at each location Average LAeq measured during the day and night time pre and post the ceiling change Reverberation times for both the unoccupied and occupied bay pre ceiling change Reverberation times estimates for both the unoccupied and occupied bay post ceiling change Ward D8 - measurement locations, time periods and patient gender Average LAeq measured for 24 hour, day and night time periods at each location Examples of noise events at the nurse station Examples of noise events in the multi-bed bays Measurement location, time interval and patient type Average LAeq measured for 24 hour, day and night time periods at each location Examples of noise events at the nurse station xiii 7 7 8 8 10 11 11 12 13 14 60 61 71 74 74 100 100 103 108 112 112 138 1 43 143 151 151 154 158 166 166 169
Acoustic Design for Inpatient Facilities in Hospitals List of Tables ___________________________________________________________________________________________________________________
Table 9.8 Table 9.9 Table 10.1 Table 10.2 Table 10.3 Table 10.4 Table 10.5 Table 10.6 Table 10.7 Table 10.8 Table 11.1 Table 12.1
Measurement location, time interval and patient gender Average LAeq measured for 24 hour, day and night time periods at each location Comparisons between measured and MLE-RT20 values Numbers of triggers recorded during the simulations SLM 1 with curtains open SLM 1 with curtains drawn SLM 2 with curtains open SLM 2 with curtains drawn Locations with data available for MLE-RT20 estimation Day time data shown in 4 hour windows; overall mean estimate with 95% confidence intervals Summary of the objective and subjective data collected during the study World Health Organisation guidelines for hospital wards and treatment rooms
181 181 205 206 207 207 207 207 208 209 217 248
xiv
Acoustic Design for Inpatient Facilities in Hospitals Acknowledgements __________________________________________________________________________________________________________
Acknowledgements
This piece of work has been funded as a Case Award by the Engineering and Physical Sciences Research Council (EPSRC) and by Arup Global Healthcare. Firstly, I would like to thank the members of the Redevelopment Team at Great Ormond Street Childrens Hospital, and those members of the Estates Teams at Bedford Hospital and Addenbrookes Hospital who facilitated this research. Without their time and support, this research would not have been possible. I would also like to thank the study ward managers for their cooperation, and all the ward staff and patients who have taken time to complete the study questionnaires. This work would also have not been possible without the unending support, patience and kindness of my supervisor Professor Bridget Shield. I would like to thank you for having faith in me and allowing me finally to fulfil my true potential. I would also like to express my gratitude to my supervisor Rosemary Glanville for sharing her extensive knowledge and experience of healthcare buildings, and for providing me with invaluable guidance and support. I would like to extend my thanks to Russell Richardson at RBA Acoustics who gave me the initial opportunity to gain some valuable experience in the field of acoustics, and to Peter Attwood of Acoustic Associates (Sussex), without whom I would never have embarked on this journey. Finally, a thank you to all my friends and family for believing in me and for always being there.
xv
Acoustic Design for Inpatient Facilities in Hospitals Glossary and definitions ___________________________________________________________________________________________________________________
Acoustic glossary and definitions

Sound and noise Sound refers to the harmonic pressure variations that we hear in air and is an important part of our everyday world. Too much sound can be annoying, even dangerous. Noise usually refers to unwanted sound. General environmental and building noise consists of sound that is composed of many different frequencies. The decibel and sound pressure The decibel (dB) is the main measurement unit in acoustics. It can be the measure of the magnitude of sound, changes in sound level and a measure of sound insulation. The decibel is not an absolute unit, but the ratio of two levels expressed in logarithmic form. A 1 dB increase in level is un-noticeable in everyday life. A 3 dB increase would be barely perceptible (even though it is actually a doubling of sound energy). A 10 dB change in level is perceived as the doubling in loudness.
The pressure fluctuations caused by sound waves in air are called sound pressure. The lowest sound pressure level which can be heard is 0 dB, known as the threshold of hearing. The highest level which can be tolerated is called the pain threshold and is around 120 dB. The response of the human ear and A-Weighting The response of the human ear is dependent upon the frequency characteristics of the sound. The ear is not equally sensitive to sound at all frequencies, being less sensitive at low and very high frequencies, with peak response around 2500 to 3000 Hz The vast majority of noise measurements made are in A-weighted decibels (dBA). The A-weighting is an electronic frequency weighting network which attempts to build the human response to different frequencies into the reading indicated by a sound level meter, so that it will relate to the loudness of the noise. The measured readings are denoted with either an A as in 90 dBA or a subscript in the case of LAeq, LA90, and LAmax. The subscript Z denotes that the measured sound level is unweighted. A- Weighted Equivalent Sound Pressure Level (LAeq) Most measured noise is not steady, but fluctuates significantly in level over a short period of time. It is not easy to find a measure which accurately quantifies what is heard with a single number. The LAeq is the A-weighted equivalent continuous sound pressure level. It is an average of the total amount of sound energy measured over a specified time period (commonly a 1 hour period). If noise is measured for discrete periods of time, the overall LAeq,T can be calculated using the following equation: LAeq,T = 10log[(t1.10L1/10 + t2.10L2/10 + t3.10L3/10 + . TN.10LN/10)/T] Where t1 is the time at noise level L1 dBA t2 is the time at noise level L2 dBA t3 is the time at noise level L3 dBA, etc. .and T is the time over which the value is required.
LAmax, LAmin and Statistical Parameters The LAmax is the A-weighted maximum sound pressure level during a measurement period. The LAmin is the A-weighted minimum sound pressure level during a measurement period. Statistical parameters (sometimes called noise percentile levels) are the sound pressure levels exceeded for a certain percentage of the measurement period. LA10 and LA90 are the most commonly used, where LA10 is the level exceeded for 10% of the time and LA90 is the level exceeded for 90% of the time. LA90 is used to represent background noise levels. Fast and Slow Time Weightings Time weightings determine how quickly a sound level meter (SLM) responds to changes in the sound pressure level. Most measurements are now made with a fast weighting selected, unless otherwise specified by the relevant standard. When sampling is set to fast the SLM is sampling over a number of 0.125 s periods and all parameters are calculated from these measurements.
xvi
When sampling is set to slow the SLM is sampling over a number of 1 s periods and all parameters are calculated from these measurements. As the SLM is responding more slowly when set to slow, impulsive measurements such as LAmax would give lower sound pressure level readings (as a rule of thumb 5dB less could be expected) and LAmin values would be higher than expected. The difference between LAeq measurements would not be so significant (perhaps 1dB depending on the specification of the SLM). Frequency (Hz) Frequency is defined as the number of oscillations per second and is measured in Hertz (Hz). A healthy young person can hear frequencies from 20 Hz to 20,000 Hz (the lower the value the lower the pitch). Loudness Loudness is a measure of the subjective impression of sound Reverberation Time (RT) The reverberation time is defined as the time it takes for sound to decrease by 60 dB. Long reverberation times usually exist in spaces with hard surfaces and minimal sound absorption (like curtains, carpets and soft furnishings). Sounds within reverberant rooms may seem hard-edged and even echo-y. With high levels of background noise present, the nature of a reverberant room is such that speech intelligibility may be affected, causing voices to be raised. To determine the length of the reverberation time, different parts of the reverberation curve are used. This is illustrated by the graph below:
Sound pressure level (dB)
Time (secs) To calculate the Early Decay Time (EDT), the time taken for sound to decrease by 10dB is used and multiplied by a factor of six. The EDT is known as the early reverberation time and is considered to better reflect how we perceive the reverberance in the room. The descriptors T20 and T30 are usually called late reverberation times as they measure the later parts of the curve. When calculating T20, the time taken for the sound to decay by 20dB is used and is trebled to give the reverberation time. It should be noted that the evaluation does not start until after the sound level has already fallen by 5dB. When calculating T30, the time taken for the sound to decay by 30 dB is used and is doubled to give the reverberation time. As with T20 calculations, the evaluation does not start until after the sound level has already fallen by 5 dB. If the reverberation curve is straight, the EDT, T20, T30, will all produce the same value. However, the reverberation curve is usually not straight (shown by a dashed line on the graph), which means that the descriptors will differ. The room volume, room shape, and the amount of sound absorbing material present all have an effect on the reverberation time. Sound Absorption Coefficient The sound absorbing properties of a material are expressed by the sound absorption coefficient, , as a function of the frequency. ranges from 0 (total reflection) to 1.00 (total absorption).
xvii
Sound Absorption Class In accordance with international standard EN ISO 11654, the absorption classes are designated A-E, where absorption class A has the highest sound absorption. The graph below illustrates the difference in properties of each class.
Sound absorption table
Absorption coefficient ()
Frequency (Hz) 1. Absorption class A 2. Absorption class B 3. Absorption class C 4. Absorption class D 5. Absorption class E 6. Unclassified Reflective building materials have low absorption coefficients (), for example plastered walls have a value as low as =0.02. Sound absorbing materials have a relatively high absorption coefficient, for example some ceiling tiles can have value as high as =0.95 at some frequencies.
xviii
Acoustic Design for Inpatient Facilities in Hospitals Abstract ___________________________________________________________________________________________________________________
Abstract
There is an increasing body of research into the acoustic environment of hospitals which provides evidence of the detrimental effects of noise on the well being and comfort of patients and on staff, and of a significant rise in hospital noise levels in recent years. Much of this evidence has focused on specific areas of healthcare such as critical care and operating theatres, with comparatively few studies carried out within general inpatient wards and in UK hospitals. The current study aims to investigate, through objective and subjective surveys, the noise climate and acoustic design within general inpatient facilities in the UK, and their influence on the acoustic comfort of patients and staff. Noise and acoustic surveys have been carried out in six inpatient wards in three major UK hospitals, with corresponding questionnaire surveys of staff and patients. Noise measurement data has been analysed to build up a comprehensive understanding of the contributing factors to noise in both single room and multi-bed patient accommodation, and at the main ward nurse stations. Comparisons are made between patient accommodation types; medical and surgical wards; building construction types; ward layouts; and finishes. The potential impact of the design for infection control on acoustic comfort is also examined. Patient and staff perceptions of noise are investigated, with the identification of the most annoying and disturbing noise sources. Attitudes to noise and factors such as age, length of stay, bed location and length of service are considered where appropriate. The problem of hospital noise in inpatient wards is found to be very complex in nature, with many different factors affecting the noise climate. The study concludes that a multi-faceted approach is required if any significant improvement is to be achieved. This should be centred on three main areas (i) optimising the acoustic design of the ward, (ii) minimising the disturbance caused by equipment in use on the ward and (iii) modifying the behaviour of those on the ward. Discussion of these areas is provided and potential areas of noise control investigated. The observational culture of nursing in UK hospitals is also considered in relation to ward design. Two further pieces of work have been carried out in addition to the main study. The first investigates the effects of changing a non acoustic suspended ceiling for one with good acoustic properties. Noise levels and reverberation times prior to and after this change are measured and improvements found. The second piece seeks to validate an estimation method for reverberation times in occupied spaces. Using noise data captured during the main study, the estimated data is found to demonstrate similar accuracy to standard measurement techniques, and as such the method could potentially be used to provide reverberation time estimates in occupied areas where real time measurements are not practical or possible.
Acoustic Design for Inpatient Facilities in Hospitals Introduction ___________________________________________________________________________________________________________________
Introduction
1.1. Background
Concern about noise in healthcare is not a recent phenomenon. In her book Notes on Nursing (Nightingale, 1860), first published over 150 years ago, Florence Nightingale devoted an entire chapter to noise and its negative effects. She warns Unnecessary noise, then, is the most cruel absence of care which can be inflicted on the sick and well. Research into the subject area of hospital noise is surprisingly diverse, with Ulrich et al (2004) citing no less than 130 studies which focus on the subject. A clear trend of rising hospital noise since the 1960s has been identified by Busch-Vishniac et al (2005), with average increase in noise levels per year of 0.38 dB during the day and 0.42 dB at night. This increase in noise levels appears to be universal in nature, with many international studies showing similar trends. There is also increasing evidence of the detrimental effects of noise on patient wellbeing and on staff, with noise induced stress being linked to burnout of critical care nurses (Topf and Dillion, 1988). Studies have also linked noise levels to patient recovery rates (Fife and Rappaport, 1976) and associated improvements in acoustic design with reductions in patient re-admission rates (Hagerman et al, 2005). It should be noted, however, that much of the evidence concerning the impact of noise on patients and staff is focused on a small number of frequently cited studies. A review of the literature has found that research into hospital noise has tended to concentrate on busier areas within hospitals, such as critical care units, intensive care units and operating theatres, with limited research carried out in inpatient wards. Patients in inpatient wards are generally recovering from either a severe infection or from surgery, and so require restful conditions that are beneficial to their recovery. Research on the impact of noise in this area is felt to be of at least equal importance to research into noise in critical care areas.
1.2. Aims and objectives

The aim of this study is to provide a comprehensive insight into the noise climate in inpatient hospital wards. The research involves both objective measurements which help build up an understanding of noise levels and the sources of high level noise; and questionnaire surveys of staff and patients on the wards, which explore their perceptions of noise. Six inpatient wards in three major hospitals are the subject of the study, with a mixture of medical and surgical wards and types of building stock. Hospital buildings of differing ages, construction type, ward design, and finishes are all considered. The findings of the study will help to inform the decision making process in both new building design and the refurbishment of existing wards, and to provide general advice around the choice of ward
equipment and technological systems to be commissioned. The findings also enable a critical review of current standards to be carried out to ascertain their applicability in relation to occupied buildings.
1.3. Overview
This thesis consists of 12 chapters. Chapter 2 discusses relevant guidance and standards for healthcare buildings and is followed by an extensive literature review in Chapters 3 and 4. The findings of these three chapters inform the study methodology which is discussed in Chapter 5 and is trialled during the pilot study detailed in Chapter 6. The main study locations and results of the subsequent objective and subjective studies are discussed in Chapters 7 and 9, with a further overall analysis of these provided in Chapter 11. A full discussion of the study findings in relation to noise control are provided in Chapter 12. Chapters 8 and 10 are concerned with the results of a ceiling change carried out in one particular ward, and the validation and use of a reverberation time estimation method for occupied spaces. In order to understand the role of acoustic guidance documents and standards in healthcare buildings, a review is carried out in Chapter 2. Specific guidance on the acoustic design of healthcare buildings was found to exist as early as 1966, and the changes to standards over time are considered. Comparisons with European standards are made; the World Health Organisation guidelines are assessed; and specific design criteria in relation to control of infection on hospital wards is discussed. In relation to this study, which is concerned with occupied buildings, the relevance of much of the documentation in this chapter is found to be minimal. Chapters 3 and 4 aim, through critical evaluation, to build up a thorough understanding of previous research carried out in the field of hospital noise and acoustic design. For the purposes of clarity these chapters are divided into a number of categories: noise measurement studies; sleep studies; the effects of behaviour modification on hospital noise; the effects of room acoustic design modifications on hospital noise; the effects of noise on healthcare staff; and the effects of noise on patients. Within each category a number of papers are summarised and some critical appraisal is made of methods used where appropriate. Discussion of the study findings, design limitations and areas found to be lacking in research is provided. Conclusions drawn from the literature review in Chapters 3 and 4 were seminal in informing the design of the study. Chapter 5 outlines the aims and objectives of this study and provides further detail of the objective and subjective survey methods used. The preliminary work involved in obtaining ethical approval and the necessary permissions to carry out the study within occupied ward environments are also discussed. To trial the study methodology, a pilot study was carried out in a post surgical inpatient ward at Great Ormond Street Childrens Hospital, London. Chapter 6 discusses the considerations required when working in a healthcare environment, including the appropriate location of measurement equipment 3
and the distribution of questionnaires. Detailed results of the objective noise level measurements are presented and the subjective perceptions of the noise climate of both patients and staff examined. The main study was undertaken in two inpatient wards at Bedford Hospital and three inpatient wards at Addenbrookes Hospital, Cambridge. The results from these study wards are reported in Chapters 7 and 9 respectively. Details of the objective measurements made in each ward are discussed and the subjective perceptions of staff and patients are compared. Further analysis of these results are presented in Chapter 11, which considers factors affecting noise levels, and staff and patient perceptions. Two further pieces of work were carried out in addition to the main study. Planned refurbishment works were due to take place in the medical ward at Bedford Hospital, which had been the subject of objective and subjective surveys. The works were scheduled to start several weeks after the end of the study data collection, and this enabled a ceiling intervention study to be carried out in which non acoustic ceiling tiles were replaced by tiles with good acoustic properties. Noise levels and reverberation times were investigated prior to and after this change and the results are reported in Chapter 8. The second piece of work was the validation of a reverberation time (RT) estimation method. This method, known as the Maximum Likelihood Method, was developed by the University of Salford for use with recorded speech or music. Since making RT measurements in occupied hospital wards is not practical using standard methods, an alternative method to estimate the RTs of occupied wards was considered to be extremely useful. Validation of this method was carried out using data captured from the study wards and is discussed in detail in Chapter 10. Chapter 12 summarises the findings of this study in relation to noise control in inpatient hospital wards; the applicability of standards to occupied hospital buildings; and the usefulness of particular acoustic measurement parameters in the reporting of hospital noise. Study conclusions and recommendations for further work are discussed in the final chapter of this thesis.
Acoustic Design for Inpatient Facilities in Hospitals Acoustic standards and guidance ___________________________________________________________________________________________________________________
2.
Acoustic standards and guidance
2.1. Introduction
This chapter reviews the history of guidance relating to acoustic design of UK healthcare facilities up to the present time. The differences between UK and European guidance are explored, and the relevant section of the World Health Organisation Guidelines for Community Noise summarised. The final section presents further discussion on other aspects affecting acoustic design in healthcare.
2.2. UK design guidance

In the UK, design of healthcare buildings is governed by guidance published by the Estates and Facilities Division within the Department of Health. Essentially, the guidance falls into the following two categories: Health Building Notes (HBN) which provide advice to project teams designing and planning new buildings and refurbishing existing buildings. Health Technical Memoranda (HTM) which provide estates and facilities professionals with guidance on the design, installation and running of specialised building service systems. It appears that some guidance was available as early as 1966, with information on noise control provided in the Ministry of Health Hospital Design Note 4: Noise Control (Her Majestys Stationery Office, 1966). As with more recent guidance this design note provided general advice on plant noise, the performance of internal partitions and external wall construction. More surprisingly, suggestions for effective noise control in occupied wards were given for a variety of items such as quiet curtain tracks; hospital trolleys with good quality rubber tyred wheels; the use of steel sinks coated with rubber on the underside to minimise noise; and the suggestion that TVs in multi-bed wards should be wired through patients headphones only. For the purposes of this review, the three most recent guidance documents affecting acoustic design of healthcare buildings are examined in further detail. These are HTM 2045 Acoustics: Design Considerations (NHS Estates, 1996), HTM 56 Partitions (NHS Estates, 1997) and HTM 08-01 Acoustics (The Stationary Office, 2008). HTM 2045 Acoustics: Design Considerations (NHS Estates, 1996) contained partition performance requirements as well as advice on many other aspects of acoustic design including mechanical services noise, impact noise, vibration, faade sound insulation, and reverberation times. Not only did the guidance provide specification of acoustic design criteria, it also provided information on the sources of noise and provision of noise control, and an eleven page section on the principles of 5
acoustics. At a total 53 pages in length, it was considered by some as rather impractical (Popplewell, 2008). HTM 56 Partitions (NHS Estates, 1997) provided general design guidance on the construction and performance of internal partitions, with some specific performance criteria to ensure adequate privacy between rooms in a healthcare setting. The partition performance criteria set out in this guidance document were less stringent than those provided in HTM 2045 as they were specified in terms of sound reduction, with no implied requirements for flanking control or need to pay attention to junction detailing. Confusingly, HTM 2045 was intended to take precedence over HTM 56, although it was published a year earlier. However, in the construction industry both guidance documents were considered to be current. Many healthcare buildings were built to the less comprehensive and less stringent criteria set out in HTM 56, as this was considered to be a less expensive option. To end this confusion HTM 56 was finally revised in 2005. All guidance on acoustic performance was removed and the standard was re-written to refer to HTM 2045, which was itself superseded in 2008 by HTM 08-01, and is discussed in the following section.
2.2.1. HTM 08-01

The latest acoustic design guidance HTM 08-01 Acoustics was published in 2008 (The Stationary Office, 2008) and superseded HTM 2045. Popplewell (2008) explained the drivers for change behind the latest standard and discusses some of the difficulties associated with the implementation of HTM 2045. A number of examples were provided by the author to illustrate these difficulties, following his own experiences of working with several large Private Finance Initiative healthcare projects, and are shown below. The reverberation time criterion was felt to be unrealistic in areas where absorptive finishes were not appropriate for the reasons of infection control, for example in an operating theatre. The building services noise criterion was felt to be impractical due to the need for dedicated air handling systems in specific areas. The impact sound performance requirements of HTM 2045 could not be guaranteed by contractors. A number of aims of the new standard were suggested by Popplewell (2008), with the emphasis on creating a simple, practical and effective standard. HTM 08-01 was designed to: simplify and shorten the text of HTM 2045 make sure recommendations were practical and appropriate 6
clarify aspects which were considered to be open to misinterpretation in the previous standard remove specific limits where they either imposed unnecessary costs or where they were unachievable from a practical perspective allow for the incorporation of new technologies As published, the HTM 08-01 is indeed a more streamlined version of the previous standard. It recommends acoustic criteria for both noise intrusion and mechanical services noise. Footfall, plant vibration and internal sound insulation requirements are also considered. Examples are shown in Tables 2.1 & 2.2. Table 2.1 Example of criteria for intrusive noise from external sources, HTM 08-01 (The Stationary Office, 2008)
Table 2.2 Example of criteria for internal noise from mechanical and electrical services, HTM 08-01 (The Stationary Office, 2008)
Several matrices are provided to simplify the calculation of the internal sound insulation requirements. These matrices take into consideration the privacy requirements and the potential noise generated within each room type. Tables 2.3 and 2.4 provide examples.
Table 2.3 Example of sound insulation parameters of rooms, HTM 08-01 (The Stationary Office, 2008)
Table 2.4 Example of sound insulation ratings (dB, DnT,w) to be achieved on site, HTM 08-01 (The Stationary Office, 2008)
One of the main differences between this and the previous standard is the lack of specific guidance on room acoustic design. HTM 2045 specified reverberation times for rooms of different functions with volumes less than 1000 m , as shown in Figure 2.1, where room types A, B and C are consulting rooms, multi bed wards and bathrooms respectively.
3
Figure 2.1 Recommended RTs for different room functions and volumes, HTM 2045 (NHS Estates, 1996) The latest standard is much more general, providing guidance regarding the amount of acoustic absorbency used, but does not include precise guidance on room reverberation times, advising that a reverberation-time criterion should be agreed depending on the specific requirements for use of the space. The guidance acknowledges that the use of acoustically absorbent materials can have a dramatic effect on the acoustic comfort in a room and is particularly necessary where speech intelligibility is a requirement. It states that sound absorbent treatment should be used in all areas, including corridors, and recognises that there may be issues surrounding the use of sound absorbent materials where cleaning, Control of Infection, patient safety, clinical and maintenance requirements allow. The guidance suggests that the most appropriate area for acoustically absorbent material should be a ceiling, with the minimum absorption area equivalent to 80% of the area of the floor when using a Class C absorber (see Glossary). Occupied hospital buildings It is important to note that HTM 08-01 is applicable to a newly built or refurbished unoccupied healthcare facility, where specific systems can be measured alone. Most of the values specified however cannot be compared with measurements made in an occupied building, where systems cannot be isolated. Nevertheless, there is some general guidance which is relevant for occupied buildings and is listed below: Medical Equipment The standard states that ideally it should be chosen so that it does not adversely affect the use of the surrounding space and that quiet equipment should be chosen.
Nurse-call Systems Guidance is given regarding the choice of nurse-call systems to the effect that nurse-call systems can disrupt sleep; therefore non-audible systems should be considered; especially at night. Also it is suggested that audible alarms intended for staff should be located such that they cause minimum disruption to patients. The standard also contains a checklist for the most important acoustic issues. Some internal fit-out equipment and management issues listed are relevant when considering noise levels in occupied wards, and the relevant extract of the checklist is shown in Table 2.5. Table 2.5 Extract from Chapter 8 Checklists, HTM 08-01 (The Stationary Office, 2008)
2.3. European healthcare design guidance

Bergmark and Janssen (2008) compiled an overview of international standards focussing on those portions which dealt with room acoustics within healthcare buildings. Comparison was made between standards from Sweden, Germany, Finland, Norway, UK, Denmark and The Netherlands. The UK was found to be the only country with a specific healthcare guidance document. Sweden, Finland, Norway and Denmark have healthcare specific sections within wider guidelines, whereas Germany and the Netherlands use general workplace criteria. The study noted that the UK standard, HTM 08-01, was the only standard which did not provide specific guidance on either room reverberation times or speech intelligibility. It was also noted that, with the exception of UK and Denmark, comfort classes were mentioned in each of the standards. These comfort classes aim to define the level of acoustic comfort for the user by using certain acoustic parameters such as reverberation times and levels of building services noise. Table 2.6 provides a summary of the seven European standards and guidelines reviewed.
10
Table 2.6 Standards and guidelines for healthcare design in Europe (Bergman and Janssen, 2008)
Table 2.7 summarises the acoustic parameters specified by the standards for the purposes of acoustic room comfort. Table 2.7 Acoustic parameters (Bergman and Janssen, 2008)
2.4. World Health Organisation guidelines

The most recent edition of the World Health Organisation (WHO) Guidelines for Community Noise was published in 1999 (Berglund et al, 1999). The guidelines seek to consolidate actual scientific knowledge on the health impacts of community noise and to provide guidance to environmental health authorities and professionals trying to protect people from the harmful effects of noise in nonindustrial environments.
11
In relation to noise in hospitals, the guidelines state that the critical effects of noise are on sleep disturbance, annoyance and the communication interface, including interference with warning signals. Values of LAeq are provided for day time, which is listed as 16 hours from 07.00 23.00. LAeq and LAmax values are provided for night time, which is listed as eight hours from 23.00 07.00, with LAmax values measured on a fast setting. A summary of the guidelines is shown in Table 2.8. Table 2.8 World Health Organisation guidelines for hospital wards and treatment rooms
Specific Environment Critical Health Effects LAeq (dB) Time Base (Hours) LAmax (dB)
Hospital, ward rooms, indoors Hospital, ward rooms, indoors Hospital, treatment rooms, indoors
Sleep disturbance
30
Night time (8 hours)
40
Sleep disturbance
30
Day time and evenings (16 hours)
Interference with rest and recovery
As low as possible
The guidelines also suggest since patients have less ability to cope with stress that the LAeq level should not exceed 35 dB in most rooms where patients are being treated or observed.
2.5. Control of Infection

In the UK, Healthcare-Associated Infection (HCAI) has become an increasingly high profile issue over recent years. The result has been increased pressure on UK hospitals to clean more frequently, more thoroughly and with stronger cleaning agents. Several guidance documents have been produced which provide advice on the use of suitable materials to withstand the new cleaning regimes. These materials include flooring, wall coverings and ceiling finishes, the most relevant to the current study being the use of acoustic ceiling tiles. The relevant paragraphs of each document are explored in the sections below. Control of Infection teams have been set up in all hospitals to make general decisions in relation to hospital HCAI policies. These teams interpret the guidance documents in different ways, some more vigorously than others.
2.5.1. HTM 60
HTM 60 (NHS Estates, 2005) deals with all aspects of suspended ceilings, including fire performance, ceiling tile composition, wind loading, and details of grid construction. Only two sections are of relevance in terms of HCAI and these detail the physical characteristics of the ceiling tiles to be installed, and provide advice on hygiene and cleaning.
12
The guidance divides room types into six different categories and lists suitable ceiling finishes to be used for each. For example, Category 1 applies to operating theatres; Category 4 applies to multi-bed bays or single patient rooms; and Category 5 to storerooms. Table 2.9 shows that within a Category 4 room, all ceiling types may be used, whereas in a Category 1 room only smooth, imperforate, jointless ceilings are advised. Table 2.9 Recommended ceiling characteristics for hospital room types, HTM 60 (NHS Estates, 2005)
In the section entitled Hygiene and cleaning, the guidance discusses a new model cleaning contract for hospitals which, it states, has three key aspects: 1. The National Standards of Cleanliness. This document discusses possible measures for HCAI cleaning and disinfection. 2. NHS Cleaning Manual. This manual sets out best practice methods for cleaning. 3. Cleaning frequencies. These should be determined to address the element of risk identified within the National Standards of Cleanliness and should take into account any further advice and guidance in the model cleaning contract and the NHS Cleaning Manual. At the time of writing, the NHS Cleaning Manual is no longer readily available, even though HTM 60 is still current and has not been amended to reflect this.
2.5.2. National Standards of Cleanliness for the NHS

There are several relevant sections in the National Standards of Cleanliness for the NHS (NHS Estates, 2001) which list the cleaning requirements and cleaning frequencies for walls, skirting boards and ceilings. The guidance states that internal and external walls and ceilings are to be free of dust, grit, lint, soil, film and cobwebs and that walls and ceilings are free of marks caused by furniture, equipment or staff. Further information can be seen in terms of cleaning time scales for walls, skirting boards and ceilings. Table 2.10 provides information on cleaning frequencies by area type.
13
Table 2.10 Cleaning frequencies in terms of area type priority Priority A B C D Frequency Constant, cleaning critical Frequent, cleaning important and requires maintaining Regular, on a less frequent scheduled basis and as required in between Infrequent, or on a scheduled or project basis Time Frame for Rectifying Problems Immediate 0-48 Hours 2-7days 1-4 weeks
Priority A applies to operating theatres, ICU and similar units and Protective Isolation Areas. Priority B applies to sterile supply areas, A&E, pharmacy, general wards and daily activity areas, rehabilitation areas, residential accommodation, pathology, kitchens, outpatients clinics, treatment and procedure rooms, cafeteria and public thoroughfare. Priority C applies to general pharmacy, laboratories, mortuary, medical imaging, waiting rooms and administrative areas. Priority D applies to non-sterile supply areas, record archives, engineering workshops, plant rooms and external surrounds.
2.5.3. HFN 30 Infection Control in the built environment

HFN 30, Infection Control in the Built Environment Design and Planning (NHS Estates, 2002) states that if the burden of healthcare-associated infection is to be reduced, it is imperative that architects, designers, and builders be partners with healthcare staff and infection control teams when planning new facilities or renovating older buildings. In relation to ceilings, the guidance stresses the importance of high quality finishes and recommends that ceilings with smooth, hard, impervious surfaces are installed in theatres and isolation rooms. The guidance warns that during maintenance work, suspended ceilings can allow dust to fall onto the area below and therefore this type of ceiling should therefore be avoided in isolation rooms, operating theatres and treatment rooms. There is no mention of ceiling types for general inpatient wards.
2.6. Discussion
The latest UK design guidance has been simplified to be as practical as possible. Popplewell (2008) mentions that in particular the internal sound insulation matrices have been well received by contractors who have found them simple to use and understand. Design guidance of this type aims to go some way to minimising external noise break-in, plant noise, and noise transmission between rooms and when put into practice appears to be successful in these areas, as shown by Boulter (2007). 14
The relevance of the HTM guidance to the current study is fairly limited, as the study is investigating noise levels in occupied wards. Only guidance regarding the use and choice of a number of internal systems has some application. The omission of specific guidance relating to reverberation time values or target values for speech intelligibility potentially provides designers and contractors with a get out option, which is simply to ignore these criteria completely. This could have a negative impact on noise levels and acoustic comfort of wards. Comparison of the UK guidance with six different European design guidelines indicates that it is considered important by those countries to specify at least one measure of acoustic comfort. This again suggests that this omission in the latest HTM could be detrimental to the acoustic environment in UK hospitals. The noise levels suggested by the World Health Organisation are relevant in terms of occupied wards. It would, however, seem that the validity of these levels is questionable as all noise studies referenced (see Chapter 3) have found levels to be above the WHO guidelines in general. Guidance documents are available which provide advice on the types of ceiling finishes required and the cleaning frequencies. The documents are open to interpretation and therefore may be interpreted more or less stringently at different locations. Some inconsistencies do exist, with one major guidance document referenced, no longer available. With regards to the type of acoustic materials for use in areas where there are concerns about HCAI, no clear design guidelines appear to be available. Acoustic ceiling tiles which will withstand
bleaches and high pressure washing are readily available on the market. It is important that not only are contractors and designers made aware of these products, but also the hospital Control of Infection teams, who have an increasing influence on the internal finishes used. Without this knowledge the use of non acoustic ceiling tiles may become more prevalent, having a detrimental effect on the acoustic comfort of a room.
2.7. Conclusions
It has been shown that the UK guidance is on the whole less stringent than guidance in other European countries or the WHO guidelines. In the current study objective levels measured are compared with current guidelines where appropriate. The following two chapters review previous research on hospital noise and its effects on staff and patients. Many of the studies reviewed show that current guideline values are exceeded.
15
Acoustic Design for Inpatient Facilities in Hospitals Previous research on hospital noise ___________________________________________________________________________________________________________________
3. Previous research on hospital noise

3.1. Introduction
There has been a significant body of research into various aspects of hospital noise and acoustics in the past 50 years. This chapter aims, through critical evaluation, to build up a thorough
understanding of previous research carried out in the field of hospital noise and acoustic design. For the purposes of clarity this review is divided into four categories: noise measurement studies; sleep studies; the effects of behaviour modification on hospital noise; and the effects of room acoustic design modifications on hospital noise. Within each category a number of papers are summarised and some critical appraisal is made of methods used where appropriate. Discussion of the study findings and design limitations are provided at the end of each section.
3.2. Noise measurement studies

The hospital environment is an extremely complex one, consisting of many different areas, including Accident and Emergency (A&E) departments, operating theatres, intensive care units (ICUs) and general inpatient wards. In each area, different activities are taking place with patients requiring different levels of care. Much of the available literature investigating noise levels in hospitals concentrates on specific measurement locations. It appears that many of these locations were chosen because of the perception that they were more noisy. This makes it difficult to build up an overall picture of the noise climate across all areas of the hospital. A large review of previous studies dating from 1960 was carried out by Busch-Vishinac et al in 2005. In order to identify whether a trend in hospital noise exists, the authors compiled data from all comparable studies post 1960 which listed LAeq noise measurement values. Although it was acknowledged that there were some discrepancies in the data (for example, no indication of sampling rates or the period of time averaging), the study yielded some interesting results. The findings were three-fold: i. Not one single study showed a hospital which complied with the WHO guidelines for hospital noise, raising the question of the validity of these particular guidelines. ii. The study showed less variation in results than was expected. This was surprising given that the data was gathered from widely differing sources different types of medical units and hospitals situated in a number of different countries throughout the world. It led the authors to conclude that the problem of hospital noise is a universal one.
16
iii.
A clear trend was shown for rising hospital noise since 1960. The data showed an increase of 0.38 dB per year for day time levels and 0.42 dB per year for night time levels, with a rise in measured LAeq values from 57 dB in 1960 to 72 dB in 2005 during day time hours, and from 42 dB in 1960 to 60 dB in 2005 during night time hours. Figures 3.1 and 3.2, reproduced from Busch-Vishinac et al (2005), show the A-weighted equivalent day and night sound pressure levels as a function of year of study publication. The error bars indicate that data was given as a range spanned by the error bars.
Figure 3.1
LAeq values measured in hospitals during day time hours as a function of the year of study publication. (Busch-Vishinac et al, 2005)
Figure 3.2
LAeq values measured in hospitals during night time hours as a function of the year of study publication. (Busch-Vishinac et al, 2005)
The remainder of this section focuses on the more recent relevant studies of hospital noise, which are summarised in Table 3.1.
17
In 1996, McLaughlin et al considered noise levels in a cardiac ICU at the Royal Group of Hospitals, Belfast, and concluded that they were consistently higher than those stipulated in the WHO guidelines. LAeq values for the measurement period were found to be above 60 dB at all times and LAmax values greater than 80 dB were measured as early as 5am. Two studies of noise in hospitals in the US were published in 1999 and 2001. Holmberg and Coon (1999) examined noise levels within adult and adolescent day rooms in a state psychiatric hospital in Indiana, and found levels to exceed those measured in other studies of medical, surgical and intensive care units. Noise levels in A&E departments in four Phoenix hospitals were measured by Buelow (2001), who concluded that levels were higher than those in which an individual can comfortably work. The levels were thought to approach or exceed those that can cause feelings of annoyance. Tsiou et al (2008), recorded sound levels during surgical procedures carried out in the operating theatres of nine Greek hospitals. Comparisons between sound levels measured during nonorthopaedic surgery and orthopaedic surgery were made. This extensive study showed orthopaedic surgery to be a particularly lengthy, noisy process and recommended that personnel make use of hearing protection and undergo regular audiometric tests. Sources of noise were identified and their sound levels noted.
3.2.1.
Limitations of noise measurement studies
Measuring noise within a healthcare setting introduces a number of challenges. Patients are in hospital because they require care; staff are busy and often working beyond their capacity. To take measurements in a non-intrusive manner without causing annoyance or suspicion is often difficult. The Hawthorne Effect One known issue when undertaking a measurement study in an occupied building, is the reaction of those people in the vicinity of the measurement equipment. If it becomes known that a study is being undertaken, people may react in a way that might affect the results. This phenomenon was identified by Henry A. Landsberger in 1955 when analyzing results from a set of experiments carried out from 1924 1932 at the Hawthorne Works, and has become known as the Hawthorne Effect (cited by Bailey and Timmons, 2005). Landsberger defined the Hawthorne effect as a short-term improvement caused by observing worker performance. The Hawthorne effect was thought by Bailey and Timmons (2005) to be significant in their study of noise levels in paediatric ICU in a large UK teaching hospital. It was noted that once
18
Table 3.1 Measurement data from studies cited in Section 3.2 Author
McLaughlin et al
Date
1996
Location
Cardiac ICU, Belfast
Measurement Period
24 hr
Measured Levels
LAeq > 60 dB, LAmax > 70 dB (for measurement period) , LAmax =100.9 dB
Holmberg and Coon
1999
State psychiatric hospital, US
36.5 hrs in total at different times of day
Mean (arithmetic) = 75.7 dBA, LAmax = 92.5 dB ; Sound peaks consistently between 85 and 90 dB
Buelow et al
2001
A&E departments at 4 hospitals in Phoenix, US
16.00 20.00 (4 hours)
LAeq = 69.1 dB, 70.1 dB, 71.1dB, 65.0 dB LAmax = 76.6 dB, 73.4 dB, 73.0 dB, 75.2 dB
Bailey and Timmons
2005
Paediatric ICU in a UK teaching hospital
24 hr
Loudest voices measured between 68 and 72 dB; General conversation measured between 50 and 65 dB; Equipment alarms measured between 65 and 83 dB
Busch-Vishinac et al
2005
5 locations within the Johns Hopkins Hospital
3 x 24 hr measurements at each location
LAeq : 50 60dB (PICU showing the highest noise levels)
Kracht et al
2007
38 operating theatres at the Johns Hopkins Hospital, Maryland, US
Orthopaedic surgery LAeq = 66 dB 24 hr Neurosurgery, urology, cardiology, gastrointestinal surgery LAeq : 62 - 65 dB Neurosurgery and orthopaedic surgery LAmax >100 dB for over 40% of the time, with peaks >120 dB
Orellana et al
2007
Adult A&E at the Johns Hopkins Hospital
24 hr
Triage Area LAeq : 65 73 dB General A&E LAeq : 61 69 dB
Pre-surgical LAeq : 61.1 - 78.2 dB, LA90 : 49.2 - 61.2 dBA, LAmax : 83.6 - 99.4 dB Tsiou et al 2008 Operating theatres at 9 Greek hospitals During 43 different procedures Surgical LAeq : 57.4 - 70.1 dBA, LA90 : 48.2 - 58.7, LAmax : 84.7 - 100 dB Post-surgical LAeq : 60.5 - 74.1 dBA, LA90 : 49.7 - 60.7, LAmax : 78.8 - 106 dB Connection / disconnection of gas supply responsible for loudest sound peak of 106 dB
members of staff became aware of the study, they changed their behaviour accordingly. The researchers found that the female staff lowered their voices and made an effort to keep noise levels down; whereas the male staff made deliberate attempts to make noise by shouting at the microphone and banging equipment (these impulsive noises were ignored in the results). It has been shown that it is possible to minimise the influence of the Hawthorne effect. For example, in the study of noise in a cardiac ICU by McLaughlin et al (1996), discussed in the previous section, the SLM was concealed in a dummy box which had time, temperature and humidity displays. It was perceived that these displays were sufficient to satisfy the curiosity of the staff regarding this new piece of equipment placed in their work environment. Similarly, Kracht et al (2007), who measured operating theatre noise at the Johns Hopkins Hospital in Baltimore, placed their sound level meter (SLM) so that it did not interfere with the operations. It was wrapped in a plastic bag to avoid contamination and this may have made the meter less conspicuous. It was observed that the staff were generally unaware that the meter was present and so did not attempt to control conversation levels or the playing of music during the operations. Acoustic inconsistencies and omissions Partially due to the restrictions that are inherent in working within an occupied ward, it was found that many of the studies reported measurements in ways that often prevent study comparison. It was also noted that many of the studies were undertaken by healthcare professionals with little or no knowledge of acoustics. In a number of studies certain acoustic criteria were not considered, or important key elements were omitted from the report. Further details of typical omissions and inconsistencies are given below: i. Calculation of the time averaged sound pressure level (SPL) Some of the studies appeared to calculate the average SPL recorded over a period of time arithmetically, rather than logarithmically. This method of time averaging yields a lower value and therefore levels may not be comparable with those studies that provide a logarithmically averaged LAeq,T value. ii. Equipment Sampling Rates Many studies do not state the sample rate settings used on the SLM. For some measurements this setting can make a difference and therefore studies which do not specify this cannot be used for comparison purposes.
20
iii.
Unspecified, linear or A-weighted measurements Measurements reported in some studies were either unspecified or listed as linear unweighted values. Without frequency band data, conversion of linear to A-weighted values for comparison with other studies is not possible.
iv.
Only minimum and maximum values reported Some studies did not list either background levels of the noise or time averaged noise levels, choosing only to list maximum and minimum values. Without time averaged measurements it is impossible to put maximum and minimum values into context
v.
The position of the sound level meter whilst undertaking measurements. Some studies were concerned with the noise levels experienced by the patient and positioned the sound level meter by the bed head, while other studies measured the general sound levels within an area and in some cases positioned the meter by the nurse station. Alternatively some studies took readings in a number of different areas to build up an overall picture of the noise climate. On occasions this positioning information was omitted from the paper completely.
3.2.2.
Understanding the overall hospital noise climate
Due to the singular nature of many of the studies, it has been difficult to build up a full picture of noise within a hospital as a whole. Measurements are provided for specific settings, but have no context in which they can be compared with other areas in the same hospital. In their wide-ranging study, Busch-Vishniac et al (2005) took measurements in five different locations within the Johns Hopkins Hospital in Baltimore, USA. The locations included a paediatric ICU, an adult medical / surgical ward and a ward for immuno-compromised patients. The locations were in a variety of buildings of differing ages (the most recent having being built in 1999, the oldest in 1950). To build up a full picture of the noise climate in each location, measurements were made in patient rooms, hallways and at nurse stations. The study found that there was little difference between the sound levels measured in different locations, although it did cite hallways as the noisiest areas, followed by nurse stations and patient rooms. The average recorded levels exceeded the levels specified by WHO guidelines by 20 dB and by at least 15 dB for LAmax. The authors expressed surprise that the newer building (where noise had been a consideration during the design and construction) was not particularly quieter than the older buildings. This study also examined the frequency spectra of the measured sound. Analysis of the component parts of the noise can provide extremely useful information regarding the sources of noise. The
21
spectra were found to be similar in shape for each of the locations and the following observations were made: The low frequency noise was almost certainly related to the air handling systems Due to the amount of talking observed, the shape of the mid frequency spectrum could easily be explained. The high frequency noise was thought to be predominantly caused by alarms and mobile medical equipment. The high velocity airflow system was also thought be influential. To add to this picture of the hospital noise climate, two further studies were conducted at the Johns Hopkins Hospital: Kracht et al (2007) measured noise levels in the 38 operating rooms at the Johns Hopkins Hospital. Noise levels occurring during each type of surgical procedure were captured and frequency spectra were also analysed. For neurosurgery and orthopaedic surgery, peak levels were found to be above 100 dBA for over 40% of the time. The highest recorded peaks were in excess of 120 dBA. The results raised two concerns: the potential for hearing loss and the disruption to clear speech communication. Orellana et al (2007) recorded LAeq measurements within seven different areas of an adult A&E department. Levels were found to be 5 dB higher than those recorded in other inpatient units within the hospital, with the triage area of the department found to be the noisiest. The study raised concerns regarding speech communication without errors, with additional concerns for the medical staff, since speaking in a raised voice can in itself be tiring.
3.2.3.
Identifying noise sources
In addition to building up a picture of the noise climate within a hospital in terms of noise levels, some studies endeavoured to provide a list of the main sources of noise. Some earlier studies relied on an individual observer making lists of those sounds which were perceived to be the loudest. This was of course difficult to later tally with data measurements and relied on subjective opinion and accuracy of observation. It also introduced the possibility of the Hawthorne effect as the observer would probably need to inform those around them of their purpose. However, measuring noise levels during an un-manned study presents a different problem, namely how to identify the sources of high level noise. Hodge and Thompson (1990) tackled this problem by making an audio recording of the entire surgical procedure as well as making measurements with a sound level meter. This allowed sound peaks to be identified. However, this method introduced more intrusive equipment into a very sensitive area and also required accurate synchronising of the equipment, potentially leading to later errors. The study
22
found that during surgery the main sources of high level noise were the sucker and the ventilator, with the anaesthetic alarms and intercom also contributing. In their noise surveys of operating theatres, Kracht et al (2007), reported that it was not possible to link peak sound pressure levels with specific events (for example the use of a bone saw).
3.2.4. Discussion
The review of the literature on hospital noise levels has indicated that the problem of hospital noise appears to be universal in nature, with a clear trend of rising noise levels both day and night since 1960. Without exception, all noise measurement studies reviewed found that hospital noise levels exceeded both the World Health Organisation guidelines and the standards set within their own particular countries. This finding surely makes the validity of the standards questionable. The majority of studies evaluated have targeted perceived noisier areas of hospital care, with a bias towards patient noise exposure. Few studies concentrate on inpatient ward noise levels and fail to give adequate consideration to the working environment for staff. Some consideration should be given to the potential impact of the Hawthorne effect within the design of a future study. However, this raises the issue of the ethics of studies undertaken in secret. Previous studies have failed to build up a robust picture of the noise climate within an entire hospital ward, with most using only a single microphone position. Measurement intervals also tend to be short (24 hours or less), with no representative interval established. Identification of noise sources was often found to be unscientific, with acoustic inconsistencies and omissions in reported data making meaningful study comparisons difficult. The new generation of sound level meter may overcome the problem of identification of noise sources without observers. Due to the large data storage capacity now available, these meters are able to record a short digital audio file whenever sound levels exceed a certain threshold. The audio files are automatically synchronised with measured sound levels, and so on playback it is possible to identify the source and level of a particular sound.
3.3. Sleep studies

Sleep is widely thought to be necessary for the restorative and energy conservation processes (Adam and Oswald, 1983; Berger and Phillips, 1995), and sleep disturbances have been shown to exacerbate the pain felt by hospital patients (Raymond, 2001). Many reviewed studies considered the effects of environmental and medical factors on sleep, but few looked directly at the relationship between noise and sleep in a hospital environment. This section reviews four studies that were felt to be relevant to the current study.
23
Aaron et al (1996) conducted a study to test the hypothesis that nocturnal (midnight to 6am) sound peaks would be associated with an increase of EEG arousals from sleep in patients in the respiratory ICU at the Rhode Island Hospital, Providence, USA. A significant difference was found between the number of sleep arousals in quiet periods and the number in very loud periods. However, due to the small sample size of six patients and the differing health factors (which prevented realistic comparison), the study found only an indirect link between sleep disturbance and environmental noise. Freedman et al (2001) studied 22 critically ill patients with continuous polysomnography (PSG) to characterize the sleep-wake patterns and objectively determined the effect of environmental noise on sleep disruption. This study was deemed to be unique at the time as the measurement output from the sound level meter was relayed to the PSG so that the results could be simultaneously evaluated. Findings showed that environmental noise was responsible for 11.5% of overall arousals and 17% of awakenings, but concluded that noise was not responsible for the majority of sleep fragmentation and therefore may not be as disruptive as previously thought. Gabor et al (2003) monitored critically ill patients in an ICU and compared the results with a sample of healthy, unattended individuals who volunteered to take part in the study in ICU to see the effect of noise on their sleep quality. The study also investigated the effectiveness of a noise-reduction strategy by monitoring subjects in a single bed room. All subjects were monitored with continuous and attended PSG. This study made some interesting findings: Fewer than 30% of arousals and awakenings in the ICU patient group were identifiably due to noise and patient care activities. This suggested that other elements of a critically ill patients environment should be investigated as causes of sleep disruption, and as with Freedman et al (2001), this is in contradiction to traditional hypotheses. The healthy individuals in the sample slept relatively well in the ICU. Noise was responsible for the majority of their sleep disruptions, although this was unsurprising, as other potential interruptions present. A quantitative improvement of sleep in single bed rooms was found, but sleep architecture was nearly identical. As with the majority of sleep studies, this study was limited by its small sample sizes. such as patient care activities and respiratory ventilation were not
3.3.1. Modification of room acoustics and its effects on sleep

Berg (2001) monitored the sleep patterns of subjects exposed to different sounds in an acoustically altered hospital room. PSG monitoring was carried out on twelve healthy subjects between 20 and 25 years old with no prior history of sleeping difficulties. The room was a refurbished former three bed surgical ward in a Swedish hospital with a suspended ceiling. The ceiling comprised sound reflecting 24
tiles during the first two nights of the study. On the third night the tiles were replaced with visually identical sound absorbing tiles. There were minor differences measured in the SPL before and after the intervention. The reverberation time was found to decrease by an average of 0.12 seconds (200 to 5000 Hz). Twelve different sounds of varying frequencies were played at different levels (27 to 58 dBA). The study found no significant difference in sound induced sleep stage changes, but did find fewer EEG sleep arousals in the less reverberant room. This study did not however deal directly with hospital noise. The type of environmental sounds did not reflect those that patients would be exposed to in a healthcare setting, and the level at which they were played was not necessarily representative of that found in a hospital. However, the study showed that the room acoustic design modification appeared to have some effect on sleep quality and as such is considered to be relevant.
3.3.2. Discussion
Study findings suggest that, contrary to traditional hypotheses, noise is not responsible for the majority of sleep disturbances of critically ill patients; however with such small sample sizes, no definitive conclusion is possible. The large numbers of uncontrollable factors make it extremely difficult to obtain realistic and comparable data within this category of studies; especially when studying patient groups. Each patient is unique, with a different physicality, different health issues, and taking differing amounts and types of medication. Even when considering two healthy individuals, they would sleep differently in the same environment, and so, with the introduction of so many additional variables, it is very hard to draw meaningful conclusions.
3.4. The effects of behaviour modification on hospital noise

Human behaviour has been identified as being responsible for a surprisingly large percentage of high level noise within hospital settings. It appears that through simple methods, at minimal cost, noise levels can be lowered by making individuals aware of the direct and indirect impact of their actions. Elander and Hellstom (1995) measured noise levels for routine activities within a neonatal ICU of a Swedish University hospital. It was found that many of the high levels were attributable to human behaviour, for example staff laughter, conversation and careless closing of doors, incubators and drawers.
25
As part of the study an education programme was presented to the ICU staff which consisted of three parts: A videotape showing a child's post-operative period (filmed from the child's view-point; highlighting the childs reaction to various sounds) - one nurse was surprised to find an infant wake and start to cry at the sound of her voice. Sound level values for various activities were provided to help nursing staff to put the levels into context. Detailed discussions with staff were carried out to identify realistic ways of modifying behaviour. Using a dosimeter, noise levels measurements were made in an infants cot both before and after the education programme. The study found an average decrease of 8 dB LAeq following the programme. Kahn et al (1999) conducted a two part study which sought to limit noise in the ICU of a Rhode Island hospital by behaviour modification. Firstly, twelve of the loudest sources of noise were identified. It was found that half of these sources were attributable to human behaviour and thus could be potentially modified (with talking and the television as the most prominent). A staff training programme was devised and implemented, following which a behaviour modification programme was enforced for a three week period. The study found that the programme was effective in reducing the noise levels and recommended that it be used as one part of a larger noise control programme. A study conducted by Johnson and Thornhill (2006) came to the same conclusions as Kahn et al, but stressed the need for support from management in any attempt at long term behaviour modification. A team effort in noise reduction was required for success.
3.4.1. Discussion
Studies have shown that as part of a training programme staff initially make efforts to modify their behaviour, especially if the training period is being monitored. However, after this initial period it is potentially difficult to motivate staff to continue. Staff must be enthusiastic for behavioural modification programmes to work. These programmes need frequent re-evaluation, education and feedback to reinforce behavioural change. This is only possible with 100% support throughout the entire staff hierarchy. Management must realise the benefits of noise reduction, otherwise this is not a realistic proposition. For a culture of quiet to be re-adopted within hospitals, a complete change of attitude is required. This is not only necessary from a staff and management perspective, but also required of visitors who need to be aware and respectful of the needs of the other patients around them.
26
3.5. The effects of room acoustic design modifications

Good room acoustic design essentially means that the acoustics are adapted to the activities being carried out in the space. In theory, hospital room acoustics should not only aim to reduce the sound level, but give priority to both speech clarity (as clear communication between staff, and staff and patients is paramount), and speech privacy for patients. Hospitals rooms are generally made up of hard, easy to clean surfaces. Carpets and curtains are rarely used. As mentioned in Section 2.2.1 a suspended ceiling is generally the only feasible area that can be used for the placement of sound absorbing materials. Ceilings provide a relatively large area for sound absorbency and this can have the following effects: (i) reduction in the room reverberation time; (ii) reduction in the measured sound level of the room; (iii) improvement in speech intelligibility. This section reviews studies where physical changes have been made to room acoustics and the subjective and physiological impact of patients examined.
3.5.1. Control of Infection and room acoustics

As discussed in Section 2.5, there has recently been a great deal of concern regarding the use of acoustically absorbent materials in areas where Control of Infection is important. The study described below by McLeod et al (2007), demonstrates one method of introducing effective, long term noise reduction, whilst meeting healthcare standards and minimising costs. The study was carried out in an immuno-suppressed Haematological Cancer ward of the Johns Hopkins Hospital, Baltimore, US, which had been built with a reflective, solid ceiling after concerns that the small holes typically found in an acoustic ceiling might harbour bacteria. The chosen approach was to add custom-made sound absorbing panels to the walls of the unit. These were made by the research team and consisted of glass fibre wrapped in anti-bacterial fabric, as at the time only a single vendor was found to be selling suitable material; a review of manufacturers literature shows that this has since improved. Objective and subjective data were collected before and after the installation of the sound absorbers. It was found that there was an approximate drop of 5 dB in the measured LAeq and the reverberation time was more than halved. The subjective view was that the unit had changed from one perceived to be 'very noisy' to one that was 'relatively quiet' Based on anecdotal evidence, it was concluded that the immediate impact of the sound absorbing panels was to permit patients, staff, and visitors to lower the level of their voices whilst still being well understood. It was felt that this probably accounted for the majority of the sound level drop. Staff commented on how loud the telephone and overhead paging system sounded after installation of the panels, and subsequent steps were then taken to lower the volume of these systems. The drop in sound level was also felt to promote a safer environment with greater confidence in understanding speech.
27
3.5.2. Physiological response to acoustic modification

The physiological response of neonatal babies was examined by Johnson (2001). Previous studies had found higher sound levels inside an incubator than in the open neonatal intensive care unit. It was thought the primary causes were the incubator operating motor and care giving equipment, such as the ventilator and suction tubing. Acoustic foam was added to an incubator and found to significantly reduce environmental noise. The response of the neonatal babies to this noise reduction was measured as changes in oxygen saturation. It was found that there was a significant correlation between environmental noise and levels of oxygen support therapy required by the neonates. A study by Hagerman et al (2005), examined the role of room acoustics on patients with coronary artery disease admitted to the Huddinge University Hospital, Sweden. The study focussed on changes of physiological parameters which were previously shown to be sensitive to physiological arousal. These parameters were heart rate, heart rate variability, blood pressure (systolic and diastolic), and pulse amplitude. A subjective response from patients was also analysed; this took the form of a questionnaire with a number of questions about the quality of care. The study took place over a period of eight weeks. During the first four weeks the ceiling tiles in the patient rooms and the main work area consisted of sound reflecting plaster tiles (bad acoustics). For the last four weeks these ceiling tiles were changed to Class A sound absorbing tiles (good acoustics). The tiles were visually identical. A total of 94 patients were analysed during the eight week study period. It should be noted that the average stay within the unit was 17 hours, so no patients were present during both bad and good acoustic periods. Impacts of the changes were: During the 'good acoustics period sound pressure levels fell marginally, but the reverberation times were halved. Speech intelligibility (as measured by the RASTI method and by subjective reports from staff) was found to improve considerably. There was found to be a significant difference in pulse amplitude at night between groups during the bad and good acoustics periods, together with a significantly greater need for extra intravenous beta-blockers for patients (suggested to be an indication of pain) with bad acoustics. Patients treated during the good acoustics period considered staff attitude to be much better than those treated during the bad acoustics period. There was a higher incidence of re-hospitalisation at both 1 and 3 months in the group with bad acoustics compared to the good acoustics group. Early mortality was not found to differ.
28
3.5.3.
Discussion
The study carried out by McLeod et al (2007), showed an innovative way of making sound absorbers which met Control of Infection standards. As discussed earlier in Chapter 2, it seems that no clear design guidelines are readily available regarding the choice of acoustic materials for use in areas where there are concerns about infection control. The review by the authors suggests that this is not just an issue in the UK, although the study indicated that improvements are being made regarding the choice of products available, and this had been found to be the case from a recent review of manufacturers literature. The improvement in the acoustic design of a space (by the addition of sound absorbing materials) has been shown to improve both objective measurements and subjective perceptions. Speech intelligibility is also enhanced, resulting in the lowering of voices and hence a quieter noise climate. Only a single study was found which attempted to link the change of acoustic design with a lowering of re-hospitalization rates. It is felt that with a sample set of patients suffering from complex and serious conditions, there are too many variables involved to draw any meaningful conclusions in this regard. There appear to be few studies which systematically vary acoustic conditions. This is an area which could usefully be investigated further, however in practice this would be difficult to achieve in a working hospital environment.
3.6. Conclusions
This chapter has reviewed studies which have objectively measured sound levels in hospitals, and examined the effects of changing the room acoustics on both noise levels and on the physiological responses of patients and staff. Many of the results and the limitations of the previous studies have been used to inform the design of the current study, which is discussed in Chapter 5. The following chapter discusses previous research which has investigated the effects of noise on staff and patients.
29
Acoustic Design for Inpatient Facilities in Hospitals The effects of noise on staff and patients ___________________________________________________________________________________________________________________
4. The effects of noise on staff and patients

4.1. Introduction
This chapter focuses on previous research which looked at the effects of noise on staff and patients in healthcare environments. For the purposes of clarity this review is divided into two categories: (i) the effects of noise on healthcare staff and (ii) the effects of noise on patients. Discussion regarding study findings, design limitations and areas which appear to be lacking in research is provided at the end of each section. It should be noted that in some areas very few studies have been undertaken and these few are regularly cited in the research literature. This is especially true of studies investigating the effects of noise on patients.
4.2. Effects of noise on staff

4.2.1. Stress levels and burnout
Topf and Dillon (1988) investigated whether noise-induced stress was a predictor of burnout in critical care nurses. Two university hospitals on the west coast of America were involved in the study, with 100 critical care staff from a range of backgrounds surveyed. The surveys were designed to build up a picture of the stress that the healthcare staff felt they were under, and were not only related to noise, but assessed other factors including stress caused by life events and occupational stress. The results supported the hypothesis that a greater degree of noise would be linked with a greater degree of burnout. The study also found that nurses with an intrinsic sensitivity to noise were no more at risk from burnout linked with noise induced stress than those intrinsically less sensitive. As part of the study staff were asked to indicate which noises were felt to be the most disturbing. These were compared to those found in a previous study by the authors examining patients after surgery. Staff cited beeping monitors, equipment alarms and telephones as the most disturbing sources; patients cited loud talking in the corridor at night and other patients coughing and snoring. The study raised the interesting point that the noises most disturbing to nurses may be perceived by patients as necessary for recovery.
4.2.2. Cognitive function / memory

The studies in the following paragraphs attempt to show the effects of hospital noise on cognitive function and short-term memory of staff, by simulating surgical environments. The findings are contradictory; hence it is thought that a simulation of this type of environment in a laboratory is not wholly representative of a hospital situation. Murthy et al (1995) examined the effects of noise on cognitive function and short-term memory of a group of twenty anaesthetists. Following a measurement period within an operating theatre, a 90 minute audio cassette was created to be representative of typical operating room noise. This was 30
played to the anaesthetists who undertook a series of tests. The study concluded that exposure to the recorded noise caused deterioration in mental efficiency and short-term memory in the subjects of the study. Moothy et al (2005) evaluated the effect of noise on the performance of a complex laparoscopic task. Twelve surgeons undertook this task under three controlled laboratory conditions quiet, noise at 80 to 85 dB and background music. A validated motion analysis system was used to assess performance. The noise used was monotonous repetition of background operating theatre noise and did not involve any sudden bursts of sound. It was found that neither the noise nor the music had any significant effects on task performance. It was considered likely that surgeons had learnt to effectively block out the presence of the auditory stimuli.
4.2.3. Effects of acoustic design on the work environment

The study by Hagerman et al (2005) cited in Section 3.5.2, not only examined the influence of different acoustic conditions on patient physiology, but also on the work environment and the staff in a Coronary Critical Care Unit. During the first four weeks the ceiling tiles in the patient rooms and the main staff work area were sound reflecting plaster tiles (bad acoustics). For the last four weeks these ceiling tiles were changed to Class A sound absorbing tiles (good acoustics). The tiles were visually identical. Thirty six regular members of staff were asked to participate in the investigation of the psychosocial environment and emotional states, and were required to complete a questionnaire at the start and end of each shift. The questionnaires were designed in line with the 'demand-control-support model', frequently used in healthcare to analyse work related stress. During the good acoustics period it was found that staff (particularly on the afternoon shift) experienced significantly lower work demands and reported less pressure and strain. Staff also reported feeling more relaxed and less irritable, and considered speech intelligibility to have improved. Caution must be taken in interpreting results of studies of this nature. Although objective measurements showed a change in some acoustic parameters, there are many other contributing factors such as work load and tiredness, which may have an effect on the staff mood and their perceived stress levels. This study did not appear to take these other factors into consideration, and as such the findings may be compromised.
4.2.4. Discussion
There appear to be very few studies which deal with the effects of the acoustic design on staff outcomes such as job stress, work demands, fatigue, and quality of patient care. Although it appears that each member of staff is individual in their tolerance to and their perception of noise, one regularly cited area of disturbance is that of equipment noise.
31
Interestingly, one study indicates that the noises most disturbing to nurses may be perceived by patients as necessary for recovery. Concerns are raised regarding staff speech communication without errors and the fatiguing effects of having to communicate with a raised voice. Some evidence exists that staff may be able to effectively tune out auditory stimuli whilst performing tasks which require a high degree of concentration. However, laboratory studies examining the effects of noise on surgical task performance proved to be contradictory.
4.3. Effects of noise on patients

4.3.1. Recovery rates
In an opportunistic study, Fife and Rappaport (1976) took advantage of construction work being carried out outside the University of Minnesota hospital to examine the effects that this might have on the recovery rates of patients. The building works were situated outside the rooms of patients recovering from a cataract operation. Comparisons were made between discharge rates a year prior to the study and discharge rates one year later. It was assumed that discharge dates were determined by wound healing. The study sample chosen were patients that were undergoing simple cataract surgery, who were free of any diagnoses that were likely to cause complications. This made the results of the study less subject to variation. It was found that the difference between the average length of stay during the noisy period and the pooled quiet periods was statistically significant, increasing from 8.7 days to 9.9 days (p<0.05, one tail test).
4.3.2. Subjective response to noise

Allaouchiche (2002) undertook a multi-disciplinary study looking at the effects of noise on patients recovering from the effects of anaesthesia. The study, carried out in the post anaesthesia care unit in a hospital in Lyon, France, involved 26 adult patients. Objective measurements were taken with the sound level meter positioned close to the patients head. Patients were interviewed two hours after discharge and asked to complete two questionnaires to assess their experiences on the unit. The first questionnaire was unstructured; the second was structured and asked questions about common complaints, including noise. The study concluded that high levels of noise were present and that the majority of this noise could have been prevented. However, noise was not perceived by patients as the main cause of discomfort, with only 19% (five patients) identifying noise as an important factor. Of these, four complained about conversation and one about equipment alarms. It was shown that approximately 55% of sound peaks greater than 65 dBA were caused by conversation.
32
Pugh et al (2007) note an interesting point in their review of noise studies in ICUs. Each patient is individual in their tolerance for, and how they view, noise. Some patients like the reassurance of hearing alarms and having people talking around them because it makes them feel safe. The review concluded that the impact of noise should be reduced by a three way approach - modifying staff behaviour and practices, minimising the disruption caused by equipment and alarms and optimising the acoustic design of the ICU.
4.3.3. Speech privacy

Barlas et al (2001) examined whether patients perceived less privacy in A&E curtained treatment areas than in walled rooms and concluded that patients from curtained areas did report significantly less auditory, visual and overall privacy than those in rooms. The study made several points worthy of note: 85% of patients reported a high degree of respect for privacy from the staff. A small percentage of patients in the curtained areas withheld portions of their medical history or refused part of their medical examination because of privacy concerns. Older patients believed that they could hear others' conversations with a physician or nurse more often than younger patients. It was felt that due to the sensitive nature of some of the questions posed, responses may not represent the true feelings of the individuals.
4.3.4. Single bed patient rooms

Van de Glind et al (2007) noted that an increasing number of hospitals have taken the decision to provide single bed patient rooms. It is not clear if these policy decisions are based on scientific evidence. The study reviewed the literature currently available on the benefits of single bed rooms, and examined the following: privacy and dignity, patient satisfaction with care, noise and quality of sleep, hospital infection rates, recovery rates and patient safety issues. The study concludes that due to the lack of research, there is currently not enough evidence to prove that the introduction of single bed rooms is beneficial.
4.3.5. Discussion
Only one single, opportunistic study was found which attempted to link patient recovery rates to noise levels. Further research in the area would be beneficial, but the number of variables involved in carrying out studies in healthcare makes meaningful results difficult to obtain. Although it appears that each patient is individual in their tolerance for and how they view noise, one regularly cited cause of patient disturbance is staff conversation. It was found that some patients like
33
the reassurance of hearing alarms and having people talking around them because it makes them feel safe. Privacy concerns may cause patients to withhold portions of their medical history or refuse part of their medical examination. This appears to be particularly relevant in the case of more elderly patients.
4.4. Conclusions
This chapter has highlighted the lack of extensive research studies examining the effects of noise on staff and patients, and the difficulty of obtaining reliable and meaningful data in both hospital environment and simulated laboratory studies. This lack of data has influenced the current study which aims to further investigate the subjective perceptions of staff and patients to noise in a range of ward types. This is discussed further in the following chapter.
34
Acoustic Design for Inpatient Facilities in Hospitals Study Design ___________________________________________________________________________________________________________________
5. Study design
5.1. Introduction
The aims and objectives of the current study were informed by the literature review, with input from the industrial partner, Arup Global Healthcare. This chapter outlines the aims and objectives of the research and discusses the objective and subjective survey methods used in more detail. The preliminary work involved in obtaining ethical approval and the necessary permissions to carry out the study within an occupied ward environment of a hospital are also discussed.
5.2. Study outline aims and objectives

The following conclusions drawn from the literature review were seminal in informing the proposed study design: Very few studies have been carried out in general inpatient hospital wards. The majority of previous studies have been undertaken by healthcare staff with little knowledge of acoustics. This has led to inconsistencies in the use of acoustic parameters; short or incomplete measurement periods; unknown microphone positioning and a general lack of rigour. Few studies have compared noise levels in multi-bed and single bed patient accommodation; nor have they attempted to build up an overall picture of the noise climate of a ward. There is a noticeable lack of studies carried out in UK hospitals. Only one study was found which investigated the relationship between acoustic design and design for infection control purposes. Further studies exploring patients perceptions of privacy would be beneficial. The current UK acoustic design guidelines are effective in terms of general construction advice, but have less relevance in terms of occupied ward areas. The current study therefore aimed to address many of these issues, and this is discussed in the following sections. It was decided that the proposed research would be both objective and subjective in its nature. The objective study would consist of an acoustic survey to obtain data on the noise levels and acoustic conditions in inpatient hospital wards. The subjective component would aim to build up an understanding of staff and patient perceptions of noise in the same ward environments by use of questionnaire surveys. It was felt that the hospitals involved in the study should be chosen to reflect a broad range of building and ward designs, different types of inpatient care and a mixture of surgical and medical wards. This 35
would allow many useful comparisons to be made. Buildings undergoing refurbishment were of special interest for pre and post intervention studies. Individuals working in healthcare estates with links to the Medical Architecture Research Unit at London South Bank University were contacted to locate possible study sites. Three potential sites were identified: Great Ormond Street Childrens Hospital, London; Bedford Hospital, Bedford; and Addenbrookes Hospital, Cambridge.
5.2.1. Acoustic survey

The main aim of the acoustic survey was to build up a picture of the noise climate within general inpatient care wards by making a comprehensive series of noise measurements. The data captured would include average, maximum and background noise levels and the identification of the sources of high level noise. Noise levels during the day and night would be investigated to allow for comparison with relevant standards. Particular consideration would be given to building construction, ward layout, the amount of acoustic absorbency provided and how the design for control of infection affected the acoustic comfort within the space. Where possible, factors such as ceiling finishes, ventilation systems and glazing were also to be investigated: It was thought that the use of technology on the ward might have a negative impact on the noise climate. Staff systems, medical equipment and patient entertainment (including TV and radio) were therefore investigated to build up an understanding of the effects of their use on the noise climate. In addition to noise measurements other room acoustic parameters such as reverberation times were considered to provide further indication of the acoustic comfort of the space. It was felt that the data captured during these objective studies would be invaluable in understanding the key elements of the noise climate in inpatient care and link directly into the following areas: understanding of the physical and behavioural factors that significantly affect the noise climate the effects of acoustic design changes on the acoustic comfort of a space exploration of the conflicts between design for infection control and acoustic comfort
5.2.2. Questionnaire surveys

Initially it was hoped that a subjective assessment of the noise climate in inpatient care could be undertaken using a semi-structured interview approach, talking to both staff and patients. However it became apparent that to obtain ethical approval within the project time constraints, a questionnaire survey approach would be more suitable. Further details concerning the need for ethical approval can be found in Section 5.5. 36
Two questionnaires were designed, one for the ward medical staff and another for patients whose stay on the ward was longer than 24 hours. The aim of the questionnaires was to build up an understanding of the perceptions of both staff and patients regarding noise in their environment, and ultimately to establish whether any relationships existed between the objective data collected on the wards and the perceptions of the ward users. Good questionnaire design is paramount if meaningful data is to be collected in a survey of this type. The use of leading words or questions should always be avoided. As such, a great deal of thought was given to the type of information that was required, and many questions were discarded before reaching the final versions. The length of time taken to complete the questionnaire was also considered. Staff are very busy and unlikely to complete a survey which may take longer than five minutes. Patients, also, would find a long questionnaire daunting, especially if they were feeling unwell or weak. With this in mind the questionnaire was kept relatively short and the layout was designed for clarity and ease of completion. Questionnaires in their final form were trialled throughout the pilot study (see Chapter 6). Responses were reviewed and any questions that were felt to be ambiguous were rewritten. Further details on the design of the questionnaire surveys can be found in section 5.4 of this chapter, and sample questionnaires can be viewed in Appendix A.
5.2.3. Comparison studies

The ultimate aim of the study was to make use of the objective and subjective data captured in the following ways: comparison studies of single and multi-bed inpatient accommodation comparison studies of inpatient wards situated in buildings of differing age, construction and layout a comparison study of a surgical and medical inpatient ward examination of the perceptions of the acoustic environment of patients and staff identification of the dominant; most annoying; and most disturbing noise sources analysis of noise, acoustic and subjective data in order to suggest methods of noise control, particularly in the areas of equipment and human interface
37
5.3. Acoustic survey methodology

5.3.1. Equipment
The following equipment was used throughout the study: Norsonic 140 Class 1 Sound Level Meter Norsonic Sound Calibrator Type 1251 (114 dB @ 1000Hz) Norsonic Environmental Case with two additional heavy duty batteries Additional heavy duty batteries to allow for quick equipment rotation 5 m microphone extension cable Mini microphone tripod / 300 mm ceiling bracket To allow for longer term measurements to be made, the sound level meter (SLM) was placed in an environmental case with two heavy duty batteries. The life of the batteries was such that one weeks worth of data could be collected at each measurement position before replacements were required. A five metre extension cable allowed the microphone to be placed away from the environmental case, which afforded flexibility regarding its positioning, see Figure 5.1.
Figure 5.1 Sound level meter, environmental case and associated equipment
5.3.2. Control of Infection

Due consideration was given regarding the choice of equipment and whether it could be easily wiped clean if necessary. It was felt that the environmental case, which was made of tough plastic, was easily cleanable. However the microphone, being extremely sensitive, would not be cleanable to any degree. It was therefore decided that if the microphone was positioned out of reach, it was unlikely to be touched and contaminated, and thus unlikely to need specific cleaning for purposes of infection control. Meetings with a member of the hospital Control of Infection team were held at each study site
38
to ensure that the use of the noise measurement equipment was acceptable and met with hospital infection control policies.
5.3.3. Acoustic parameters

To allow an acoustically robust picture of the noise climate to be built up, the following parameters were measured: LAeq,1hr, LAmax, LA90, LAmin and LZ(SPL) and reported where appropriate.. The third octave frequency band spectrum of the sound was also included. Throughout the study the SLM was set on a fast time weighting. At each change of measurement location the data captured was downloaded onto a laptop computer and reviewed further using the software provided by Norsonic, the manufacturer of the SLM. This software, NorReview, allowed all captured data to be viewed graphically; to be analysed in detail; and subsequently exported into a Microsoft Excel spreadsheet for reporting purposes.
5.3.4. Presentation of sound levels

To build up an understanding of the noise climate in each ward location, sound levels are presented throughout the study in a number of different ways: The measured LAeq,24hr, LAeq,16hr and LAeq,8hr quoted in tables for each ward are arithmetic averages of each metric over the number of days in the measurement interval. For example, if five days worth of data have been collected, the reported LAeq,16hr is the arithmetic average of the five daily measured LAeq,16hrs. Where average LAeq,1hr and LA90,1hr levels are shown graphically over 24 hours, this is again the arithmetic average of each metric over the measurement period. For example, for a five day measurement period, for the time interval 11.00 to 12.00, the LAeq,1hr would be the arithmetic average of five LAeq,1hrs measured from 11.00 to 12.00. Any other sound level presented is defined and labelled.
5.3.5. Measurement interval

Many of the measurement intervals in the reviewed studies were either short (less than 24 hour) or incomplete. It was felt that to build up an accurate picture of noise in inpatient care it was extremely important to establish a representative measurement interval. This interval could then be used throughout the main study. This is discussed further in Section 6.14.
5.3.6. Measurement locations

To fully understand the noise climate of a ward, measurements were undertaken in each type of patient accommodation and at the nurse stations. For example, if a ward consisted of four bed and single room accommodation, it was decided that at least two single rooms and two 4-bed bays should
39
be measured. This would ensure that the data captured was typical of the accommodation type and enable comparisons to be made.
5.3.7. Identifying sources of high level noise without an observer

Previous studies conducted in healthcare environments often made use of a team of observers to note down the sources of high level noise. Identifying these sources is important, as it allows some noise control or other remedial measures to be put in place to counteract unnecessary noise. Of course the use of observers is only possible in the short-term. Where longer measurement periods are proposed, this is not realistic or practical. The SLM used in the study incorporated built in level above trigger functionality. By enabling this feature, a short sound file is created as soon as the value of LAmax exceeds a specified threshold. Once the data is downloaded for analysis, each sound file can be reviewed and the sources of the high level noise identified.
5.3.8. Reverberation times

Reverberation time (RT) is generally used as an indicator of the acoustic comfort of a space and is an important measurement in the field of room acoustics. There are a number of methods used to measure the RT value. The two widely used methods listed in the British Standard BS EN ISO 3382:2 (2008) are the Interrupted Noise Method or the Impulse Response Method. Both methods rely on generating high level noise, which would generally be unacceptable in an occupied hospital ward. It was decided that if an opportunity arose to make RT measurements in unoccupied patient accommodation, then the Impulse Response Method should be used with a balloon burst as the source. This was purely down to the logistics of carrying bulky equipment into a hospital. Loudspeakers and amplifiers would be required if the Interrupted Noise Method was chosen, and this was not practical. Where the Impulse Response Method was used, the number of source and receiver positions stipulated by British Standard BS EN ISO 3382-2 (1998) for engineering work was adhered to with RT20 values for octave frequency bands from 250 Hz to 4000 Hz reported. For this category at least two source and at least two receiver positions were stipulated for each measurement. Where on site RT measurement was not possible, an estimation method was used which made use of level above trigger sound files. This is discussed in further detail in Chapter 10.
5.4. Questionnaire survey design

This section describes the design of the staff and patient questionnaires used in the study. Examples of each questionnaire can be seen in Appendix A.
40
5.4.1. Staff questionnaires

The first section entitled About You was designed to categorise certain staff attributes by identifying sex, age bracket, staff grade, length of service on the ward and at the hospital. The questions were very general, without asking anything that could be deemed as too personal or intrusive. It was thought that with a large overall dataset, these categories may help to establish relationships between, for example, the length of service and noise annoyance. The second section entitled About Your Environment examined perceptions of noise annoyance and noise interference with the ability to work. It was felt that it was important to not only examine noise annoyance but also whether staff felt that their work was impacted by certain sounds. The questionnaire sought to identify the sources of both annoyance and interference by providing a list of noises (which were identified from initial observations made in the pilot study ward). Staff were asked to rate the annoyance / interference of each noise source listed on a scale of 0 to 4 (where 0 indicated no annoyance / interference and 4 indicated a great deal). Several lines were left blank at the bottom of the lists for staff to add and rate additional noise sources. The third section aimed to aid understanding of which sounds were felt by staff to be important in order to carry out their jobs effectively. Staff were asked to rate the sounds on a scale of 0 to 4, where 0 indicated not at all important and 4 indicated extremely important. Again, several lines were left blank at the bottom of the lists for staff to add and rate additional sounds that they considered important. A comments section was left at the end of the questionnaire for any additional feedback.
5.4.2. Patient questionnaires

The first section entitled About You was designed to categorise certain general attributes by identifying sex, age bracket, length of stay on the ward and bed number. As with the staff questionnaires the questions were very general, without asking anything that could be deemed as too personal or intrusive. The following provides the reasoning behind this line of questioning: The sex of a patient may yield information regarding differences between men and women regarding sensitivity to noise. The age group of the patient may be related to their sensitivity to noise, as hearing generally deteriorates with age. How long a patient has been on the ward may indicate whether there is a correlation between length of stay and becoming more acutely aware of noise, or whether an individual becomes more used to the noise levels.
41
The patients bed number provides location information. Relationships may be shown to exist between bed location and specific sources of noise. The second section titled About Your Environment considered noise annoyance both during the day and at night. The questionnaire sought to identify the sources of noise that may annoy or disturb patients. Respondents were given a list of noises (which were identified from initial observations made in the pilot study ward), and were asked to rate the annoyance / disturbance on a scale of 0 to 4 (where 0 indicated no annoyance / disturbance and 4 indicated a great deal). Several lines were left blank at the bottom of the lists for patients to add and rate additional noise sources. The third section of the questionnaire contained a number of questions designed to investigate the acoustic environment further, including perceptions of speech privacy. The following paragraphs discuss the contents of this section further: Sound was examined in a positive rather than in a negative light, with patients asked if there were any sounds that they actually found comforting. Three blank lines were provided for a response. Communication between nursing staff and patients was investigated, by asking patients whether they could clearly hear what was said to them by the medical staff. The aim of this question was to highlight high levels of background noise and poor acoustics, but of course a patient suffering from a hearing impairment would have difficulty hearing for other reasons. To take this into consideration, patients were also asked if they had a hearing impairment. It was felt that asking details of the impairment would be deemed too personal, and so a yes or no response to this question was provided. Conversational privacy was investigated by asking whether a patient felt that they could have a private conversation at their bedside. If the response was in the affirmative, a further question was asked to see if the patient would feel comfortable speaking normally or whether they would need to lower their voice or take some other precautionary measure. Finally, respondents were asked if they felt that there was ever too little sound in a room. This question was asked as some previous study findings suggest that patients may feel isolated if it is too quiet. A comments section was left at the end of the questionnaire for any additional feedback.
5.5. Preliminary work

5.5.1. Building relationships with hospitals and Healthcare Trusts
Following the agreement of the project proposal, meetings were held with estates staff at each of the proposed locations. It was considered very important that the estates teams supporting the study felt that it would provide them with useful data. Study feedback could potentially provide information on
42
the current performance of the occupied buildings on site, and could be used to inform the design of future site developments or refurbishments. Feedback from the initial meetings was positive, and further meetings were held with the members of the estates teams and senior clinicians to ascertain the best inpatient ward locations in which to conduct the study at each site. It was agreed that a pilot study would initially take place at Great Ormond Street Childrens Hospital. It was felt that the study could yield useful information for the hospitals redevelopment team about the acoustics of the Octav Botnar Wing, whose design was heavily influenced by the need for infection control, with hard, easily cleanable surfaces.
5.5.2. Ethics and Trust approval

Before any part of the study could be started, the necessary permissions needed to be granted by each Trust. These permissions involved a personal police check of the researcher who would be working at each site, and a statement of ethical approval from both the central NHS Ethics Service and London South Bank University.
National Research Ethics Service All proposed healthcare studies require a level of scrutiny by the National Research Ethics Service (NRES). The advice published by the service is biased towards clinical trials and is hard to interpret. Some short-term projects have failed entirely due to the length of time involved in granting ethical approval. After consulting with NRES, it became apparent that there would be no ethical issues surrounding the proposed objective study, which would be viewed as an audit. However, ethical approval would be required if staff and patients were to be interviewed. This would mean a very lengthy process waiting for committee decisions. Further discussion with NRES suggested that if, rather than interviewing staff and patients, an anonymous questionnaire was used, submission to an ethics committee might not be required, with the study being classed as a service evaluation. A document was prepared explaining the planned use of staff and patient questionnaires. This document along with details of the objective study, a poster advertising the study, staff and patient information sheets and copies of the questionnaires were all submitted electronically to NRES. The response classed the study as a service evaluation and confirmed that a Research Ethics Committee review was not required. Copies of all documents mentioned above can be found in Appendix A.
43
London South Bank University Ethical Review Following the response from NRES, an application for ethical review was submitted to the Executive Dean of the Faculty of Engineering, Science and the Built Environment at London South Bank University. The London South Bank University Code of Practice for Investigations on Human Participants deems that Class 1 Investigations are any investigation taking the form of a general survey / questionnaire / interview (including telephone surveys) which do not involve the request or receipt of personal information, as defined by the Data Protection Act 1998, from the participant. The Code of Practice states that Executive Deans may approve Class 1 Investigations providing these comply with this Code of Practice. It was felt that this study fell into the category of a Class 1 Investigation and as such the relevant documentation was sent directly to the Executive Dean for review. Ethical approval from the University was given forthwith and evidence is provided in Appendix A. Hospital specific permissions Once ethical approval had been received from NRES and London South Bank University and the police checks had been finalised, temporary contracts of employment and security passes could be issued by each Trust. The onsite study was then able to commence.
5.6. Conclusions
This chapter has described the design of the objective and subjective surveys, plus the preliminary procedures that were necessary in order to carry out research in occupied hospitals. The following chapter describes the pilot study that was undertaken to validate and further develop the methods discussed in this chapter.
44
Acoustic Design for Inpatient Facilities in Hospitals Pilot Study __________________________________________________________________________________________________________________
6. Pilot Study
6.1. Introduction
A pilot study involving acoustic and questionnaire surveys was carried out in a post surgical inpatient ward in a five year old building at Great Ormond Street Childrens Hospital, London. The aim of the pilot study, which took place over a four month period from September to December 2009, was twofold: to test the methodology to be used in the main study to ensure that meaningful results could be obtained in line with the research proposal; and secondly, to provide useful feedback for the redevelopment team and the ward manager on site. The design of this particular building was heavily influenced by the need for infection control, with hard, easily cleanable surfaces. The redevelopment team were interested to find out how the building performed acoustically post occupation, and whether the design compromised this performance in any way. Particular consideration was given to the following aspects of methodology to identify the optimal dataset to be collected in the study and to ensure it would be as robust and reliable as possible. The choice of suitable microphone positions to allow for meaningful comparisons to be made between patient accommodation types. Use of the level above trigger functionality built into the sound level meter as a means of identifying sources of high level noise. The choice of a representative measurement time interval. To check for ambiguous or misleading questions in the staff and patient questionnaires. This chapter begins by looking at the background of Great Ormond Street Hospital for Children, providing an overview of the acoustic design considerations of the study ward and exploring the hospital policies and equipment usage that may affect noise levels. The chapter continues by considering the most effective ways of positioning the measurement equipment and also ensuring adequate publicity of the study. Objective results from each ward are reported, and staff and patient perceptions of the noise environment are explored. The results of the study were reported back to the ward staff in several meetings; their observations on the findings and possible actions are discussed at the end of this chapter.
6.2. Background
Great Ormond Street Hospital for Children (GOSH) was established in 1852, and was first located at 49 Great Ormond Street, London. With its motto the child first and always, the hospital has become the leading UK tertiary paediatric hospital, providing the widest range of specialist paediatric services in the country. 45
As part of an ongoing redevelopment plan for the site focusing on the delivery of a new model of care, the construction of The Octav Botnar Wing was completed in early 2006, and is shown in Figures 6.1 and 6.2. This building was designed to provide a unique, uplifting environment for both patients and staff by maximising the use of natural light, bright colours, and innovative designs. The Octav Botnar Wing houses a number of specialist centres including an International Patient Centre; Medical Day Care Centre; Orthopaedic Ward and Biomedical Engineering Centre.
Figure 6.1 The Octav Botnar Wing
Figure 6.2 Main entrance to the Octav Botnar Wing
6.3. Sky Ward

Out of the four specialist centres situated in the Octav Botnar Wing, only the orthopaedic ward (known as Sky Ward) fitted the research project criterion of general inpatient care. Length of stay here is generally from one day to two weeks, with patients undergoing a number of different types of surgery including limb lengthening procedures, and spinal, hip and foot surgery. Patient ages vary from infancy up to 18 years of age. The new clinical facilities were designed to provide greater space for patients, more comfortable surroundings for a parent to stay by their childs bedside and more efficient use of space for nursing teams. The ward is built on a "racetrack design" that positions patient rooms on the outer part of each floor and locates the health care resources in the centre of the building. Consisting of three 4-bed bays and six single patient rooms, there are a total of 18 patient beds. Rooms facing east and south include floor to ceiling glazing looking out onto a balcony area (which is locked and not available for use). Rooms facing west have smaller areas of glazing. All 4-bed bays and single rooms have ensuite shower and toilet facilities. Each patient bed also has its own flat screen television, hoist, and a bed for parents to sleep next to the patient (the single rooms have a pull down bed; the 4-bed bays have chairs which convert to beds). Figure 6.3 shows the brightly coloured ward reception area, and Figure 6.4 shows a patient bed in a 4-bed bay. 46
Figure 6.3 Sky Ward reception
Figure 6.4 Typical four bed bay
Healthcare resources are situated in the centre of the ward and include a clean and dirty utility room; a plaster room; a sensory room; kitchen; assisted bathroom; equipment store; adolescent room; ward managers office; and two nurse stations. A plan of the ward is shown in Figure 6.6 on page 51.
6.4. Building acoustic design considerations

The Octav Botnar Wing was built to conform to the Health Technical Memorandum HTM 2045: Acoustics Design Considerations (NHS Estates, 1996), which contained partition performance requirements as well as advice on many other aspects of acoustic design (as discussed in Chapter 2). It was therefore assumed that floors, walls, windows and doors were of a reasonable specification in terms of sound insulation and sound attenuation.
6.4.1. Nurse stations and common areas

Within the corridors and around the nurse stations there was no visible acoustic absorbency. The ceilings at the nurse stations were solid plaster with round inspection hatches to access services. The corridor ceilings were also solid plaster with a strip of metal ceiling tiles running down the centre to provide access to services. Floors were heavy duty vinyl and walls were plasterboard on a metal grid system.
6.4.2. Patient accommodation

The 4-bed bays and single patient rooms all had suspended ceiling grids with Armstrong Ultima ceiling tiles. The properties of these tiles are shown in terms of their sound absorption coefficients () at different frequencies in Figure 6.5.
47
Figure 6.5
Ultima ceiling tile sound absorption coefficients () over a range of frequencies Source Manufacturers datasheet
All patient accommodation has vinyl flooring and plasterboard walls mounted on a metal grid system. Additional acoustic absorbency is provided by window curtaining, upholstered upright chairs, patient and parent beds and privacy curtaining that can be pulled around each bed (4-bed bays only). Full length curtains are also provided to pull around the parent bed in the single patient rooms for additional privacy.
6.5. Ward routines

The ward manager of Sky Ward was enthusiastic and supportive of the study. He had some concerns about several of the systems in place on the ward which he considered to be excessively loud. It was decided that an investigation of these systems could be easily incorporated into the study. To help inform the study design, some time was spent on the ward to build up an appreciation of ward layout, to meet the staff and to make on-the-spot sound level measurements. Several discussions with the ward manager helped to build up an understanding of the day-to-day running of the ward and any events that could potentially affect the noise levels. This initial discussion process was found to be useful and was used throughout the main study. Information and events which were thought be of some significance are discussed in the following sections.
6.5.1. Staffing and patient levels

Due to the nature of care in this particular ward, and the timings of the operations, staffing levels and ward occupancy are generally at their highest during weekdays. Weekday staffing levels generally consist of between three and five clinically trained staff, up to three students and up to two health care assistants for a day shift, with fewer staff at night. During a weekend, ward occupancy rates drop to less than 50%, staffing levels are lower and often one 4-bed bay and several single rooms are left empty. 48
Due to this drop in occupancy, it was decided that measurements made during the weekends would not be representative of the typical use of the ward and it was therefore decided that analysis would only be carried out on weekday measurements for the pilot study.
6.5.2. Staff shift patterns and ward rounds

Staff day shifts start at 07.45 and end at 20.15, with night shifts starting at 19.45 and ending at 08.15. There is a 30 minute shift overlap between 07.45 and 08.15 and 19.45 and 20.15 with potentially more staff gathered at the nurse stations for handover sessions. Higher levels of noise may be attributed to these changeover periods. Ward rounds generally start around 08.15 (after the completion of the shift handover), with many of the children on the ward requiring hourly checks by medical staff.
6.5.3. Cleaning
Cleaning usually begins at 08.00. Daily cleaning generally consists of a felt floor mop to remove dust (floors are buffed once every two weeks); the emptying of the two rubbish bins in each of the bays and single rooms twice daily (general and chemical waste); the cleaning of the ensuite bathrooms; and bed changing.
6.5.4. Meal times

Meal times are at 12.00 and 17.00 and last approximately one hour. Meals are individually served to patients, with no meal round with a trolley. It was felt that it was unlikely that noise levels would be impacted greatly by the serving of meals to patients.
6.5.5. Medical equipment with alarms

Three types of medical equipment are used on the ward which may contribute to noise levels: The fluid pump. This has a high pitched alarm if intervention is required. Heart rate and oxygen level monitor. This has a lower pitched bong alarm if intervention is required. A nebuliser which creates a mist of medicine which is breathed in through a mask or mouthpiece. This is commonly used to give high doses of reliever medicine and when in use makes a low level bubbling sound.
6.5.6. Access to patient accommodation

All doors to the 4-bed bays are left open both day and night for staff observation. This is how the staff at this hospital are trained to care for patients in multi-bed accommodation. If doors to a bay are shut, it is assumed that there are no patients present. 49
The doors of the single patient rooms are generally left open during the day (depending on the patients preference and condition), but closed at night if there is a relative staying with the patient. Patients given single room accommodation are generally infants under one year old, those with special medical needs or infectious patients. Staff are happy for the doors to single rooms to be closed if a patient is infectious or if the parent is with the child and can call for help if the need arises.
6.6. Measurement locations

It was considered important that the measurement locations chosen were those that could be easily repeated within the main study and that the locations reflected the study aim: to build up a comprehensive picture of noise levels in inpatient care. Figure 6.6 shows the ward layout which is discussed further in the following sections.
6.6.1. Nurse stations

There were two nurse stations on the ward and these were located at either end of the central healthcare resource block. It was felt that for comparison purposes it was important to measure noise levels at both. Nurse station 1 was closest to the ward entrance at the junction of several corridors. One corridor provided access to the ward managers office and ward reception, with the other corridor running down the length of the ward. A 4-bed bay was located opposite this nurse station. As the doors to this bay were always left open, patients could potentially be affected by noise from the nurse station. Figure 6.7 shows the nurse station, and Figure 6.8 shows the corridor running to the ward managers office and ward reception.
Figure 6.7 Nurse station 1
Figure 6.8 Internal corridor
50
Figure 6.6 Layout of Sky Ward with microphone positions
Microphone Position Single Patient Room B 4 Bed Bay B
Nurses Station 1 Nurses Station 2
Ward Entrance Single Patient Room A
Reception Desk
4 Bed Bay A
Waiting Area
Acoustic Design for Inpatient Facilities in Hospitals Pilot Study __________________________________________________________________________________________________________
The nurse station was a semicircular desk with a number of drawers, on which a computer, printer, telephone, security monitor and the nurse call control panel were installed, as shown in Figure 6.9. There was a small grill on the wall behind the desk covering the loud speaker to which the nurse call system and doorbell were piped. This loud speaker can be seen labelled in Figure 6.10.
Speaker grill
Figure 6.9 Internal telephone & security monitor
Figure 6.10 Wall mounted speaker grill
Nurse station 2 was much larger than its counterpart and tended to be busier, with more staff. It was located at the opposite end of the ward to nurse station 1, with two single patient rooms directly opposite and a 4-bed bay to the right. Due to the open door observation policy the 4bed bay could be potentially affected by noise from the nurse station. To either side of the nurse station were sets of double doors but these were left open at all times, except in the event of fire, when they would automatically be closed. As with nurse station 1, this location was a semicircular desk with a number of drawers, on which several computers, a printer, a telephone, security monitor and the nurse call control panel were installed. This can be clearly seen in Figures 6.11 and 6.12. Again there was a small grill on the wall behind the desk covering the loud speaker to which the nurse call system and doorbell were piped. Patient notes were kept in ring binders and there were often a number of ring binders laid out on the top of the desk.
52
Figure 6.11 Nurse station 2
Figure 6.12 Nurse station 2 desk
6.6.2. Four bed bays

Two different 4-bed bays were chosen for comparison purposes. The rooms were different shapes, had different bed positioning and differing amounts of glazing. Identical facilities were available for patients. 4-bed bay A faced out onto nurse station 2 and was accessed through a set of open double doors. To the left hand side of the ward entrance was a hand washing sink and two rubbish bins for chemical and general waste. There were two beds positioned on the left hand side of the room and two on the right. Each patient bed could be curtained off from the ward, which was often the case when the bed was occupied, presumably for reasons of privacy. Parents were provided with a chair which converted to a bed so that they could sleep next to their child at night. At the back of the bay were floor to ceiling windows and a glazed door opening out onto the balcony area. The windows and door did not open; this bay was mechanically ventilated only. In the back right hand corner a set of metal lockers were provided for parents to store valuables, next to which was the door to the ensuite shower room. 4-bed bay B (shown in Figure 6.13) was situated at the opposite end of Sky Ward facing the smaller of the two nurse stations. Unlike bay A, this bay was L shaped, with two beds situated on the right hand side of the room and two beds on the back wall. This room was both mechanically and naturally ventilated, with openable windows. As with bay A, the bay was accessed through a set of double doors, which were left open at all times. To the left of this entrance were two rubbish bins for chemical and general waste, and round the corner was a hand washing sink, door to the ensuite shower room and lockers provided for storage of valuables, as can be seen in Figures 6.14 and 6.16. All the same patient and carer facilities existed in both bay A and bay B, including flat screen televisions for each bed and patient hoists. Figure 6.15 clearly shows these patient facilities. 53
Figure 6.13 4-bed bay B
Figure 6.14 Hand washing sink, door to shower room and lockers
Figure 6.15 Patient bed and fold down chair
Figure 6.16 Ward entrance with rubbish bins
54
6.6.3. Single patient rooms

Two different rooms were chosen for measurements. The rooms were situated on different sides of the building and were slightly different in terms of room design and the amount of glazing used. Single patient room A was opposite nurse station 2. This room was mechanically ventilated only and had full length windows and a patio door looking out onto a balcony area. As with 4-bed bay A, the windows and door could not be opened. The room had its own ensuite shower and toilet, and a pull down bed was provided for the childs parent. For privacy purposes, full length curtains could be pulled around this bed. Separate curtaining was provided to cover the external windows and patio door and also block out the light shining through the glazed panel of the door to the corridor. To the left hand side of the room was a hand washing sink and two rubbish bins for chemical and general waste. Several upright chairs were available for visitors and an easy chair next to the bed for patient or parent use. Figures 6.17, 6.18 and 6.19 show the bed head services, glazed balcony door, and pull down bed with the room sink and rubbish bins.
Patient hoist
Figure 6.17 Patient bed showing bed head services
Flat screen television
Figure 6.18 Locked door onto balcony 55
Figure 6.19 Door to ensuite, pull down bed, sink and rubbish bins
Single patient room B was located halfway down one of the main ward corridors, opposite the dirty utility room and the assisted bathroom. This room was mechanically ventilated, but unlike room A, the windows could also be opened. Room B was slightly smaller in area than room A, but had the same facilities available for patients and their carers including a pull down bed, flat screen television and a patient hoist, which could be used to hoist the patient out of bed and as far as the ensuite shower room if necessary. Figures 6.20, 6.21, 6.22 and 6.23 show the patient bed, sink and rubbish bins, the pull down bed and the patient hoist in room B.
Figure 6.20 Patient bed and opening windows
Figure 6.21 Rubbish bins and hand washing sink
Figure 6.22 Pull down bed
Figure 6.23 Flat screen television
6.7. Equipment and microphone positioning

Apart from the issue of cleanability, which was discussed in Chapter 5, care was taken over the positioning of the microphone and associated equipment so as to minimise its impact on staff duties and patient care. It was decided that to record comparable noise levels the 56
microphone should be located in similar positions in similar locations (e.g. similar positioning in two single rooms). Figure 6.6 shows the ward layout and microphone positions which are discussed in the followed sections.

As the nurse stations were busy areas, it was important that the microphone was positioned where it was not likely to be knocked, and yet could collect comparable measurement data. In both cases the microphone was placed on a shelving unit, 2.1 m high, pointing down at the nurse station. The microphone position is shown by the star symbol on Figures 6.24 and 6.25.
Figure 6.24 Microphone position at nurse station 1
Figure 6.25 Microphone position at nurse station 2
6.7.2. Four bed bays

The four bed bays presented more of a problem regarding microphone positioning. It was not possible to position the measurement equipment close to a bed head, as there was too much medical apparatus situated there. It was also felt by the ward manager that the patient / family would find having the microphone so close rather intrusive. Another consideration was that when moving the beds, equipment would easily be knocked and potentially damaged. As measurements were to be made in two 4-bed bays, it was also necessary to find a position in each ward that would yield comparable sets of measurements. A set of lockers 2 m high, which were used by the parents for storage of valuables, were identified as a possible location. These lockers, shown in Figure 6.26, were located at the back of each ward next to the ensuite bathroom door and offered a comparable position in each bay.
57
Figure 6.26 4-bed bay B with microphone placed on top of lockers

As with the 4-bed bays it was not practical to position the microphone close to the bed head. A fixed cupboard housing the parents pull down bed was identified as a comparable location in each room. As shown in Figures 6.27 & 6.28, the microphone was positioned 2 m from the ground pointing down into the rooms.
Figure 6.27 Single patient room 1 with microphone position shown
Figure 6.28 Single patient room 2 with microphone position shown
6.8. Other considerations

6.8.1. Identifying the optimal level above setting for trigger files
As discussed in Chapter 5, the use of observers to identify sources of high level noise was not practical in this study. An alternative method of identification was to make use of the built in level above trigger functionality of the sound level meter which creates a short audio file 58
each time the LAmax parameter exceeds a specified threshold. Once the data is transferred to a PC for analysis, each audio file can be reviewed and the sources of the high level noise identified. To determine the settings needed for the optimal use of this feature, various threshold and audio quality settings were tested during the first measurement periods. Due to the limited storage capacity of the sound level meter (2 Gb) it was important that the number and the size of the trigger files created did not exceed this capacity before the completion of the measurement period. If this did occur the sound level meter would simply stop part way through a five day measurement period, resulting in a loss of data. After some experimentation, a threshold of 70 dB LAmax was found to be the most workable setting, which meant that each time the value of LAmax exceeded 70 dB an audio file was created. Further detailed technical information on the use of trigger files can be found in Section 10.2.
6.8.2. Publicising the study

Some weeks before the study commenced, five laminated advertising posters were displayed throughout the ward common areas. These posters explained in simple terms why and how the study was being undertaken, and were aimed at both staff and parents / patients. In addition to these posters the ward manager personally discussed the study with all his staff during staff meetings. It was felt that it was of utmost importance that as much information as possible was provided. This would help to avoid any unnecessary suspicion or animosity once the microphone was visibly introduced into the ward environment. If staff and parents / patients were informed some time before the equipment was introduced, it was unlikely that their behaviour would change as a result (known as the Hawthorne Effect, discussed in Section 3.2.1), as they would fully understand the reasons for the study and not feel under scrutiny themselves. A copy of the publicity poster can be seen in Appendix A.
6.8.3. Reverberation time measurements

As occupancy levels in Sky Ward reduced during the weekend, leaving several single rooms and one 4 bed bay empty, it was possible to take advantage of this and make some reverberation time measurements using an Impulse Response Method with balloon bursts as the source. Measurements could not be made in the common areas such as nurse stations and hallways, as the ward was still occupied in part and this would have disturbed both staff and remaining patients.
59
6.9. Questionnaire survey considerations

Information sheets and questionnaires were handed out to the staff on an away day out of the ward. This was suggested by the ward manager as it was thought that this approach would yield the best response rate. The sample set for the staff study was fairly small, as the total number of full time staff working on this ward was only 12. Every full time member of staff completed the questionnaire survey. Many of the children on this ward were too young to complete the questionnaire themselves. After discussion with the ward manager, it was decided that the questionnaires would be given to the accompanying parents, who were given the option of completing the questionnaires on their own or jointly with their child (age and medical condition permitting). It was felt important that the parent / patient had been present on the ward for over 24 hours to give sufficient time for the individual(s) to form an opinion of the noise environment during the day and the night. The first section of the patient questionnaire was changed slightly to capture information about both the parent and patient, with an additional question to establish whether the parent completed the questionnaire with or without input from their child. The patient information sheets and questionnaires were handed out by the ward clerk to the parents of the patients on the ward. In total 31 completed parent / patient questionnaire responses were received. Copies of the patient / parent and staff questionnaires are shown in Appendix A.
6.10. Overall acoustic survey results

Table 6.1 shows the locations and the time periods of all measurements made. It should be noted that 1 week is defined as one set of week day data (Monday to Friday), weekends having been excluded from the measurement period for the reasons given in Section 6.5.1. Table 6.1 Measurement location and time interval Position Nurse station 1 Nurse station 2 4-bed bay A 4-bed bay B Single room A Single room B Length of measurement period 2 consecutive weeks (10 days) 1 week (5 days) 2 consecutive weeks (10 days) 1 week (5 days) 1 week (5 days) 1 week (5 days)
Overall noise measurements of A-weighted equivalent sound pressure levels (LAeq) for 24 hours, day time and night time were recorded, the day and night periods being defined by the WHO guidelines (Berglund et al, 1999), where day time is specified as 07.00 to 23.00 and 60
night time as 23.00 to 07.00. Table 6.2 shows the average LAeq measured for 24 hour, day and night time periods at each location. Table 6.2 Average LAeq measured for 24 hour, day and night time periods at each location
Position in ward
Weekday average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Nurse Station 1 Week 1
56.6
58.3
47.2
54.3
56.0
46.2
58.9
60.4
51.6
4-Bed Bay A Week 1
50.2
51.7
43.4
4-Bed Bay A Week 2
52.3
54.0
41.9
4-Bed Bay B
50.4
52.2
39.5
Single Patient Room A
50.4
52.2
34.8
Single Patient Room B
56.6
58.2
47.8
A summary of the day and night time average levels (averaged over all the measurement days for each location) are presented in Table 6.2 are presented graphically in Figure 6.29 for clarity. It can be seen that without exception, all levels exceed those suggested in the WHO guidelines (30 dBA LAeq for day and night). Levels measured at the nurse stations and in the 4-bed bays are shown to be fairly consistent, with night time levels on average 10 dB lower than those measured during the day. Single patient rooms, however, are much less consistent during both the day and night. Detailed results of levels measured at the nurse stations are discussed in the next section, with further results from the 4-bed bays shown in Section 6.12 and from single patient rooms in Section 6.13.
61
70
Day time Night time

60
50
Sound Pressure (dBA)
40
30
20
10
0 Nurse Station 1 Nurse Station 2 Single Room A Single Room B 4 Bed Bay A 4 Bed Bay B
Figure 6.29 Average day and night LAeq levels measured at each location
6.11. Nurse stations

Figure 6.30, shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours at the two nurse stations (means of total number of measurement days). It can be seen that the levels at nurse station 2 are consistently higher than those measured at nurse station 1. This is as we would expect as this nurse station is significantly larger in size, with more staff working in this area. The levels follow very consistent patterns suggesting that the daily ward routines which contribute to the noise levels are similar at both locations. Interestingly, background levels at night (shown in terms of LA90,1hr) are slightly lower for nurse station 2, even though the LAeq,1hr levels are higher at this nurse station. This could be affected by the level of airflow from the mechanical ventilation system. Although controlled centrally in the ward, noticeable differences were found between levels of airflow in different ward locations. This is discussed further in Section 6.15.2.
62
70 Night time Day time
Sound Pressure (LAeq, 1hr)
60
50
40
30
Time (24h:00) Nurse Station 1 LAeq Nurse Station 2 LAeq Nurse Station 1 LA90 Nurse Station 2 LA90
Figure 6.30
Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse stations
6.11.1. Sources of high level noise

As discussed in Section 6.8.1, sources of high level noise were identified from the trigger files captured by the sound level meter during the measurement period. The threshold for this data capture was set to 70 dB LAmax. By reviewing each file it was possible to build up an understanding of the types of high level noise sources present, and, by analysing the data further, understand the impact of a particular noise event on the average noise levels. A summary of high level noise sources identified at the nurse stations is given below: Staff to staff conversation Staff talking on the telephone Staff talking with patients Patients talking Ward doorbell Nurse call Internal telephone ringing Patients crying out Desk drawers The ward manager was particularly interested in capturing occurrences of a number of ward systems which were considered to be excessively loud. The nurse call system, internal telephone and the door bell were all cited. Trigger files collected at the nurse stations were analysed to find the average maximum levels of these systems. This data was captured during the first three weeks of the pilot study and in some cases was incomplete due to initial equipment configuration problems. However, 63 Footsteps Laughter Furniture scraping on the floor Coughing Replacing the telephone receiver Medical equipment alarms Mobile phones ringing Closing ring binders
Acoustic Design for Inpatient Facilities in Hospitals Pilot Study ___________________________________________________________________________________________________________________
enough data was collected to provide a good indication of the true levels of these systems. This is discussed further in the following paragraphs. Nurse call When a patient presses the nurse call button by their bed, a light flashes outside their bay or room, a tone is emitted through the speaker behind each nurse station and information is displayed on the console on the nurse station desk, as shown in Figure 6.31. The tone continues until the nurse attends to the patient and cancels the call by the bedside.
Figure 6.31 Nurse call console The microphone at nurse station 1 was positioned 3 m from the wall speaker. The levels and number of occurrences of the nurse call which created trigger files were noted and the maximum levels were arithmetically averaged over the five day measurement period. The distribution of the LAmax levels is shown in Figure 6.32. In total there were 115 instances of the nurse call tone captured, resulting in an average maximum value of 81.3 dB LAmax. The highest percentage (45%) of occurrences fell into the LAmax level category of 82 to 84 dB. There was no noticeable difference between the levels measured during the day or night, which was interesting as it was thought that the system had a night time setting which lowered the volume of the tone emitted.
60
50
Number of occurences
40
30
20
10
0 72 - 74 74 -76 76 -78 78 - 80 80 - 82 82 - 84 84 - 86
LAmax range (dB)
Figure 6.32
The number and levels (LAmax) of occurrences of the nurse call system at nurse station 1, measured at 3 m over 5 days 64
Only a small amount of data was available for review from nurse station 2. In total 13 instances of the nurse call were captured over a 19 hour period, again with the microphone positioned 3 m away from the wall speaker. An average maximum value of 80.2 dB LAmax was calculated. The highest percentage (54%) of occurrences fell into the LAmax level category of 82 - 84 dB as with nurse station1. Internal Telephone The microphone at nurse station 1 was positioned 3 m from the telephone on the desk. Over the five day measurement period there were ten separate occurrences of the ringing telephone which created trigger files. The resulting average maximum of the internal telephone was found to be 72.3 dB LAmax. As the average maximum was close to the trigger threshold of 70 dB LAmax, it is felt that some occurrences of the ringing telephone may not have been captured, so as such this figure may not be accurate. No data exists for the internal telephone at nurse station 2, but it is expected that similar levels would have been measured. Ward Doorbell When the ward reception desk is unmanned, ward visitors ring the doorbell and a member of staff remotely opens the door to the ward, which is locked for security purposes. As with the nurse call, the ward doorbell is piped to the small speaker mounted behind each nurse station. Only a small amount of data was available for review in this case. This was for nurse station 2 with the microphone positioned 3 m from the speaker. In total, 42 instances of the doorbell were captured over a 19 hour period, resulting in an average maximum value of 80.6 dB LAmax. The highest percentage (52%) of occurrences fell into the LAmax level category of 80 to 82 dB. The distribution of the LAmax levels is shown in Fig 6.33.
25
20
Number of occurences
15
10
0 74 -76 76 -78 78 - 80 80 - 82 82 - 84 84 - 86
LAmax range (dB)
Figure 6.33 The number and levels (LAmax) of occurrences of the ward doorbell at nurse station 2, measured at 3 m over 19 hours 65
Although data was only available from nurse station 2, it is expected that similar levels would have also been measured at nurse station 1. All the reported levels for the three ward systems are typical of those to which a staff member would be exposed when sitting at the nurse station desk. Figure 6.34 illustrates the difference between these calculated maximum levels and the average day and night time levels measured at the nurse stations. It can be seen that the average LAmax levels exceed the day time LAeq,16hr by between 14 and 23 dB, and the night time LAeq,8hr by between 24 and 33 dB.
90 85 80 75
80.8 dB LAmax 72.3 dB LAmax
80.6 dB LAmax
70 65 60 55 50 45 40 35 30
Nurse Call
Internal Phone
Doorbell
Average day time level at nurse station: 58.2 dB LAeq,16hr Average night time level at nurse station: 48.3 dB LAeq,8hr Figure 6.34 Average LAmax of the nurse call system, internal telephone and ward doorbell When questioned, staff cited the ward doorbell, nurse call and internal phone as the most annoying sources of noise (see Section 6.16.2). Staff also rated these systems as the noise sources which most interfered with their ability to carry out their job effectively.
6.12. Four bed bays

Figure 6.35 shows the averaged LAeq,1hr and background levels (LA90,1hr) levels over 24 hours for the two 4-bed bays. It can be seen clearly that the averaged LAeq,1hr levels were very consistent over time, with a day time level of around 53 dB LAeq,16hr, over 20 dB higher than the WHO guidelines. The night time average was found to be 11 dB lower than that measured during the day, around 42 dB LAeq,8hr. This was still over 10 dB higher than the acceptable level stated in the WHO guidelines. Surprisingly, there is a 5 dB discrepancy in background levels between the two bays during the night. This can be partly explained because of the ventilation systems in use on the ward. Bay A is mechanically 66
ventilated only, and in bay B the windows can also be opened. It is possible that opened windows may account for part of this discrepancy. The figure also shows that the WHO day / night division is a poor fit, with noise levels tailing off earlier in the evening, at around 21.00. This of course could be attributed to the fact that this is a childrens ward and as such a day and night division specifically for a ward of this type may be more appropriate.
70 Night time 60 Day time
Sound Pressure (LAeq,1hr)
50
40
WHO GUIDELINES
30
20
Time (24h:00)
4 Bed Bay A LAeq
4 Bed Bay B LAeq
4-Bed Bay A LA90
4-Bed Bay B LA90
Figure 6.35 Average LAeq,1hr and LA90,1hr levels over 24 hours for 4-bed bays A and B

The sources of high level noise were identified from the trigger files captured by the sound level meter during the measurement periods. The overall numbers of files for the three periods were similar, with 942 and 1002 files recorded in 4-bed bay A during measurement weeks 1 and 2 respectively, and 870 recorded in 4-bed bay B. Building up a full picture of the sources of high level noise by reviewing the trigger files was at times difficult. Many of the sources were fairly close to the microphone making it hard to build up a complete picture of what was happening in the bay. As will be seen in Section 6.13 this was not the case in the single rooms, where it was possible to identify some patterns with only one patient and parent present. Figure 6.36 shows the numbers of occurrences of each noise source type as a percentage of the total number of files captured. It can be seen that for each measurement period a high percentage of trigger files were listed as unidentified. These were commonly caused by visits by a clinician to a patient. Often these visits were to provide patients with their medication; to undertake medical examinations; or to redress wounds. During this time bed rails were moved and beds were readjusted 67
and repositioned, leading to the creation of a number of trigger files which were very difficult to identify accurately. With younger patients especially, a visit by a clinician often led to a great deal of screaming or crying. Patients were then made comfortable, leading to another set of high level noises which were again difficult to identify and categorise. On occasions it was obvious that certain noise events were related to a patient procedure, and were noted as such; however this was not always the case. Specific high level noise events which were clearly identifiable in these two bays were the door to the ensuite facilities in 4-bed bay A (because of the loud locking mechanism); medical equipment alarms; patients crying out; conversation; coughs and sneezes; and the use of rubbish bins, both in the ensuite shower room and on the ward.
Children's entertainer Parent shouting f or nurse Patient vomiting Accessing lockers Patient procedures Door to ensuite bathroom Visiting time Meal time Cleaning Desk drawers / cupboard doors Squeak of shoes on f loor Furniture Scraping Medical Equipment Parents / patients talking Curtains Crying Laughter Dustbin Cough / sneeze Conversation between staf f and parents Conversation between staf f Unidentif iable 0 10
4 Bed Bay A Week 1 4 Bed Bay A Week 2
20
4 Bed Bay B
30
Figure 6.36 Percentages of high level noise events by type measured in 4-bed bays A and B Figure 6.36 shows that particularly large differences can be seen between the percentages of trigger files created over the measurement intervals in the two bays, especially those caused by patient procedures, medical equipment and patients crying. All these are of course dependent on the severity and type of the patients condition. Patients in Sky Ward undergo different levels of surgery and as such it is not surprising to find large variation in the numbers of occurrences of high level noise events. There are notable differences in the use of the door to the ensuite bathroom from week 1 to week 2 in 4-bed bay A. The use of the ensuite may be related in part to the mobility of the patients and also to the numbers of parents staying on the ward, so these factors may account for the change. 68
As discussed earlier in this section, many of the trigger files categorised as unidentified are as a result of visits by clinicians. It is interesting to note that the week with the highest percentage of unidentified trigger events is the week with the highest percentage of medical equipment alarms and patient procedures, suggesting more clinical activity on the ward during this week. Surprisingly, trigger files caused by visiting time are only captured during week 2 in 4-bed bay A. On reflection, this may be very dependent on the location of the microphone. Only visitors to the bed situated next to the microphone are likely to cause noise of a sufficient level for the trigger files to be created.
6.13. Single patient rooms

Figure 6.37 shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours for the two single patient rooms. It can be seen that, unlike the 4-bed bays, the levels in the two single patient rooms were very inconsistent, with an average difference in the LAeq levels of 6 dB during the day and 13 dB at night. The night time background levels in single room B can be seen to be low, at around 29 dB LA90,1hr, around 5 dB less than for single room A. This may suggest differences in the consistency of airflow of the mechanical ventilation system, which is discussed further in Section 6.15.2. The dotted line indicates the WHO specified average noise level for ward accommodation. It can be seen that the night time levels in single patient room A were close to the recommended WHO levels, but levels at other times were much higher, with a day time average LAeq,16hr of 58.2 dB in single room B; almost 30 dB higher than the WHO recommendations.
Sound Pressure (LAeq, 1hr)
50
40
30
WHO GUIDELINES
20
Time (24h:00) Single Room A LAeq Single Room B LAeq Single Room A LA90 Single Room B LA90
Figure 6.37 Average LAeq,1hr and LA90,1hr levels over 24 hours for single patient rooms A and B
69
The next section looks at the sources of high level noise in each room in an attempt to understand the inconsistencies found.

By identifying the sources of high level noise from the trigger files, it was possible to build up a picture of the type of events which caused the measured levels to be so different between the two single rooms. Figure 6.38 shows the percentages of each noise source type as a percentage of the total number of files captured. The overall numbers of high level noise events in each room were very different. There were nearly five times more recorded high level noise events in room B (2898) than room A (608). This can be explained in part by the severity and type of the patients condition. The patient occupying room B for the majority of the measurement period required a large amount of clinical intervention during their stay which resulted in 35% of all high level noise events being attributed to medical equipment alarms and 16% to patient procedures (visits by clinicians to provide a level of care). It can also be seen that the use of televisions and mobile phones caused a number of high level noise events in room B, but not in room A. During the measurement period in room B there were 124 instances (4%) where the television level was greater than 70 dB, and there were 224 instances (8%) where conversations on mobile phones were measured at this level or above.
Patient procedures TV Talking on mobile phone Visiting time Meal time Cleaning Furniture Scraping Medical Equipment Parents / patients talking Crying Laughter Dustbin Cough / sneeze Conversation between staff and Conversation between staff Unidentifiable 0 5 10 15 20 25 30 35
% occurance of event type Single Patient Room B Single Patient Room A
Figure 6.38 Percentages of high level noise events by type for single patient rooms A & B To illustrate the impact of certain high level noise events on the average noise level within a room, certain typical events in single patient room A were analyzed in further detail. Table 6.3 presents the 70
event LAeq and LAmax. To put these levels into context the day time average noise levels measured in this bay were 52.2 dB LAeq,16hr. Table 6.3 Average and maximum noise levels of identified events in single room A
Event (dB)
Fluid Pump Alarm Room Cleaning Visit to a patient Bin Bag Changing Patient Procedure Rubbish Bin
Event LAeq
71.6 54.4 57.6 60.4 59.2 Impulsive
Event LAmax
89.1 79.3 75.7 75.5 81.5 80.6
Level above room day time LAeq LAeq

19.4 2.2 5.4 8.2 7.0 Impulsive
LAmax
36.9 27.1 23.5 23.3 29.3 28.4
It must be stressed that the events in Table 6.2 are shown for illustration purposes. They are not necessarily representative of every event of that type. However, it is interesting to see that the maximum noise levels measured were as high as 89 dB LAmax. Both the fluid pump and the rubbish bins caused levels which were over 80 dB LAmax and exceeded the average day time noise level by nearly 30 dB.
6.14. Establishing a representative measurement interval

One aim of the pilot study was to establish a representative measurement period which could then be used throughout the main study. Many reviewed studies measured noise levels for a single 24 hour period. It was considered unlikely that a randomly chosen 24 hour interval would be representative of typical noise levels on a hospital ward. This is further illustrated by Figure 6.39 which shows the noise level fluctuations over 24 hours for five days at nurse station 1. It can be seen that if, for example, Thursday had been chosen for a 24 hour measurement interval, it would have yielded very different results than if Tuesday had been chosen. In fact there would be a difference of 5.4 dB during the night time LAeq,8hr and 5 dB difference during the day time LAeq,16hr.
71
70
60
50
40
30
Time (24h:00)
Friday LAeq
Monday LAeq
Tuesday LAeq
Wednesday LAeq
Thursday LAeq
Figure 6.39
LAeq,1hr levels measured over five consecutive days at nurse station 1
Given that a single 24 hour period did not appear to be representative, a five day measurement period was then considered (weekdays only due to low occupancy during weekends). Data was available for two consecutive five day intervals at both nurse station 1 and 4-bed bay A. Figure 6.40 shows arithmetically averaged LAeq,1hr values from Monday to Friday, for two consecutive weeks at nurse station 1. It can clearly be seen that the averaged levels are similar in both level and fluctuation. A goodness of fit test showed that the two datasets do not differ significantly at the 1% level. Therefore it may be assumed that a five day measurement period gives reliably representative data to describe the noise climate at this nurse station.
70
Night time 60
50
40 Day time 30
Time (24h:00) Nurse Station 1 Week 1 Nurse Station 1 Week 2
Figure 6.40
Average LAeq,1hr levels over 24 hours for week 1 and week 2 at nurse station 1
72
Data was also available for two consecutive five day intervals for 4-bed bay A. Figure 6.41 shows averaged LAeq,1hr values from Monday to Friday, over a period of two consecutive weeks in 4-bed bay A. It can clearly be seen that the averaged levels are similar in both level and fluctuation, again suggesting that a weekday measurement interval may be representative. The goodness of fit test again showed no statistically significant difference between the two datasets at the 1% level.
70 Night time 60
50
40
30 Day time 20
Time (24h:00) 4-Bed Bay A Week 1 4-Bed Bay A Week 2
Figure 6.41
Average LAeq,1hr levels over 24 hours for week 1 and week 2 for 4-bed bay A
It has therefore been shown that, as with the nurse station, a five day measurement period is a suitably representative interval for 4-bed bay A. These results suggest that a five day period at a nurse station or in patient accommodation is sufficiently long to give reliable noise level data.
6.15. Other measured acoustic parameters

6.15.1. Reverberation times
It was decided that it would be possible to make some reverberation time (RT) measurements in the unoccupied ward accommodation during the weekend, without causing undue disturbance to staff and patients. Balloon bursts from thick latex 14 cm balloons were chosen as the noise source. As discussed in Section 5.3.8 the number of source and receiver positions used were in accordance with the British Standard BS EN ISO 3382-2 (1998) for engineering work. Table 6.4 shows the measured RT20 values (to the nearest 0.05 s) in three rooms, together with room volume, ceiling area (which is the main area of acoustic absorbency) and the glazing area. This building was built in line with the previous acoustic design guidance, HTM 2045 (NHS Estates, 1996), 73
as discussed in Section 6.4. All the measured RT values would be considered to be very low; well within the recommendations of this guidance. Table 6.4 Reverberation times measured in different ward accommodation
Room Description
Volume (m3) 168.5 57.7 40.3
Ceiling area (m2) 62.4 19.5 14.9
Glazing area (m2) 9.1 10.5 5.6
RT20 (s) @ 1kHz
4-bed bay B Single room A Single room B
0.28
0.32 0.26
It can be seen that the RT in single room A is slightly longer that the RTs in the other two areas. This is probably due to the larger area of glazing in room A, as described in Section 6.6.3.
6.15.2. Ambient noise levels

30 second measurements of ambient noise levels were made in two empty single patient rooms and one empty 4-bed bay during the day time. The results are shown in Table 6.4. Table 6.5 Ambient noise levels measured in unoccupied patient accommodation
Room description
Ambient noise level (LAeq) 35.0 dB
4-Bed Bay B windows closed
4-Bed Bay B windows open
37.5 dB
Single patient room A windows closed
31.6 dB
Single Patient Room B windows closed
37.7 dB
Single patient room B windows open
40.3 dB
Table 6.5 shows a difference of 6.1 dB between the quietest and noisiest rooms with the windows closed. This is thought to be predominantly due to the amounts of low level air flow through the ceiling vents. In some rooms the air flow was much more noticeable than in others. The mechanical ventilation system on this ward is controlled centrally, and not on a room-by-room basis, and so it would be expected that airflow in all rooms would be constant. However, in discussions with ward 74
staff, the differences in air flow and also heat levels were mentioned on several occasions. This suggests the mechanical ventilation system may not be working as designed. Rooms situated on the west side of the building have windows that can be opened. It can be seen in Table 6.4 that there is an increase in ambient noise levels of approximately 2.5 dB when a window is opened. This increase is thought to be mainly caused by traffic noise, as the hospital is situated in central London.
6.16. Results of the staff questionnaire surveys

The design of the questionnaire surveys was discussed in detail in Section 5.4 and the administration of the questionnaires in Section 6.9. In total 12 staff completed the questionnaires. The following sections discuss results from the questionnaires and show the differences between staff perceptions.
6.16.1. Staff profile

To establish certain attributes about the staff, the first section posed a number of basic questions. Out of the 12 respondents only 17% were male and 83% were female. The majority of staff were relatively young, with 75% in the age group 20-30, 17% in the age range 31-40, and 8% in the age range 4150. The average length of time that staff had worked on the ward was just over 2 years, but the length of time working at Great Ormond Street Hospital was longer for some staff, suggesting internal transfer from other wards.
6.16.2. Noise annoyance and interference

General feelings of noise annoyance were investigated by asking staff to what extent they were annoyed by noise. It can be seen from Figure 6.41 that 18% of respondents felt moderately annoyed by noise, with 45% of respondents very much annoyed by noise in their work environment. Thus in total 63% of staff were moderately or greatly annoyed by noise. It is interesting to note that no-one selected the extremely annoyed category. It should be remembered that all 12 permanent members of staff on the ward completed the questionnaire so it is not the case that responses were received only from those concerned about noise.
75
Figure 6.41 Distribution of the extent of staff annoyance Staff were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Figure 6.42 shows the percentages of staff who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little annoyed by the event.
Visiting time Talking on mobile phones Mobile phones ringing TV / radio Meal times Trolleys Rubbish bins Cleaning Medical Equipment Footsteps Doorbell Nurse call General conversation Staff talking on the telephone Internal telephone Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
% of staff rating annoyance event 2 or above
Figure 6.42 Percentage of staff rating an annoyance noise event with a 2, 3 or 4 Figure 6.42 clearly shows that the four most annoying sources of noise to staff are the ward doorbell, nurse call, internal telephone and medical equipment alarms, which were all rated by over 80% of staff as annoying. This response is consistent with the views of the ward manager with regards to 76
these systems. Visiting time and talking on mobile phones are found to be the next most annoying sources of noise, rated by over 60% of staff. Respondents were asked to what extent noise interfered with their ability to work effectively. Figure 6.43 shows that 42% felt noise moderately interfered with their ability to work; with only 8% feeling that noise interfered very much. It appears from these responses that noise interference is perceived to be less of an issue than noise annoyance.
Figure 6.43 Distribution of the extent of noise interference with work Staff were also asked to rate how much each noise event interfered with their ability to carry out their job effectively (again the rating scale of 0 to 4 was used). Figure 6.44 shows the percentage of staff who rated a noise event with a 2, 3 or 4, and as such it could be said that this noise event interfered to some extent with their ability to carry out their job effectively. Figure 6.44 shows clearly that as with noise annoyance, the four sources of noise which were felt to cause the most interference were the nurse call and internal telephone (both rated by over 80% of respondents), and the ward doorbell and medical equipment alarms, (both rated by over 70% of respondents). Talking on mobile phones is ranked fifth, as with noise annoyance, with 50% of respondents finding this activity interferes with their ability to carry out their job effectively.
77
80.6 dB LAmax 77.9 dB LAmax
72.3 dB LAmax
% of staff rating interference event 2 or above
Figure 6.44 Percentage of staff rating an interference noise event with a 2, 3 or 4 The corresponding LAmax values for the doorbell, nurse call and internal telephone are indicated on the figure. It can be seen that all the values for the doorbell and nurse call are higher than would be expected in a ward environment, which is again consistent with the views of the ward manager.
6.16.3. Important sounds

To aid understanding of which sounds were felt by staff to be important to be heard in order to carry out their jobs effectively, staff were asked to rate different noise events on a scale of 0 to 4, where 0 indicated not at all important and 4 indicated extremely important. It can be seen in Figure 6.45 that medical equipment alarms are considered by staff to be the most important noise events with a mean value of nearly 3.5 out of the maximum 4, followed by the nurse call and patients calling out with means of around 3.0. Given the annoyance / interference felt by the staff with regards to the nurse call, perhaps a different type of alert is more suitable, perhaps making use of silent technologies such as a personal handset which vibrates. Given the necessity for staff to be aware of the nurse call, careful consideration must be given to ensure that a suitable system is found that is not so subtle that it could be missed by staff.
78
Figure 6.45 Mean importance rating of certain noise events
6.17. Patient questionnaires

In total 31 parents / patients completed the questionnaires. The following sections discuss results from questionnaires and examine the parent / patient perceptions of the noise environment.
6.17.1. Parent / patient profile

Out of the completed questionnaires, 55% were filled out by parents alone and 45% by the parent with input from their child. Of those questioned, 10% of the parents were male and 90% were female, with 39% male patients and 61% female. Figures 6.46 and 6.47 show the distribution of ages of both parents and patients. It can be seen in Figure 6.46 that the majority of parents were aged between 31 and 50 years, with Figure 6.47 showing a fairly even split in relation to the patients age, apart from children under five years old.
50
40
% of repondents
% of repondents
40 30 20 10 0 20-30 31-40 41-50 51-60 60+
30 20 10 0 <5 5-10 11-13 14-18
Age range (years)
Age range (years)
Figure 6.46 Parents by age bracket
Figure 6.47 Patients by age bracket
Out of those questioned, 87% were staying in a 4-bed bay, with an average length of stay of four days. 79
6.17.2. Noise annoyance

The next section of the questionnaire considered day time noise annoyance and night time disturbance. The questionnaire sought to identify the sources of noise that may annoy or disturb patients. Respondents were given two lists of noises and were asked to rate the day time annoyance and night time disturbance on a scale of 0 to 4 (where 0 indicated no annoyance / disturbance and 4 indicated a great deal). Several lines were left blank at the bottom of the lists for patients to add and rate additional noise sources. Parents / patients were first asked how they perceived the day time noise environment on the ward. Figure 6.48 shows that the highest number of those questioned, 58%, felt that the ward was a little noisy during the day, 32% felt that the ward was quiet or very quiet, while 10% felt it was very noisy. It may be interesting to note that no-one selected the extremely noisy category.
Day time annoyance
Figure 6.48 Distribution of the extent of parent / patient annoyance during the day time
Although there was a high percentage of people that considered the ward to be a little noisy only 23% of respondents were actually annoyed by noise during the daytime. The patients who had indicated that they were annoyed by noise during the day, were then asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Figure 6.49 shows the percentage of patients who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little annoyed by the event. As can be seen clearly the percentages of those annoyed by any events were very low, with the largest percentage of respondents (25%) annoyed by medical equipment alarms, followed by noise from TV and radio use rated by 19%, and mobile phones ringing rated as annoying by 16% of respondents.
80
% of parents / patients rating annoyance event 2 or above
Figure 6.49 Percentage of parents / patients rating an annoyance noise event with a 2, 3 or 4 Patients were next asked how they perceived the night time noise environment on the ward. Figure 6.50 details the responses, showing that during the night 45% of those questioned felt that the ward was either very quiet or quiet. The highest percentage, 42%, found the ward to be a little noisy and 13% felt that the ward was very or extremely noisy.
Night time disturbance
Figure 6.50 Distribution of the extent of parent / patient disturbance during the night time
81
When questioned whether they were actually disturbed by noise at night, 70% of respondents said they were disturbed. Patients who had indicated that they were disturbed by noise during the night were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Figure 6.51 shows the percentage of patients who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little disturbed by the event.
Other patients cying out Talking on mobile phones Mobile phones ringing TV / radio Trolleys Rubbish bins Medical Equipment Footsteps Doorbell Nurse call General conversation Staff talking on the telephone Internal telephone Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
% of parents / patients rating disturbance event 2 or above
Figure 6.51 Percentage of parents / patients rating a disturbance noise event with a 2, 3 or 4 As with day time annoyance, medical equipment alarms were again rated as the most disturbing noise source, in this case, by over 50% of respondents. The second most disturbing noise source was banging doors (35%), followed by TV and radio usage (29%).
6.17.3. Positive sounds

Looking at sound in a positive rather than in a negative light, parents / patients were asked if there were any sounds that they found comforting. 94% of answers were left blank with only two completed responses: the footsteps of nurse coming to reset instruments; and people talking / relatives visiting. Respondents were asked if they felt that there was ever too little sound in a room. Only 7% of those completing the question felt there was, but surprisingly these respondents were in 4-bed bays.
82
6.17.4. Privacy and ease of hearing

Parents / patients were asked whether high background noise might make it difficult to hear doctors and nurses who talk to them. The majority of respondents, 81%, said they could always clearly hear what people say, with only 19% feeling that occasionally high levels of noise can make it hard to hear. Interestingly one of these respondents was staying in a single room. No respondent reported any hearing impairment. Conversational privacy was investigated by asking whether the parent / patient felt that they could have a private conversation at their bedside. 29% said that they did not feel they could speak privately, all of whom were in 4-bed bays. Out of those who said they felt they could speak privately, 38% of people said they would use their normal voice, with 62% feeling that they would need to lower their voice.
6.17.5. Patients questionnaire comments

Parents / patients were invited to make additional comments at the end of the questionnaire if they wished. Many of the comments made were in relation to the use of radios and TVs without headphones. A detailed list of these comments is shown in Appendix B.
6.18. Summary of results

This section summarises the main findings from the pilot study: Measurements made at the nurse stations and 4-bed bays were found to be consistent, with day time levels at the nurse stations around 58 dB LAeq and levels in the 4-bed bays around 52 dB LAeq. Levels measured in the single patient rooms averaged around 50 dB LAeq, but were much less consistent. The differences in measured levels were affected in part by the amount of care required by the patient, and by the behaviour of the individuals in the room. For example, the use of television at high volume and mobile phone conversations both contributed to higher levels. It was more difficult to build up a complete picture of high level noise events in the multi-bed accommodation. In the single rooms it was possible to identify some patterns as there was generally one patient and parent present, but in the 4-bed bays it was thought that more localised sounds were creating the trigger files. Night time levels were in general found to be 10 dB lower than day time levels. Reverberation times measured in patient accommodation were very low, due in part to the good acoustic properties of the ceiling tiles used. However, the use of solid plaster ceilings in hallways and at the nurse stations may lead to longer reverberation times and higher noise levels in these areas. 83
Ventilation noise was on occasions noticeably loud, and was not controlled on a room-byroom basis. 63% of staff questioned were moderately or very much annoyed by noise in their work environment, and 50% felt noise moderately or very much interfered with their ability to work. Staff questionnaire responses consistently found that the internal telephone, nurse call, ward doorbell and medical equipment alarms caused the most annoyance and interference. This suggests that a level of noise control should potentially be applied to these systems, or alternatives sought. 70% of parents / patients were disturbed by noise at night, but only 23% annoyed by noise during the day. Medical equipment alarms, banging doors and the use of television and radio were rated as the most disturbing sources of noise to patients / parents at night. Soft door closers and the use of headphones for television and radio usage would be a simple and relatively cheap solution to reduce some of this annoyance. The use of medical equipment alarms could be studied further. Some alarms may be un-necessary or are set too loud; however, a careful balance must be sought as these alarms are rated as the most important noise sources for staff to hear. Lower percentages of parents / patients rated noise annoyance and disturbance events than the staff. This may be partly due to their not wanting to be critical as their child is unwell and they are grateful to the staff and hospital, or because they are focussed on their childs care and wellbeing and noise is less noticeable than it may be in other situations. The pilot study aimed to provide useful feedback for the redevelopment team and the ward manager on the site. Following its completion, a full report was provided for each team member involved, and the ward manager of Sky Ward was also informed of the relevant findings. Follow up meetings were held with ward staff, which are discussed in the following section, and further highlighted areas that could potentially be improved with regards to noise control.
6.19. Follow up discussions

Several meetings were held with the staff of Sky Ward to present the pilot study findings and discuss changes that could potentially lead to an improvement in certain areas. Members of the redevelopment team also attended the meetings. A summary of the discussions are shown below. Television usage on the wards Many of the patient comments indicated that one of the main issues was around the use of television / radio without the use of headphones. Further discussion with staff regarding this issue yielded some interesting findings and attitudes towards the use of televisions on the ward: 84
There are four flat screen TVs in each of the 4-bed bays, all suspended from the ceiling at the end of each patients bed. A remote control is supplied with each TV, but it appears that all remote controls are identical, so they control not only the patients TV but all four installed in the bay. The consequence of changing a channel or adjusting the volume on one television could be that the other televisions in the bay are affected. Staff also mentioned that the default volume when the televisions are switched on is set very loud. The type of television installed on the ward requires wired headphones, which is a difficult option as they are positioned so far from the bed head. There are some portable DVD players currently available which could be watched with headphones. Staff felt that many of the patients with special needs would not be able to wear headphones and they appeared slightly reluctant to enforce the general use of headphones. It was felt that some patients would be used to having the TV on all the time at home and it would not be fair to prevent the patient from behaving in the same way on the ward; even to the extent of falling asleep to the noise from the TV. Younger members of staff did not appear to consider extensive and loud use of TVs as unacceptable and did not seem to be aware that that one patients behaviour might negatively impact the others on the bay. Banging doors Banging doors were listed as a disturbance in the patient questionnaires. Staff mentioned that the quiet closers on most doors did not seem to work effectively, resulting in a loud thud as the door closed. This had recently become such a problem that staff have draped towels over doors to stop them banging (although this has since been stopped by the ward manager). The door of the dirty utility room was mentioned as being particularly loud and prompted further discussion on the difficulty of entering the dirty utility room whilst carrying spillable objects. It was felt that a kick bar on the base of the door might be more effective than an ordinary handle. The necessity for security for the utility room was also discussed with staff feeling that the use of a swipe card or key code system could potentially lead to even greater access problems. Many staff felt that it was unnecessary to have security for this room at all. Single patient room 11, the staff room and kitchen were all cited as having particularly loud doors. The doorbell The ward doorbell had been noted by the ward manager as being extremely loud and was shown to be a source of interference and annoyance for staff. As a consequence, a new system had been installed with a volume control. Although the ward manager had turned down this volume, no difference had been perceived by the staff, who still felt it was extremely loud, especially at night. The use of the doorbell was discussed further. Many of those ringing out of hours tended to be visiting or staying with their children on the ward and seem to ignore the sign asking for the bell to be rung 85
once. It was also mentioned that if someone leaves their finger on the buzzer for a long period, it continues to ring. This annoys the staff immensely, especially late at night when fewer staff are present on the ward and may be busy with other duties. Staff felt that it would be useful if the doorbell could somehow be limited to ring once every 30 seconds. The possibility of issuing passes was investigated. However, this has been trialled previously and many passes were never returned to the ward. With each pass costing between 5 and 10 this option was considered too expensive to be workable. It was also mentioned that the security camera pointed at the ward entrance, which is installed so that staff can identify those ringing the doorbell, is pointed at the back of the heads of those ringing. The cameras validity in terms of security was felt to be questionable. Nurse Call This was another system that was identified as being extremely loud, with no perceptible volume change between the day and night time setting. This was still considered to be problem by the ward staff. Miscellaneous Alarms The Controlled Drugs cupboard has both a light and an alarm. This alarm is found to be very annoying by the staff. Internal Telephones Internal phones at the nurse stations were also shown to be an annoyance and to interfere with staff duties. Staff felt that the phones ring a great deal and are loud. According to the members of staff present, they can be turned down with the exception of the emergency phone, which is a fixed volume (very loud). Various ideas for replacements were discussed, including bleepers and portable phones. If the ward clerk is away from her desk, the main ward phone diverts onto the ward and consequently the ward staff have to deal with the calls and pass messages back. The possibility of installing a voice mail system for the ward clerk was discussed and considered to be a positive idea by the staff. This is an illustration of a simple, cost effective solution that can be found by directed discussion. Doctors office alarm As discussed, Sky Ward is mechanically ventilated and the system is controlled centrally. Staff mentioned that certain rooms are particularly warm and some particularly cold, especially if the beds are situated directly under the ceiling vents. One room that is always very warm is the doctors office, which leads to the door being constantly propped open. With the door open a very high pitched alarm is activated, which is continually reset. This is extremely annoying to staff.
86
6.20. Conclusions
The main purpose of the pilot study was to trial the proposed methodologies to ensure that meaningful results could be obtained in line with the research proposal. Both the objective measurements and the subjective questionnaire surveys were generally felt to be successful, with a number of specific aspects discussed in further detail below. The pilot study showed that suitable microphone positions could be found, which would allow for meaningful comparisons. However a degree of flexibility was required so as to minimise the impact of the microphone and associated equipment on staff duties and patient care. Trigger files were successfully used to identify sources of high level noise, but analysis was found to be extremely time consuming. It was felt that further consideration was needed to ascertain the best way to present this data in the main study. A working week was shown to be a representative measurement interval. This interval was to be further validated during the main study if possible. Both staff and patient questionnaires generally worked well, with only two questions requiring a small amount of re-wording to ensure complete clarity. Both the study findings and results of the follow up meetings are currently being used to positively influence / inform the choice of systems and ward design for the next phase of the Great Ormond redevelopment. Improvement to some of the existing systems on Sky Ward is also being investigated, and it is hoped that some noise control measures can be taken. The following chapters describe the main study which involved noise and questionnaire surveys in a medical and surgical ward at Bedford Hospital and in three wards at Addenbrookes Hospital, Cambridge. As part of the Bedford Hospital study, a ceiling intervention study was also carried out and changes to sound levels and reverberation times were investigated. This is discussed in further detail in Chapter 8.
87
Acoustic Design for Inpatient Facilities in Hospitals Bedford Hospital ___________________________________________________________________________________________________________________
7. Bedford Hospital
7.1. Introduction
Two wards at Bedford Hospital were the subject of the main study, which took place over an eight month period, from April to November 2010. Working in collaboration with the Estates team, two inpatient wards of similar layout were chosen in the main five storey ward block. For comparison purposes one ward was a surgical ward, the other medical. Within the study time frame a refurbishment of the medical ward was also planned. This was of particular interest to the Estates team, who wanted to establish the effects of changing reflective ceiling tiles for those with good acoustic properties. The results of this change are discussed in further detail in Chapter 8. This chapter begins by looking at the background of Bedford Hospital, providing an overview of the acoustic design considerations of the ward block and exploring the hospital policies and equipment usage that may affect the noise levels in the study wards. The chapter continues by examining the two wards participating in the study individually, including their design layouts and the daily routines. Objective results from each ward are reported, and staff and patient perceptions of the noise environment are explored.
7.2. Background
Bedford Hospital opened in 1803, consisting of just six beds, and employing a staff of four clinicians. Now, over two centuries later, the hospital provides 403 patient beds and has a staff of over 2000. Services are provided for around 270,000 people in mid and north Bedfordshire and include medical, surgical, paediatric and neonatal wards; A&E; an Acute Assessment Unit (AAU); and a specialist cancer centre. One of the original buildings still exists, as can be seen in Figure 7.1, but there have been many additions, many of which were built in the 1970s and 1980s (as seen in Figures 7.2 and 7.3). The hospital continues to expand.
Figure 7.1 Original building, Bedford Hospital (1803)
91
Figure 7.2 Five storey ward block
Figure 7.3 Main hospital entrance
7.3. Building acoustic design considerations

The main five storey ward block was constructed in the early 1980s under Crown Immunity, which in effect means that this building was exempt from building regulations. It is unknown if any acoustic guidelines were taken into consideration during the build. The construction of the ward block is primarily concrete, naturally ventilated and single glazed. The two inpatient wards chosen to participate in the study are of identical layout, and situated a floor apart. Each ward is designed around a central corridor which runs down the length of the ward. The main patient accommodation is situated to one side of this corridor and consists of a number of open four and six bed bays. These bays look out over the main hospital entrance, car park and several connecting roads, one of which is a busy main road; although traffic flow is relatively slow due to traffic lights. On the opposite side of this main corridor are the staff healthcare utilities, offices, storerooms, a kitchen and four single patient rooms, which overlook the new maternity wing. Ward plans are shown in Sections 7.5 and 7.6. Both study wards have suspended ceiling grids with perforated acoustic ceiling tiles in the majority of the multi bed bays, single rooms, offices and corridors; although there are exceptions, further details of which are discussed in Section 7.6.2. Other acoustic absorbency is provided by the fabric privacy curtains, which can be pulled fully around each bed; upholstered easy chairs where patients can sit when out of bed; window curtaining; mattresses and bedding. All patient accommodation has heavy duty vinyl flooring and solid plastered walls.
7.4.
Hospital policies and equipment common to both wards
Before looking at the two study wards individually, the hospital policies and equipment that are common to both wards and may have an effect on noise levels, are examined.
7.4.1. Meal times

Protected meal times are in use throughout the hospital. This means that during breakfast, lunch and evening meal times clinical visits cease and visitors are asked to leave the ward (except family
92
members if they are providing assistance with eating). This is primarily to allow the hospital staff to monitor what is being eaten and to provide a more relaxing environment for the patients. Each ward has a kitchen which is used for plating up hot meals and for washing up plates and cutlery (shown in Figure 7.4). There are also fridges and a food warmer. The kitchen may be a source of noise for patients in the opposite bay, as the kitchen door is always left open. Meals and drinks are served to patients from a trolley, another potential source of noise.
Figure 7.4 Medical ward kitchen Meal times are as follows: Breakfast is served from 08.00 to 08.30 and is followed by a tea round. Breakfast is usually cold except for porridge Mid morning tea is served from 10.00 to 10.30 Lunch is from 12.30 to 13.15, with tea served at around 13.00. Lunch is usually a hot meal which is brought up to the ward kitchen in a heated trolley. Meals are then plated up before being served to patients one bay at a time on a smaller trolley. Afternoon tea is served at 14.30 Tea (supper) is generally a selection of sandwiches and is served from 17.30 to 18.15, followed by a tea round at 18.00 Drinks are available in the evening from 20.00 to 21.00
7.4.2. Ward design

At the entrance to each ward there is a ward clerks desk which acts as a reception area. Behind this is a staff room and kitchen, both of which can be noisy areas. However, the bay directly opposite the ward clerks desk is the only bay likely to be adversely affected by noise. The nurse station is situated halfway down the main corridor. Potential high level noise sources here are the nurse call, internal telephone, staff conversation and the carrying out of administrative tasks which may affect the nearby bays and single rooms. 93
7.4.3. Occupancy levels

Unlike the pilot study hospital, both these wards have a high occupancy rate throughout the entire week, including weekends. It was therefore considered valid that noise level measurements should be made at each location over a full seven day period, rather than the five day interval used during the pilot study.
7.4.4. Shift patterns

Staff day shifts start at 07.00 and end at 19.30, with night shifts starting at 19.00 and ending at 07.30. There is a half hour overlap at the beginning and end of the shifts. During this overlap there is usually a handover session in the staff room (opposite 4-bed bay 1). Higher levels of noise may be attributed to these changeover periods.
7.4.5. Visiting hours

These are officially 14.00 to 20.00. Attitudes amongst ward sisters vary regarding the enforcement of these hours. Some staff consider that the positive effect on patients of visitors outweighs the other negative aspects of having numbers of visitors on the ward out of official visiting hours.
7.4.6. Ward access

There is no ward security to prevent anyone walking on to the ward. This removes the need for a doorbell, which could potentially be a source of noise annoyance.
7.4.7. Access to patient accommodation

All multi bed bays on the wards are open to the corridor; there are no doors. The doors of the single patient rooms are generally left open during the day for observation purposes, but are closed at night and during visiting times, where there is someone on hand to call a nurse if required.
7.4.8. Cleaning staff

In-house cleaning staff are used throughout the hospital rather than contractors. It is felt that this enables more control of cleaning regimes and promotes better work ethics and loyalty. Each ward has a dedicated cleaner who works from 07.00 until 15.00. Duties include mopping, sweeping and floor buffing.
7.4.9. Mobile phone policy

The hospital policy specifies that mobile phones should only be used in the lobby areas and not on the wards. This is partly to avoid disturbance and partly because mobile phones have camera and recording capabilities which could potentially cause a breach of patient privacy or confidentiality
94
7.4.10. Entertainment systems

A HTS Hospicom system is provided at each patients bedside. This is a pre-pay TV / radio / telephone console. Patients are issued with headphones. A photograph of the console is shown in Figure 7.5.
7.4.11. Rubbish bins

Quiet closing foot operated rubbish bins are in use throughout the wards. However, the opening mechanisms are found to be quite noisy especially if the bins are placed too close to a wall or nearby object, which then tends to be hit by the lid with some force. The body of the bins are also metal and undamped; if a heavy object is dropped inside, this can also be a source of noise.
7.4.12. Staff call

The emergency staff call system is activated by the patient pushing a button at their bedside. This causes a red light to flash on the ceiling outside the bay and sounds an intermittent alarm at the nurse station. This will continue until cancelled by a member of staff.
7.4.13. Medical equipment alarms

Alarms generally sound if fluids are low or as a warning that a piece of equipment, for example a canular, has fallen out. Mattresses used on the wards bleep if they become under-inflated.
7.4.14. Trolleys
Meals, drinks rounds, medication and dressings are all taken to the patients bedside on trolleys. Beds themselves are on wheels, as is the majority of medical equipment for ease of movement. Deliveries of fresh linens, food and other ward supplies all arrive at the ward in wheeled metal cages or on trolleys.
7.4.15. Internal telephones

A number of internal telephones can be found on each ward, in the staff office, at the ward clerks desk, the nurse station and on other staff desks in the ward corridor.
7.4.16. Hand gels

Automatic hand gel dispensers have been installed on the ward. These are motorised and work using a sensor as shown in Figure 7.6.
95
Figure 7.5 Hospicom entertainment console
Figure 7.6 Automatic hand gel dispenser
7.5. Medical ward

This ward is located on the fourth floor of the main hospital building and specialises in Gastroenterology and care of the elderly. Elderly patients admitted to this ward are often confused and may be suffering from a degree of dementia. Generalised medical care is also provided if the beds are needed by patients from other specialties, for example Oncology. Patients range in age from 18 upwards. The majority of patients have come via their GP to the AAU (either in an ambulance or their own vehicle) and then to the medical ward following their assessment. Patients on this ward are generally recovering from infections. The ward has 30 beds in total, with three 6-bed bays, two 4-bed bays and 4 single rooms (an example of which is shown in Figure 7.7). Patients are predominantly male, with one all female 4-bed bay and single rooms used for female patients when necessary.
Figure 7.7 Single patient room The medical ward was the first ward at Bedford Hospital to participate in the study, and before commencement a number of issues needed to be clarified. An initial meeting was held with the ward manager and the head of the hospital infection control team, who were shown the sound measuring equipment. There were no undue concerns regarding the cleaning of the equipment as it would be 96
located in each position for only one week. It was felt that if it did become contaminated, the equipment case was such that it could easily be cleaned using an alcohol wipe. Measurement locations were also discussed and possible microphone positions that would be acceptable to the staff and patients were identified in the multi-bed bays, single rooms and at the nurse station. For comparison purposes it was important that the microphone could be placed in similar locations in each accommodation type, and for this to be repeatable in the surgical ward. A ward plan showing the microphone positions can be seen in Figure 7.10 on page 99. To minimise the risk of theft, ward staff felt that it would be sensible to hide the environmental case containing the sound level meter either in or behind a cupboard. Given the positioning of the ward furniture and the length of cable available (5 m), this meant that the microphone would need to be situated close to the edge of a room, often in the corner. Due consideration was given to the possible increase in sound pressure due to wall reflections or corner reflections. A number of tests were carried out to investigate this, and the results can be seen in Appendix C. No significant increase in measured level was found due to the location of the microphone close to a wall or in the corner of the room. To avoid the microphone being knocked or contaminated, and to be as unobtrusive as possible, it was felt that suspending the microphone from the ceiling would be ideal. A 300 mm bracket was designed which simply clipped around the T shaped ceiling grid without disturbing the ceiling tiles. If the ceiling bracket could not be used, for example in the case of a solid ceiling, the microphone was mounted on a small tripod which was securely fastened out of reach. Figure 7.8 shows the microphone suspended from the ceiling grid and Figure 7.9 shows the microphone and tripod positioned on a light above a mirror in a single room.
Figure 7.8 Microphone suspended from ceiling
Figure 7.9 Microphone on tripod
Questionnaires were reviewed by the ward manager and it was decided that they would be distributed by the ward clerk to those patients who had been on the ward for over 24 hours and were felt to be fit enough to complete the survey. Staff questionnaires were to be left for staff to complete in the staff room. 97
As with the pilot study a number of laminated advertising posters were displayed throughout the ward common areas. These posters were aimed at both staff and patients and explained in simple terms why and how the study was being undertaken. In addition to these posters the ward manager personally discussed the study with all her staff during staff meetings.
7.5.1.
Ward specific information
Staffing levels The nursing staff levels are highest during the morning with seven dedicated nursing staff. This drops to five during the afternoon and four at night. Other dedicated ward staff include a housekeeper; three domestics; a ward clerk; an occupational therapist; and a physiotherapist. Ward routines The first patient visits by clinicians for general observations and the administration of intravenous antibiotics begin at 06.00. However, activity on the ward does not begin fully until 07.30, when the main lights are switched on. Two ward rounds begin at 09.00 (Monday and Friday), with two consultants from different specialities: Care of the Elderly and Gastroenterology. As well as the consultant ward rounds, the junior doctors work in the ward throughout the day talking to patients, checking bloods and organising discharges. In addition to the doctors ward rounds, bloods are also taken by the phlebotomists at 09.00-09.30 and 18.00-19.00. The ward lights are dimmed after 21.00. Sources of noise specific to the medical ward Patients who are suffering from dementia are given a wrist tag which causes an alarm to sound if the patient tries to leave the ward. There is a pneumatic system for the distribution of pharmaceuticals close to the nurse station and a 6bed bay.
98
Figure 7.10
Detailed plan of the medical ward showing microphone positions
Pneumatic system
Microphone position Sink
Acoustic Design for Inpatient Facilities in Hospitals Bedford Hospital __________________________________________________________________________________________________________________
7.6. Medical ward overall noise survey results

Noise level measurements were made at eight different locations on the medical ward. Table 7.1 shows the locations, measurement periods and the patient genders where applicable. Table 7.1 Medical ward - measurement locations, time periods and patient gender Position Ward entrance Nurse station 4-bed bay 1 4-bed bay 2 6-bed bay 3 6-bed bay 4 Single room A Single room B Length of measurement period 7 days 7 days 12 days 7 days 6 days 8 days 5.5 days 7 days Patient gender N/A N/A Male Male Male Male M/F M/F
The results reported for bay 1 are those made after the ceiling tiles were changed for those with better acoustic properties, as discussed in Chapter 8. This ensured that reported results were comparable with the other bays on the ward with similar ceiling tiles. Overall measurements of A-weighted equivalent sound pressure levels (LAeq) for 24 hours, night and day time were recorded at each location and are shown in Table 7.2. Table 7.2 Average LAeq measured for 24 hour, day and night time periods at each location
Position in ward Average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Ward entrance Nurse station 4-bed bay 1 4-bed bay 2 6-bed bay 3 6-bed bay 4 Single room A Single room B
53.7 54.0 49.4 49.8 54.0 49.7 59.3 52.6
55.2 54.9 51.3 51.1 55.2 51.2 60.6 53.9
46.4 50.0 41.8 44.1 50.1 42.4 51.1 47.3
A summary of the day and night time average levels presented in Table 7.2 are shown graphically in Figure 7.11 for clarity. As with the pilot study ward, all levels in patient accommodation exceed those suggested in the WHO guidelines without exception. It can be seen that day time levels measured at the ward entrance and the nurse station are similar, as are the day time levels in bays 1, 2 and 4. However, levels in bay 3 and the single rooms are less consistent. The drop between day and night 100
time levels varies from 9.5 dB in bay 1, to only 5 dB in bay 3 and at the nurse station. Possible reasons for these differences are discussed in Section 7.6.4.
65
Day time LAeq, 16hr

60 55 50 45
Night time LAeq, 8hr
Sound Pressure (dB LAeq,1hr)
40 35 30 25 20 15 10 5 0
Ward entrance
Nurse station
4-bed bay 1
4-bed bay 2
6-bed bay 3
6-bed bay 4
Single room A
Single room B
Figure 7.11 Average day and night LAeq levels measured at each location The following sections examine the noise levels recorded at different locations on the ward in further detail.
7.6.1. Nurse station and ward entrance

Figure 7.12, shows the average measured LAeq,1hr and LA90,1hr levels over 24 hours for the nurse station and the ward entrance by the ward clerks desk. It can be seen that noise levels in both locations increase steadily from around 04.30am, and at the nurse station, levels do not decrease substantially until around 01.00. The levels at the ward entrance decrease earlier than at the nurse station and remain consistently lower during the night. This is as one would expect given the nurse station is staffed 24 hours a day, whereas the ward clerk is only at the desk during day time office hours, and then the desk is only used periodically by other staff. However noise levels at the ward entrance are also affected by activity in the staff room and kitchen which are situated directly behind the ward clerks desk area, as can be seen in the ward plan shown in Figure 7.10 on page 99. The measured LA90,1hr levels provide a good indication of the variation in background noise levels over time. Night time background levels reduce from around 38 dB LA90 to around 34 dB LA90 during the quietest period, while day time background levels are fairly steady at around 42 dB LA90. Background levels at the ward entrance exceed those at the nurse station twice during the day: first around the 101
time of the morning shift handover, breakfast and start of the morning ward rounds; and secondly during lunch time, when hot meals are plated up in the kitchen.
70
60 Night time
50
40
30 Day time 20
Time (24h:00)
Nurse Station LAeq Ward entrance LAeq Nurse Station LA90 Ward entrance LA90
Figure 7.12 Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station and ward entrance Viewing averaged noise levels over time, as in Figure 7.12 above, provides valuable information with regards to level consistency and overall day and night time variation patterns, but does not illustrate the fluctuating nature of noise in the short term. Figure 7.13 shows noise levels captured at the nurse station over a ten minute time interval with the microphone approximately 2 m away from the main desk area. Using the trigger files captured when LAmax exceeds 70 dB, certain high level noise events have been identified.
FURNITURE SCRAPING ON FLOOR
INTERMITENT NURSE CALL
CLOSING ROOM DOORS
LAmax,2s LAeq,2s
Figure 7.13
LAmax,2s and LAeq,2s fluctuating over a ten minute interval at the nurse station
102
The events shown in Figure 7.13 are a good representation of the types of high level noise events recorded at the nurse station. Furniture scraping is found to be a constant source of high level noise which could be simply and cheaply controlled by fitting rubber feet or wheels to the chairs used. Doors are also a problem, with doors to two single rooms situated at the back of the nurse station and a set of large metal storage cupboards to one side. These often cause high levels of noise on closing. Again this could be easily and cheaply rectified. To further illustrate the types and noise levels of typical high level events at the nurse station, examples are presented in Table 7.3. It should be noted that the levels shown are for individual events and may not be representative of every noise event of that type. Table 7.3 Examples of noise events at the nurse station
Noise event Nurse call Door banging Furniture scraping on floor Metal cupboard door Internal phone Rubbish bin
LAmax (dB) 72.8 77.3 85.6 84.3 71.6 78.7
7.6.2. Multi-bed bays

Figure 7.14 shows the averaged LAeq,1hr levels measured over 24 hours for two six bed and two four bed bays. It can be clearly seen that levels for bays 1, 2 and 4 are very similar and follow the same general pattern of fluctuation. This suggests firstly that an increase in ward size from four to six patients does not necessarily affect the noise levels, and secondly that the daily ward routines which contribute to the noise levels are comparable. The levels in bay 3, however, are consistently higher than those in the other bays, by 4 dB LAeq on average during the day and by up to 8 dB LAeq during the night. This particular bay is opposite the nurse station, and for observation purposes the patients with the most serious conditions are placed here. Due to the type of patients on this ward, the main impact on the noise levels is both from the patients themselves, for example, crying out, coughing and groaning and the increased clinical activity in the bay. Unlike bays 1, 2 and 4, bay 3 has a solid plaster ceiling rather than a suspended ceiling with acoustic tiles. This may also have a negative impact on noise level. Figure 7.14 also shows that the WHO day / night division is not a particularly good fit. Noise levels increase steadily from around 05.30 rather than 07.00, and begin to decrease after the evening meal is served and then further decrease at 23.00. This suggests it might be appropriate to redefine the day and night time periods for hospital noise assessment.
103
Sound Pressure (dB LAeq,1hr )
50
40
WHO GUIDELINES
30
20
Time (24h:00)
Medical 4-Bed Bay 1 Medical 4-Bed Bay 2 Medical 6-Bed Bay 3 Medical 6-Bed Bay 4
Figure 7.14
Average LAeq,1hr levels over 24 hours for the multi-bed bays
Background levels, in terms of LA90, are shown in Figure 7.15. Levels in bays 1, 2 and 4 have been averaged for purposes of clarity and it can be seen that background levels in these bays are around 39 dB LA90 during the day and 32 dB LA90 at night. Bay 3 has considerably higher background levels of around 46 dB LA90 during the day and 41 dB LA90 at night, higher than those at the nurse station opposite (see Figure 7.12).
50
40
30
20
Time (24h:00)
Mean LA90 Bays 1,2,4 Mean LA90 Bay 3
Figure 7.15 Average LA90,1hr levels over 24 hours for the multi-bed bays
104

Figure 7.16 shows the average LAeq,1hr levels over 24 hours in the two single rooms measured. The levels measured in the multi-bed bays have been averaged, and the average is also shown on the graph to allow for comparison between single rooms and multi-bed bays. It can be seen that levels for single room B are similar to those of the multi-bed bays, but noise levels measured in single room A are considerably higher. Much of this high level noise was caused by visits to the patient by groups of relations who would arrive around 14.00 and often stay until 21.00. During this time conversation was constant. Noise levels were also impacted by the patients condition and a lack of cooperation with members of the nursing staff. Consideration should also be given to the fact that room A has a solid plaster ceiling rather than a suspended ceiling with acoustic tiles. This may also have had a negative impact on noise levels.
50
40
WHO GUIDELINES
30
20
Time (24h:00)
Medical ward - single room A Medical ward - single room B Average of multi bed bays
Figure 7.16
Average LAeq,1hr levels over 24 hours for the single rooms
Figure 7.17 shows the LAmax,2s and LAeq,2s between 14.00 and 17.00 in single room A and illustrates the impact of visiting time on the noise levels. Average noise levels during this period increase to 66 dB LAeq, with an occurrence of 89.3 dB LAmax.
105
VISITING TIME
Figure 7.17
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a three hour period in single room A
7.6.4. Further analysis of high level noise sources

To help build up a further picture of the sources of high level noise at each location, the numbers of occurrences of LAmax in 5 dB bands from 70 to 95 dB have been examined. Figures 7.18 and 7.19 show the average number of high level noise events per day and per night in different measurement locations.
500
70 LAmax < 75 dB
450
75 LAmax < 80 dB 80 LAmax < 85dB
Number of recorded noise events by category
400
85 LAmax < 90dB 90 LAmax < 95 dB
350
300
250
200
150
100
50
Figure 7.18 Average number of high level noise events recorded at each location per day
106
It can be seen that the numbers of high level noise events in single room A during the day exceed those in all other locations, with on average, over 450 events with LAmax between 70 and 75 dB; over 250 events with LAmax between 75 and 80; and over 100 events with LAmax above 80 dB. As discussed in the previous section, the majority of these high level events are due to loud conversation during visiting time, which raises the following question: If this patient had been in a multi-bed bay, would the visitors have felt more inclined to speak more quietly? The ward manager in this medical ward has a positive view on the benefits of visiting time, and does not strictly enforce visiting hours. However, are these long periods of loud conversation actually beneficial to the patient? Would these lengthy visits have been curtailed if this had been a daily occurrence in a multi-bed bay? Although, as discussed in Section 7.6.2, overall noise levels in bays 1, 2 and 4 are shown to be similar, differences can be seen between the bays in terms of the average numbers of high level noise events. Bay 1 is opposite the ward clerks desk, kitchen and staff room and is shown to have on average over 50 more high level noise events during the day than the next bay along the corridor. This is thought to be due to noise from the ward clerks desk area, kitchen and staff room and this is further confirmed by patient comments in the questionnaire surveys, as discussed in Section 7.11. Day and night time average noise levels for bay 3 and the nurse station which is opposite are very similar, as are the numbers of high level noise events. However, by looking at the overall noise levels it is unclear whether any of the activities at the nurse station have an adverse effect on noise levels in the opposite bay or whether they are unrelated. Further investigation appears to suggest that as first thought; the majority of sources of high level noise in bay 3 appear to be linked to increased clinical activity and patients conditions and behaviour. The nurse call appears to be the only nurse station related activity which is recorded at comparable levels in bay 3. Figure 7.19 shows a very different pattern of occurrences of high level noise events during the night. In this case it is the nurse station and bay 3 which show the highest numbers of events. This is unsurprising as the nurse station is manned for 24 hours a day, and bay 3 is used for more seriously ill patients who require constant care. Sources of high level noise at the nurse station are the nurse call, furniture scraping on the floor, doors banging and administrative tasks, while patients crying out and clinical activity are typical of high level noise events in bay 3.
107
90 70 LAmax < 75 dB 80 75 LAmax < 80 dB 80 LAmax < 85dB 85 LAmax < 90dB
70
60
50
40
30
20
10
Figure 7.19 Average number of high level noise events recorded at each location per night
For illustration purposes, typical sources of high level noise recorded in the bays and single rooms are shown in Table 7.4, together with their noise level (LAmax). It should be noted that the exact position of the noise source relative to the microphone is unknown. Where human activity is measured it can be reasonably assumed that this has occurred at the closest bed to the microphone (approximately 2 m in distance). Table 7.4 Examples of noise sources and levels on the medical ward
Noise event Trolleys (various) Checking patient's notes at the bedside (ring binder) Patient snoring Patient's mobile phone ringing Cough Loud crash (measured in bay 1) Medical equipment alarm Noisy motorbikes (measured in single room with window open) Sirens (unknown if window open or closed) Rubbish bin Changing bin bag Crash from kitchen (measured in bay 1) Dropped object in corridor (measured in bay 1) Nurse call (measured in bay 3) Sneeze
108
LAmax (dB) 77.7; 84.8 83.6 70.4 75.4 79.4 80.9 72.5 72.5 74.0 76.5 92.7 94.5 103.1 72.5 89.4
It can be seen in Table 7.4 that several noise events are measured at over 90 dB LAmax, with a dropped object generating a level of 103.1 dB LAmax. Such high noise levels would undoubtedly cause annoyance and disturbance to patients and staff nearby.
7.7. Surgical ward

This ward is located on the 3 floor of the main hospital building, directly under the medical ward, and is for elective orthopaedic procedures. This includes surgery to joints, hands, arms, shoulders and breast surgery. Patients range in age from 18+ and length of stay ranges from less than 24 hours to up to around 10 days. The ward has 26 beds in total, with four 4-bed bays, a 6-bed bay and 4 single rooms. Bays are predominantly female, with one all male 4-bed bay and single rooms used for male patients when necessary. With the equipment already sanctioned by the infection control team, only microphone positions and questionnaire distribution needed to be discussed with the ward manager, who was extremely supportive of the study. For comparison purposes it was suggested that similar microphone positions should be used to those in the medical ward. The ward manager was happy for this to be the case, and a ward plan indicating these positions can be seen in Figure 7.20 on page 111. As in the medical ward, the microphone was suspended from the ceiling where possible. Questionnaires were also reviewed by the ward manager and it was decided that they would be distributed by the ward clerk to those patients who had been on the ward for over 24 hours and were felt to be fit enough to complete the survey. Staff questionnaires were to be left for staff to complete in the staff room. As in the medical ward a number of laminated advertising posters were displayed throughout the ward common areas.
rd
7.7.1.
Ward specific information
Staffing levels The nursing staff levels are highest during the morning with five dedicated nursing staff. This drops to four during the afternoon and three at night. Other dedicated ward staff include a housekeeper; three domestics, ward clerk; dedicated occupational therapist; and a physiotherapist.
109
Ward routine Observation rounds begin at 06.00, with drug rounds and first admissions beginning an hour later. Doctors and anaesthetists arrive for morning admissions around 08.30 and may be on the ward for several hours. Porters begin to take patients to the operating theatre at this time. In the late morning, staff begin to take their breaks and patient discharges are discussed, causing more activity around the nurse station. Physiotherapists & occupational therapists carry out their duties on the ward; drug rounds continue; porters are busy taking patients for X-rays and bringing back patients from surgery. From 13.30 until 14.00 there is a shift handover which takes place both in the office and at the bedside. Sometimes this can lead to five or six people gathered around a patients bed, which can generate some noise. The second round of admissions arrive at 14.00; observations of patients back from surgery continue; bloods and x-rays are taken in preparation for surgery; doctors arrive on the ward to visit new admissions. At 16.00 staff begin to take breaks and at 17.30 day surgery patients begin to go home. Drug rounds are ongoing. Sources of noise specific to the surgical ward The following are potential sources of noise on the ward: The defibrillator self tests at 03.00 A series of bleeps from the fire alarms can be heard down the corridor possibly a self test? The fire exit door at the end of the ward is heavy and is ill-fitting in its frame. Although it is fitted with a quiet closer it vibrates loudly on closing. The cupboard, which is opposite 4-bed bay 4, has a noisy metal roller shutter.
110
Figure 7.20 Detailed plan of the surgical ward showing microphone positions
7.8. Surgical ward overall noise survey results

Noise level measurements were made at six different locations on the surgical ward. Table 7.5 shows the locations, measurement intervals, and the patient genders where applicable. Table 7.5 Measurement location, time periods and patient gender type Position Nurse station 4-bed bay 1 6-bed bay 3 4-bed bay 4 Single room A Single room B Length of measurement period 7 days 6 days 8 days 7 days 7 days 7 days Patient Gender N/A Male Female Female Male / Female Male / Female
Overall measurements of A-weighted equivalent sound pressure levels (LAeq) for 24 hours, night and day time were recorded at each location and are shown in Table 7.6. Table 7.6 Average LAeq for 24 hour, day and night time periods at each location
Position in ward Average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Nurse station 4-bed bay 1 6-bed bay 3 4-bed bay 4 Single room A Single room B
54.5 51.7 51.0 50.9 55.1 56.5
55.8 53.1 53.6 52.6 56.8 58.2
48.6 41.4 41.8 42.0 44.6 46.4
A summary of the day and night time average levels presented in Table 7.6 are presented graphically in Figure 7.21 for clarity. As with the medical ward, all levels in patient accommodation exceed those suggested in the WHO guidelines without exception. It can be seen that day time and night time levels measured in bays 1, 3 and 4 are very similar with an average night time drop of 11 dB. As found in the medical ward, single patient rooms are less consistent both during the day and night.
112
65
Day time LAeq, 16hr

60 55 50 45
Night time LAeq, 8hr
40 35 30 25 20 15 10 5 0
Nurse station
4-bed bay 1
6-bed bay 3
4-bed bay 4
Single room A
Single room B
Figure 7.21 Average day and night LAeq levels measured at each location Detailed results of levels measured at the nurse station are discussed in the next section, with further results from the multi-bed bays shown in Section 7.8.2 and from the single rooms in Section 7.8.3.
7.8.1. Nurse station

Figure 7.22 shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours for the nurse station.
70 Night time
60
50
40
Day time 30
20
Time (24h:00) Nurse Station LAeq Nurse Station LA90
Figure 7.22 Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station
113
As with the nurse station in the medical ward, noise levels increase steadily from around 04.30 and do not decrease substantially until late, around 23.30. Night time background levels are very consistent at around 30 dB LA90, while day time background levels are around 40 dB LA90, with a temporary peak at around 14.00 during the afternoon shift handover and the second round of patient admissions. Levels begin to decrease from around 19.30. Sources of high level noise seem to differ slightly from those captured at the nurse station in the medical ward. Here, analysis of the trigger files indicates that the nurse call was used much more frequently. This was confirmed by staff and patient responses to the questionnaire surveys, discussed in Sections 7.9.1 and 7.9.2 respectively. High level conversation was also the source of many trigger files, but unlike the medical ward high level noise due to furniture scraping on the floor was minimal. Administrative tasks involving the use of ring binders also created a number of trigger files especially during the night. This is further illustrated in Figure 7.23, which shows a number of peaks caused by the closing of ring binders over a 30 minute period.
RING BINDER
Figure 7.23 LAmax,2s (green trace) and LAeq,2s (red trace) measured over a thirty minute interval during the night, from 2.40am, at the nurse station Figure 7.24 shows an example of the nurse call captured with the microphone approximately 2 m away from the main desk area. In this particular instance the nurse call was activated for over eight minutes before it was reset. Each intermittent tone measured 70.4 dB LAmax, with an overall LAeq of 59.5 dB.
114
INTERMITENT NURSE CALL
LAmax,2s LAeq,2s
Figure 7.24
LAmax,2s and LAeq,2s fluctuating over a 11 minute interval at the nurse station
Further examples of high level noise levels captured at the nurse station, are presented in Table 7.7 below. It should be noted that the levels shown are the levels of individual events and so may not be representative of every noise of that type. Table 7.7 Examples of noise events at the nurse station
Noise event Nurse call Internal phone Ring binder Door banging further down ward corridor
LAmax (dB) 70.4; 72.7 76.8 82.6 70.6
It can be seen in Table 7.7 that the closing of a ring binder generates noise at levels as high as 82.6 dB LAmax. As much of the administrative work is carried out by staff during the night when the ambient noise level on the ward is low, this activity is likely to cause disturbance to those patients in the bay opposite the nurse station.

Figure 7.25 shows the averaged LAeq,1hr levels over 24 hours for one 6-bed and two 4-bed bays and the average background level (LA90,1hr) of all bays. It can be clearly seen that levels for the measured bays are very similar and follow the same general pattern of fluctuation, with night time levels falling to around 37 dB LAeq, and day time level remaining steady at around 53 dB LAeq. The average background level varies from around 32 dB LA90 at night to around 40 dB LA90 during the day, similar to three of the multi-bed bays in the medical ward.
115
Figure 7.25 also shows that, as in the medical ward, the WHO day / night division is not a particularly good fit. Noise levels increase steadily from around 05.30 rather than 07.00, and begin to decrease after the evening meal is served and then further decrease at 23.00. This suggests it might be appropriate to redefine the day and night time periods for hospital noise assessment and perhaps consider the addition of an evening period. The similarity of the measured levels between the 4-bed and 6-bed bays suggest, as with the medical ward, that an increase in ward size from four to six patients does not necessarily affect the noise levels. The consistency of the levels indicates that the daily ward routines which contribute to the noise levels are comparable.
50
40
30
WHO GUIDELINES
20
Time (24h:00) 4-Bed Bay 1 6-Bed Bay 3 4-Bed Bay 4 Average LA90
Figure 7.25 Average LAeq,1hr for each multi-bed bay and combined average LA90,1hr level for all bays over 24 hours

Noise levels were measured in two single rooms. Figure 7.26 shows the average measured LAeq,1hr levels over 24 hours and the average background level (LA90,1hr) of both rooms. To allow for comparison with levels measured in the multi-bed bays, the average LAeq,1hr background level (LA90,1hr) of all the multi-bed bays are also shown on the graph. and the average
116
70 Night time Day time
60
50
40
30
WHO GUIDELINES
20
Time (24h:00)
Single room A LAeq Single room B LAeq Multi-bed bay average Average LA90 - Single Rooms Average LA90 - Multi-bed bays
Figure 7.26 Average LAeq and LA90 levels for single rooms A and B and multi-bed bays It can be seen that, as in the medical ward, noise levels measured in both single rooms are consistent higher than those measured in the multi-bed bays. Background levels (LA90,1hr) are also higher, with levels around 2 dB higher at night and as much as 4 dB higher during the day. The patient in single room B had been on the ward for some weeks and seemed to enjoy chatting to staff and her numerous visitors. As well as noise generated as a result of her clinical care, much of the high level noise in this room was caused by conversation. This was particularly noticeable between 15.00 and 20.00, where it can be seen in Figure 7.26 that noise levels consistently exceed those measured in single room A. This is due to high levels of conversation during visiting time. In single room A conversation was again responsible for a percentage of high level noise, but medical equipment alarms also had an impact in this room. One particular type of alarm often generated high level noise with its intermittent 30 second bleep. An occurrence of this was found to continue for over two hours and a half hours before it was reset, with each bleep measuring close to 75 dB LAmax . The effect of this is further illustrated in Figure 7.27.
117
INTERMITTENT BLEEP OF MEDICAL EQUIPMENT ALARM
Figure 7.27 LAmax,2s (green trace) and LAeq,2s (red trace) showing the noise levels due to a medical equipment alarm over a 13 minute period
Single room A is situated directly behind the nurse station. As discussed previously, doors to the single rooms are generally left open to allow for easy observation of the patient. Staff talking, the nurse call and internal phone ringing can all be heard clearly in the background from this bay, with the nurse call measured at levels around 67 dB LAmax, and the internal phone measured at levels around 53 dB LAmax..

To help build up a further picture of the sources of high level noise at each location, the numbers of occurrences of LAmax in 5 dB bands from 70 to 95 dB have been examined. Figures 7.28 and 7.29 show the average number of high level noise events during the day and night in different measurement locations. It can be seen that the numbers of high level noise events in single room A during the day exceed those in all other locations, with, on average, almost 500 events with LAmax between 70 and 75 dB; around 150 events with LAmax between 75 and 80; and 30 events with LAmax above 80 dB. As discussed in the previous section, these high level events are partly due to conversation with staff and visitors, but are also impacted to a large extent by a particular piece of medical equipment with a high pitched alarm. Alarms generally sound when equipment requires some kind of attention, for example fluid levels are becoming low. Staff are obviously aware of the length of time they have before they need to respond to an alarm; however, in this case one occurrence of this high pitched intermittent bleep was observed to continue for over two and a half hours. This must have been annoying to the patient, who was trying to rest after an undergoing an operation on the previous day. Perhaps if this alarm had been in a multi-bed bay it may have been attended to sooner, especially if other patients were annoyed by the noise and alerted staff.
118
The differences in numbers of high level noise events between locations emphasise the limitations of using only LAeq levels to describe the noise climate of hospital wards. In terms of LAeq,1hr room B has higher levels (see Table 7.6 and Figure 7.26) yet in terms of individual high noise events Figure 7.28 shows that room A is noisier.
500
70 LAmax < 75 dB
450
75 LAmax < 80 dB 80 LAmax < 85dB 85 LAmax < 90dB 90 LAmax < 95 dB
400
350
Number of high level noise events
300
250
200
150
100
50
0 4-Bed Bay 1 6-Bed Bay 3 6-Bed Bay 4 Single room A Single room B Nurse Station
Figure 7.28 Average number of high level noise events captured at each location per day Many high level noise events can also be seen in single room B, with over 400 events with LAmax between 70 and 75 dB; over 200 events with LAmax between 75 and 80; and around 80 events with LAmax above 80 dB. When first admitted onto the ward, the patient in this room had been put in a multibed bay. Here she had enjoyed talking to fellow patients and was disappointed when she was moved into a single room, feeling more cut-off. This lady took every opportunity to chat with staff and visitors for company which caused the majority of the high level noise events. Although, as discussed in Section 7.8.2, overall noise levels in bays 1, 3 and 4 are shown to be similar, differences can be seen between the bays in the numbers of high level noise events. As in the medical ward, bay 1 is opposite the ward clerks desk, kitchen and staff room. Noise events from these areas have been shown to impact the noise environment of this bay; this is further confirmed by patient responses to questionnaire surveys, discussed in Section 7.11. Unlike the medical ward the numbers of high level noise events captured in bay 3, which is opposite the nurse station, were the lowest. This further confirms the fact that noise from the nurse station is having a minimal impact on the noise levels in this bay.
119
More sources of high level noise were captured at the nurse station on the surgical ward. As discussed in Section 7.8.1 many of these were due to conversation. The nurse call also appeared to be used more in this ward. The surgical ward is very different to the medical ward in terms of timings, logistics and planning of operations. It is unsurprising to find more discussion at this nurse station as the general pace of this ward is more frenetic than that of the medical ward.
110
100
70 LAmax < 75 dB 75 LAmax < 80 dB 80 LAmax < 85dB
90
85 LAmax < 90dB
80
Number of high level noise events
70
60
50
40
30
20
10
0 4-Bed Bay 1 6-Bed Bay 3 6-Bed Bay 4 Single room A Single room B Nurse Station
Figure 7.29 Average number of high level noise events captured at each location per night
Figure 7.29 shows that the numbers of high level noise events during the night are generally much lower than those observed in the medical ward. This is mainly due to the differences in the numbers of instances of patients crying out, which was very noticeable in the medical ward. This is further confirmed by patient responses to questionnaire surveys, discussed in Section 7.9.2.

Staff response was good in the medical ward with 18 questionnaires completed, but response in the surgical ward was rather poor, with only seven staff completing the survey. The following section discusses results from the staff questionnaires and examines the differences between perceptions on the medical and surgical wards. Information regarding the design of the questionnaire surveys can be found in Section 5.4.

To establish certain attributes about the staff, the first section posed a number of basic questions. Out of the 18 respondents on the medical ward only two were male, and all seven respondents in the surgical ward were female. 120
The ages of the respondents in the two wards are shown in Figure 7.30. It can be seen that in both wards the respondents are generally younger than 50, with a higher percentage of young staff members completing the questionnaire in the medical ward.
50 45 40 35 30 25 20 15 10 5 0 20 - 30 31 - 40 41 - 50 51 - 60
Percentage of respondents (%)
Medical Surgical
60+
Age band (years)
Figure 7.30 Age of respondents by band Questions were asked in relation to the length of time worked both on the ward and at the hospital, and the responses can be seen in Figures 7.31 and 7.32 respectively. What is clear is that staff turnover in the surgical ward appears to be relatively low, with nearly 60% of respondents having worked on the ward for over five years. The medical ward was more mixed, with an influx of new staff (~40%) in the last year. It also appears some staff had transferred from other wards within the hospital during their career.
80
Medical 50 40 30 20 10 0 Less than 1 year 1-2 years 2- 3 years 3- 4 years 4- 5 years 5+ years Surgical
60
70 60 50 40 30 20 10 0
Medical Surgical
Less than 1 year
1- 2 years
2- 3 years
3- 4 years
4- 5 years
5+ years
Time worked on the ward
Time worked on the hospital
Figure 7.31 Time worked on the ward
Figure 7.32 Time worked at the hospital
7.9.2.
Noise annoyance
General feelings of noise annoyance were investigated by asking staff to what extent they were annoyed by noise. Figure 7.33 shows that the highest percentage of staff in the medical ward were moderately annoyed by noise (43%), but this was not the case in the surgical ward, where the majority (56%) of those questioned felt only slightly annoyed by noise.
121
Extremely Very much Moderately Surgical Slightly Not at all 0 10 20 30 Percentage (%) 40 50 60 Medical
Figure 7.33 Staff perception of noise in terms of annoyance Staff were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Figure 7.34 shows the percentages of staff who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little annoyed by the event. It can be seen that the most annoying noise events for the staff on the medical ward were visiting time, medical equipment alarms and the internal telephone. This is similar for the surgical ward, except that there the nurse call is also rated by a high percentage of respondents. Discrepancies of between 20% and 30% can be seen between the medical and surgical ratings for cleaning, people talking and staff talking. These events were found to be annoying by staff on the medical ward, but to a much lesser extent in the surgical ward.
122
Visiting time Medical Equipment Internal telephone Cleaning People talking Rubbish bins Nurse call Talking on mobile phones Meal times Trolleys Mobile phones ringing Staff talking on the telephone Footsteps Doors banging TV / radio External noise 0 10 20 30 40 50 60 70 80 90 100 Surgical (n=7) Medical (n=18)
Figure 7.34 The percentage of staff rating an annoyance noise event with a 2, 3 or 4
Doors banging and external noise are rated more highly in the surgical ward. There is a particular heavy, ill-fitting fire door at the end of this ward that was mentioned during initial discussions with the ward manager. When this door bangs shut the noise appears to travel down the full length of the ward corridor. With regards to the external noise; the surgical ward was surveyed during the summer months when the weather was warmer, whereas the medical ward was surveyed in the spring. This may account for the difference in external noise annoyance, as more windows may have been open in the warmer weather. For 10 out of 16 noise sources, the percentages of those annoyed on the surgical ward are higher than on the medical ward. This could of course be simply down to the smaller sample size, and the possibility that only those staff who felt strongly about noise felt inclined to complete the questionnaire. However, other factors could also account for this difference. Medical and surgical wards are different, and as such may attract staff with certain personalities. Surgical wards are very busy with constant admissions for day or even half day procedures. Operations are booked in advance and efficiency and timing are key. Medical wards are slower paced and it is possible that staff annoyance of particular events could be less extreme.
123
7.9.3.
Interference with work
Respondents were asked to what extent noise interfered with their ability to work effectively. As can be seen in Figure 7.35, opinions of the respondents in the surgical ward were very split, whereas the majority of medical ward staff chose slightly or not at all.
Extremely Very much Moderately Surgical Slightly Not at all 0 10 20 30 40 50 Medical
Percentage (%)
Figure 7.35 Staff perception of the extent to which noise interferes with work
Staff were also asked to rate how much each noise event interfered with their ability to carry out their job effectively (again the rating scale of 0 to 4 was used). Figure 7.36 shows the percentages of staff who rated a noise event with a 2, 3 or 4, and as such it could be said that this noise event interfered to some extent with their ability to carry out their job effectively.
Visiting time Medical Equipment Internal telephone Staff talking on the telephone Nurse call Talking on mobile phones Trolleys Rubbish bins People talking Mobile phones ringing TV / radio Cleaning Footsteps Doors banging Meal times External noise 0 10 20 30 40 50 60 70 80 90 100 Surgical (n=7) Medical (n=18)
Figure 7.36 The percentages of staff rating an interference noise event with a 2, 3 or 4 124
As with the noise annoyance ratings, visiting time, medical equipment alarms and the internal telephone were all rated as interfering with work by over 40% of respondents in each ward. There are several anomalies worth noting. The nurse call is once again rated by a high percentage of surgical staff as well as the trolleys and meal times; however these are not rated by a high percentage of medical staff. Trolleys are used in both wards a great deal, but in the surgical ward patients are often being wheeled through the ward to and from surgery and X-ray so trolley noise may be more disruptive. It is unclear why meal times would be more disruptive in the surgical ward.
7.9.4.
Important sounds
To aid understanding of which sounds were felt by staff to be important to be heard in order to carry out their jobs effectively, staff were asked to rate different noise events on a scale of 0 to 4, where 0 indicated not at all important and 4 indicated extremely important. Figure 7.37 shows the mean ratings for each noise event. It can be seen that conversations with colleagues closely followed by conversations with patients were considered by staff in both wards to be the most important noise events. However, the average ratings were consistently high in all cases suggesting that all of these events are important for staff. As with the annoyance and interference ratings in the previous sections, staff in the surgical ward rated most events as more important than those in the medical ward, but a similar pattern can be seen.
4
2 Surgical (n=7) Medical (n=18)
0 Nurse call Conversations Conversations with colleagues with patients Medical equipment alarms Patients calling Patient activity out
Figure 7.37
Mean importance rating of certain noise events
125
7.10. Results of the patient questionnaire surveys

With the help of the ward clerks, questionnaires were distributed to those patients who had been on the ward for over 24 hours and were judged to be physically and mentally fit enough to complete the survey. In total 40 patients completed the questionnaire in the medical ward and 42 in the surgical ward. The following sections discuss results from the patient questionnaires and examine the differences between perceptions on the medical and surgical wards.
7.10.1. Patient profiles

As with the staff questionnaire, the first section aimed to establish certain attributes about the patients, beginning with gender. As discussed previously, the surgical ward was predominantly female, and the medical ward predominantly male. This is shown clearly in Figure 7.38 below.
90 80 70
Percentage (%)
60 50 40 30 20 10 0 Male Female Medical ward (n=40) Surgical ward (n=42)
Figure 7.38 Gender split by ward type Figure 7.39 shows the respondents age ranges. It can be seen that a relatively high percentage of patients were aged 60 or above, with 70% in the surgical ward and 40% in the medical ward in this age range.
80 70 60
Percentage (%)
50 40 30 20 10 0 20-30 31-40 41-50 Age range 51-60 60+ Medical ward (n=40) Surgical ward (n=42)
Figure 7.39 Patients age by band 126
Patients were asked how long they had been on the ward. Some differences can be seen between the surgical ward and medical ward in Figure 7.40 below. Nearly 80% of respondents in surgical were short term patients, having been on the ward for less than one week. It can be seen that the variation in the medical ward was more marked.
90 80 70
Percentage (%)
60 50 40 30 20 10 0 < 1 week 1- 2 weeks 2 - 3 weeks 3+ weeks Length of stay Medical ward (n=40) Surgical ward (n=42)
Figure 7.40 Length of patient stay when completing the questionnaire Hearing impairment was also explored, with 24% of respondents on the medical ward and 17% on the surgical ward indicating that they did suffer to some degree The bed number of the respondent was noted on the front of the questionnaire by the ward clerk. This number provided useful location information which is considered when investigating relationships between bed positioning and patient accommodation type with day time noise annoyance and night time disturbance, which are explored in Chapter 11. In terms of the single room / multi bed bay split, 91% of respondents in the medical ward were in multi-bed bays, with 83% in the surgical ward.
7.10.2. Noise annoyance and disturbance

The next section of the questionnaire considered day time noise annoyance and night time disturbance. The questionnaire sought to identify the sources of noise that may annoy or disturb patients. Respondents were given two lists of noises and were asked to rate the day time annoyance and night time disturbance on a scale of 0 to 4 (where 0 indicated no annoyance / disturbance and 4 indicated a great deal). Several lines were left blank at the bottom of the lists for patients to add and rate additional noise sources. Patients were first asked how they perceived the day time noise environment on the ward. Figure 7.41 details the responses, which are fairly split between quiet and a little noisy. Interestingly, when asked whether they were actually annoyed by noise, only 13% of patients in the medical ward felt annoyed, and 29% of patients on the surgical ward. 127
Extremely noisy Very noisy A little noisy

Surgical (n=42)
Quiet Very quiet 0 10 20 30 40 50 60
Medical (n=40)
Percentage (%)
Figure 7.41 Patient perception of the day time ward noise environment The patients who had indicated that they were annoyed by noise during the day, were then asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. With a relatively small number of people annoyed by day time noise the sample set was low (n=5 for the medical ward and n=11 for the surgical ward). Figure 7.42 shows the percentage of patients within these samples who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little annoyed by the event. It can be seen that patients crying out, trolleys, internal telephones and rubbish bins (to a certain degree) appear to be sources of annoyance in both wards. One particular difference is the doors banging, rated by nearly 60% of patients on the surgical ward, but no one on the medical ward. As discussed in the staff questionnaire section, there is one particularly heavy fire door at the end of the ward corridor which was mentioned as a problem in initial discussions with staff. Other noticeable differences are the annoyance caused by visiting time, footsteps, nurse call and external noise. All these events are only cited by patients in the surgical ward. Due to the nature of the surgical ward, with patients being taken up and down to X-ray and surgery, this may account in part to the increased annoyance caused by footsteps. Occurrences of the nurse call were captured more often and continuing for longer periods at the nurse station in the surgical ward (see Section 7.8.1), which explains the patient responses in this case. As discussed previously, external noise may be more of a problem during the study period in the surgical ward as the weather was warmer and more windows would have been open. Talking on mobile phones and TV / radio are the only events that are cited by medical patients only. This could be due to a lack of enforcement of mobile phone policy, and the non-compulsory use of headphones.
128
Patients crying out Trolleys Internal telephone Talking on mobile phones Mobile phones ringing TV / radio Meal times Rubbish bins Cleaning People talking Medical Equipment Staff talking on the telephone Visiting time Footsteps Nurse call Doors banging External noise 0 20 40 60 80 100
Surgical (n=11) Medical (n=5)
% of patients who rated each event 2 or above in terms of noise annoyance
Figure 7.42 The percentage of patients rating an annoyance noise event with a 2, 3 or 4 Two patients in the medical ward added an additional noise event that they themselves found to be annoying. The events were a patient attention seeking and the fan in the shower room (this was an ensuite shower room in a single room). Sirens and the photocopier in corridor being used in the evening were cited in addition by patients in the surgical ward. Patients were asked how they perceived the night time noise environment on the ward. Figure 7.43 details the responses, where again the majority of the responses were split between quiet and a little noisy, but with a noticeably higher percentage (18%) in the medical ward choosing the very noisy category than during the day. When asked whether they were actually disturbed by noise at night, 58% of patients in the medical ward felt they were, compared with 51% of patients on the surgical ward. This suggests that, in this hospital, over 50% of patients are disturbed by noise at night.
129
Extremely noisy
Very noisy
A little noisy
Quiet
Very quiet 0 10 20 30 40 50
Percentage (%)
Figure 7.43 Patient perception of the night time ward noise environment Patients who had indicated that they were disturbed by noise during the night were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Sample sets were higher than for the day time annoyance (n=23 for the medical ward and n=19 for the surgical ward), indicating a much higher level of night time disturbance. Figure 7.44 shows the percentages of patients within this sample who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little disturbed by the event. One noticeable difference that can clearly be seen is that patients crying out seems to be much more of a problem on the medical ward during the night. This is possibly related to the number of elderly patients suffering from confusion and dementia on this ward, who tend to cry out more often. It can be seen that certain events which were rated as annoying by only patients in the surgical ward during the day, cause a level of night time disturbance in both wards. Doors banging, medical equipment, trolleys and people talking are rated by similar percentages of patients on both wards. However, the nurse call, the internal telephone and external noise are all rated as more disturbing on the surgical ward. Occurrences of the nurse call were captured more often and continuing for longer periods at the nurse station in the surgical ward, which explains the patient responses in this case. As discussed previously, external noise may have been more of a problem during the study period in the surgical ward as the weather was warmer and more windows would have been open. As with daytime annoyance, Talking on mobile phones is cited as a disturbance only on the medical ward. The hospital policy specifies that mobile phones should only be used in the lobby areas and not on the wards. This suggests a lack of policy enforcement by the staff on the medical ward.
130
Patients crying out Doors banging Medical Equipment People talking Talking on mobile phones Trolleys Nurse call Staff talking on the telephone Internal telephone Mobile phones ringing TV / radio Rubbish bins Footsteps External noise 0 10 20 30 40 50 60 70 80 90 100
% of patients rating night disturbance event of 2 or above
Figure 7.44
The percentage of patients rating a disturbance noise event with a 2, 3 or 4
Six patients in the medical ward added an additional noise event that they themselves found to be disturbing at night. The events were: Moaning, groaning and talking in sleep Dripping taps Noisy bed neighbours A patient admitted at night Private conversations between night staff, especially in native language Night staff having private conversations on mobile phones One patient in the surgical ward also added the supply cupboard door as an additional noise event. This door had been mentioned by staff as a source of noise as it was a heavy metal roller shutter.
7.10.3.
Positive sounds
Looking at sound in a positive rather than in a negative light, patients were asked if there were any sounds that they actually found comforting. 70% of patients in the medical ward and 76% on the surgical ward left the answer blank; however, there were twelve completed responses, which included listening to music on the radio, knowing that the nursing staff were nearby to provide care, the tea trolley, and maintaining some connection with the outside world. The full responses can be seen in Appendix B.
131
Respondents were also asked if they felt that there was ever too little sound in a room. Only 8% of patients in the medical ward and 8% in the surgical ward said that they did. Surprisingly only one of these respondents was in a single room.
7.10.4.
Ease of hearing and privacy
Patients were asked whether high levels of background noise may at times make it difficult to hear doctors and nurses who talk to them. 27% of respondents on the medical ward and 36% on the surgical answered that this was the case. Conversational privacy was investigated by asking whether the patients felt that they could have a private conversation at their bedside. 100% of patients in single rooms said that they felt they could speak privately, with lower percentages in the multi bed bays of 67% and 64% in the medical and surgical wards respectively. Out of those who felt they could speak privately, around 40% said they felt they could talk in their normal voice, with 60% needing to lower their voice or taking some other precautionary measure similar percentages were found in both wards.
7.11. Questionnaire comments

Staff and patients were invited to make additional comments at the end of the questionnaire if they wished. Very few staff made comments, but many patients did leave some feedback which was very varied. Several patients cited the kitchen and ward entrance as a source of noise, and one patient suggested that the ward clerks desk, which acts as a ward reception, should be moved outside the ward entrance. Mobile phones ringing in the night, large groups of visitors around a bed and loud and aggressive patients were also mentioned. A detailed list of these comments is shown in the Appendix B.
7.12. Summary
This section summarises the main findings from the study of the medical and surgical ward at Bedford Hospital: Average noise levels measured at the nurse stations on both wards were similar both in level and fluctuation patterns, with day time levels of around 55 dB LAeq. However, sources of high level noise were found to differ, with the nurse call and high levels of conversation more prevalent at the nurse station in the surgical ward, and furniture scraping and doors banging found more frequently at the nurse station in the medical ward. Noise level measurements made in the multi-bed bays were very consistent in both level and fluctuation patterns in both wards, except for 6-bed bay 3 in the medical ward where the levels were higher. Patients crying out and increased clinical activity were found to account for these increased levels.
132
Levels did not appear to be affected by increased bed numbers, that is, from four beds to six beds. Single rooms were found to have less consistent patterns of noise levels and often had higher levels than those measured in multi-bed bays. Behaviour of patients, visitors and clinicians was shown to be the main cause. All measured levels were above those suggested by the WHO guidelines and the day / night division specified by the WHO did not appear to be realistic. Although average noise levels were similar, subsequent investigation of the numbers of high level noise events recorded in each bay indicated differences in the noise climate. For example, in both wards, bay 1 is shown to be affected by noise from the ward kitchen, staff room and particularly the ward clerks desk area. Questionnaire responses were found to reinforce this. Staff in both wards rated visiting time, medical equipment alarms and the internal telephone as the most annoying noise events. Cleaning, people talking and staff talking were found to be more annoying by staff on the medical ward, whereas the use of the nurse call, banging doors, trolleys and external noise were found to be more annoying by staff on the surgical ward. Only 13% of patients on the medical ward, and 29% of patients on the surgical ward were annoyed by noise during the day. Patients crying out, trolleys and internal telephones were the main sources of day time annoyance on both wards, with doors banging, visiting time, footsteps, the nurse call and external noise only cited by patients on the surgical ward. Higher percentages of patients were disturbed by noise at night, with 58% in the medical ward and 51% in the surgical ward. The main differences found between the night time ward environments were patients crying out and mobile phone use in the medical ward, and the use of the nurse call, internal telephone and external noise in the surgical ward.
7.13. Conclusions
Noise level measurements and questionnaire surveys have confirmed that noise is a problem in both medical and surgical wards. Staff responses indicate that they are annoyed by noise, and over half the patients questioned felt that they were disturbed by noise during the night, a time when they should be able to rest and recuperate. The identification of high level noise sources has shown that in this hospital building the ward design does appear to have a negative effect on patients located in some of the bays and rooms, specifically bay 1 and the single rooms behind the main nurse station. By re-siting the ward clerks desk and ensuring doors to the kitchen and staff room are kept closed, much of this unwanted sound could be prevented. However, the two single rooms located directly behind the main nurse station is a more difficult problem to address, with doors to the rooms left open to allow for patient observation.
133
Noise levels did not appear to be related to occupancy levels, with similar levels measured in four and six bed bays, and higher levels measured in single patient rooms than in multi-bed bays on occasions. Much of the high level noise identified could be reduced with changes to behaviour, correct enforcement of hospital policies, simple improvements to design and maintenance of equipment. This is discussed further in Chapter 12. The following chapter investigates the effects of a refurbishment carried out in the medical ward at Bedford Hospital. Reflective ceiling tiles in one four bed bay were changed for tiles with good acoustic properties. Subsequent changes to noise levels and reverberation times are investigated and reported.
134
Acoustic Design for Inpatient Facilities in Hospitals Ceiling intervention study ___________________________________________________________________________________________________________________
8. Ceiling intervention study, Bedford Hospital

8.1. Introduction
The medical ward in Bedford Hospital was due for refurbishment soon after the objective measurements and subjective surveys (discussed in Chapter 7) were completed. This was to include the replacement of the suspended ceiling tiles; a full repaint of all rooms; replacement of protective wall panelling; and some general tidying and maintenance work. Much of the existing suspended ceiling in the ward was known to have good acoustic properties, but the large perforations in the tiles were felt to be unsuitable in light of the current control of infection policy. Tiles of similar acoustic properties are now available on the market designed specifically for hospital use. These tiles have a smooth finish and can withstand the bleaches and detergents commonly used in hospital cleaning regimes. The author, working in conjunction with the hospital estates team, suggested that these replacement ceiling tiles should be considered throughout the ward. Although the tiles were slightly more expensive than the standard plain plaster tiles, the purchase was agreed. One bay was of particular interest to both the Estates team and the author. This bay had been refurbished more recently than the rest of the ward and had a suspended ceiling of plain plaster type ceiling tiles with poor acoustic properties. It was considered by both parties to be an excellent opportunity to study the effects of changing the ceiling tiles, on the acoustic environment. The results could then be used by the estates team to inform their decision making for further ward refurbishments. Figure 8.1 shows the bay during refurbishment.
Figure 8.1 Photographs of the bay during refurbishment
135
The study of the effects of the ceiling change in this bay discussed in this chapter, consists of two parts: firstly the investigation of the effects of the ceiling change on the noise levels; and secondly the effects of the ceiling change on occupied and unoccupied reverberation times.
8.2. Bay information

Bay 1 is the first open four bed bay immediately after the main entrance to the medical ward, and is opposite the ward clerks desk, the staff room and the kitchen, as shown in Figure 7.10 in Section 7.5. This particular bay was refurbished several years ago with a change from the original acoustic ceiling tile to Armstrong Bioguard Plain tiles. Although suitable for hospital use, these tiles have poor acoustic properties. Figure 8.2 shows the manufactuers data on absorption coefficients for octave frequency bands from 125 Hz to 4 kHz.
Figure 8.2 Absorption coefficients of Armstrong Bioguard Plain ceiling tiles Source : Manufacturers product specification sheet During the ward refurbishment, these tiles were changed for Armstrong Bioguard Acoustic tiles. Figure 8.3 shows the manufactuers data on absorption coefficients in frequency bands from 125 Hz to 4 kHz for this tile. It can be seen that these tiles provide much greater acoustic absorption than the previous ones, with particular improvement in frequencies from 500 Hz to 4000 Hz.
Figure 8.3 Absorption coefficients of Armstrong Bioguard Acoustic ceiling tiles Source : Manufacturers product specification sheet
136
8.3. Effect of ceiling tile change on noise levels

As part of the main study, occupied noise levels were measured in bay 1 over two separate weeks prior to the ceiling change, as discussed in Section 7.6.2. For comparison purposes, the microphone was suspended in the same position following the ceiling change and a further two weeks of data were collected and compiled. The first set of noise level data was collected with the bay occupied by female patients in April and June 2010. The post ceiling change data were collected during December 2010 with male patients in the bay. Apart from the time of year and the patient gender, there were no other known changes on the ward, apart from the ceiling tiles. All ward routines remained the same. Before any meaningful comparisons could be made between the pre and post ceiling change measurements, the data were analysed in detail to see if any anomalies were present. Each trigger file (created when LAmax exceeded 70 dB) was reviewed and all files were grouped by event type. The average number of triggers over a 24 hour period was calculated for each event type both pre and post the installation of the new ceiling tiles. Two anomalies caused by untypical patient behaviour were identified and were considered to be worth further investigation: unusually high numbers of triggers caused by patient cries during the pre ceiling change period; and the amount of coughing after the ceiling change. Further analysis of the trigger files showed that during the pre ceiling change measurement period there was a very confused elderly lady in the bay who screamed whenever staff tried to sit her up or get her out of bed. This caused a total of 278 individual triggers files (201 during the day and 77 at night) over the period of one week which caused a increase in overall noise levels. It was felt that this was an untypical event as there were no other episodes of this type during the subsequent weeks of data collection. Therefore, to ensure that the data was comparable, this particular event was removed for the analysis of noise levels. The noise level measurements made after the ceiling change were carried out in December. This was a particularly cold month and patients coughing caused a large number of trigger files during the measurement period (on average more than 120 triggers were caused by coughing over 24 hours compared with an average of 20 over 24 hours during the pre ceiling change period). As the ward was not one that dealt with chest infections this was felt to be untypical of the noise climate on this ward. It was decided that to provide a more realistic data comparison, the high numbers of coughs in the data should be reduced to the average number found during the pre ceiling change period. With the anomalies in the data removed, overall noise levels were calculated. Table 8.1 shows the average LAeq measured during the day and night time pre and post the ceiling change. It can be seen that day time and night time levels after the ceiling change are on average 2.4 dB and 3.4 dB LAeq respectively lower than levels before the change. 137
Table 8.1 Average LAeq measured during the day and night time pre and post the ceiling change
Position in ward
A-weighted equivalent sound pressure levels Day time LAeq, 16hr Night time LAeq, 8hr
Pre ceiling change (2 weeks) Post ceiling change (2 weeks)
53.7 51.3
45.2 41.8
70 Night time 60
50
40 Day time 30
20
Time (24h:00)
Pre ceiling change
Post ceiling change
Figure 8.4
Average LAeq,1hr levels over 24 hours pre and post ceiling change
Figure 8.4 shows the average LAeq,1hr levels measured over 24 hours pre and post the ceiling change. Fluctuations in level follow the same pattern, suggesting that ward routines are unchanged, but the levels are consistently lower. The decrease in the overall measured noise levels suggests that the acoustic ceiling tiles are having some positive effect. This can be further substantiated by looking in more detail at the high level noise sources before and after the ceiling change, with the anomalies removed. Figure 8.5 shows the number of high level noise events captured; that is events where LAmax exceeds 70 dB. Any notable differences are highlighted.
138
Mobile phone ringing Noise from corridor / kitchen Visiting time Trolley Bathroom door Meal time Tea / drinks round Cupboard door / drawers Cleaning Ring binder / admin at nurses' desk Furniture scraping Patient calling out Shoes squeaking on floor Laughter Rubbish bin Cough / sneeze / gurgling / groaning / snoring Patient procedures Patient talking on phone Patients talking Staff and patients talking Conversation between staff Unidentifiable 0 10 20 30 40 50 60 70 80 90 100 110 120 130
PRE ceiling change POST ceiling change
Average number of trigger files recorded in 24 hours
Figure 8.5 Average number of trigger files recorded over 24 hours by event type It can be seen in Figure 8.5 that there are notable reductions in the numbers of high level noise events associated with visiting time, patient procedures and staff and patients talking. These events include speech, a frequency range at which the new ceiling tiles are known to be particularly acoustically absorbent which may explain these reductions. The numbers of high level noise events at meal times and during cleaning are also reduced, as are the number of unidentifiable events. Figure 8.6 shows the frequency content of a typical high level noise event associated with meal time, that is, metal cutlery scraping on a china plate, which was recorded before the ceiling change. It can be seen that the levels are higher at the low frequencies, decreasing steadily from around 150 Hz to around 1 kHz, and then rising again in the frequency bands at which the ceiling tiles are most effective. It is therefore unsurprising that the new ceiling tiles, which provide more absorption at high frequencies, are having some positive effect on noise events of this type.
Figure 8.6 The frequency content of metal cutlery 139
8.4. Effect of ceiling tile change on reverberation time

Reverberation time (RT) is generally used as an indicator of the acoustic comfort of a space and is an important measurement in the field of room acoustics. As discussed in Section 5.3.8, any RT measurements made in unoccupied wards were carried out using the Impulse Response Method with balloon bursts as the source of noise. In occupied wards, where physical RT measurements are not possible, an estimation method is used. As explained in Chapter 1, data collected in this study have been used to validate an estimation method developed at the University of Salford (Kendrick, 2009). The validation is discussed in more detail in Chapter 10. The reverberation times for occupied wards presented in this Section 8.4.2, have been estimated using this method.
8.4.1. Unoccupied reverberation times

RT measurements were carried out in the unoccupied bay using the Impulse Response Method just prior to, and just after the ceiling tile change. In total twelve RT measurements were made, six before the ceiling change and six after. Figure 8.7 shows the layout of source and receiver positions, with the following combinations of source and receiver positions used: S1-R1; S2-R1; S3-R1; S1-R2; S2- R2; S4-R2.
S1 R2 S3 S2
S4 R1
Figure 8.7 Source (S) and receiver (R) positions used to measure reverberation time in the unoccupied bay before and after the ceiling change Although the bay was unoccupied during the RT measurements, there was some furniture piled in the centre of the bay on both occasions, as shown in the photographs in Figure 8.1. This may have 140
provided some small amount of acoustic absorbency, but was in place during both sets of measurements and as such may affect the overall RT values, but should not affect the differences. Figure 8.8 shows the spatially averaged RT20 values over third octave bands from 250 Hz to 4 kHz as stipulated in BS EN ISO 3382-2 (2008). The error bars show the 95% confidence limits of the mean values. It can be seen that at all frequencies the measured RT20 values have decreased after the ceiling tile change, by between 0.1 s and 0.4 s, with the greatest changes above 500 Hz. Given that these are the frequencies at which the tiles have the highest absorption coefficients, these results are as expected. The 95% confidence limits are generally small, particularly at the higher frequencies, suggesting little variation between the measured values at each frequency.
1.0 0.9 0.8 0.7
Mean RT20 (s)
Pre ceiling change Post ceiling change
0.6 0.5 0.4 0.3 0.2 0.1 0.0 250 315 400 500 630 800 1k 1.25 k 1.6 k 2k 2.5 k 3.15 k 4k
Frequency (Hz)
Figure 8.8 Average unoccupied RT20 measurements with 95% confidence limits (Impulse Response Method) Bork (2000) shows that in a room with an RT value of 2 s or less, the subjective difference limens are 0.1 s, and hence any change in the RT that is less than 0.1 s would not be noticeable to the listener. In this case however, the changes of between 0.1 s and 0.4 s exceed this difference limen, and as such would be perceived by the occupants of the bay.
8.4.2. Occupied reverberation times

Using trigger files captured during the measurement period before and after the ceiling change, reverberation times of the occupied bay were estimated using the Maximum Likelihood Estimation Method (MLE-RT20) which is discussed fully in Chapter 10. Over 5000 individual trigger files were processed to produce each octave band MLE-RT20 estimate from 250 Hz to 4 kHz. As would be expected the estimated values for the occupied ward are lower than the measured RTs in the unoccupied ward, by up to 0.3 s at 4 kHz both pre and post the ceiling
141
change .This is due to the absorbency of the additional furniture, bedding, curtains and the occupants themselves. Figure 8.9 shows the estimated MLE-RT20 values before and after the ceiling replacement. As with the values measured in the unoccupied bay, it can be seen that the addition of the acoustic ceiling tiles has reduced the MLE-RT20 values by up to 0.1 s in each of the octave bands estimated. The largest differences are in the 1 kHz to 4 kHz frequency bands, where the reduction in MLE-RT20 is greater than 0.1 s. This is as would be expected as these are the frequencies where the new ceiling tiles are particularly effective. The reduction in the MLE-RT20 estimate is much less at 500 Hz, and at 250 Hz the 95% confidence limits of the pre-replacement estimate are 0.13 s and therefore the results at this frequency should be ignored.
0.7
Estimated MLE-RT20 (s)
0.6 0.5 0.4 0.3 0.2 0.1 0 250 500 1000 2000
Pre ceiling replacement Post ceiling replacement
4000
Octave band (Hz)

Figure 8.9 Occupied MLE-RT20 estimates pre and post the ceiling replacement,
8.5. Comparison of unoccupied and occupied RTs

The estimated MLE-RT20 values shown in the previous section not only reinforce the effect of the ceiling tile change on the room acoustic, they also provide useful information regarding the amount of acoustic absorption in an occupied four bed bay, and the effect that this extra absorption has on the reverberation time. Tables 8.2 and 8.3 show the reverberation times measured in the unoccupied bay alongside the estimated MLE-RT20 values when the bay is occupied, both before and after the ceiling change.
142
Table 8.2 Reverberation times for both the unoccupied and occupied bay pre ceiling change
Frequency (Hz)
250 500 1k 2k 4k
Pre ceiling change - measured Pre ceiling change - estimated Decrease MLE-RT20 (s) in occupied bay (s) RT20 (s) in unoccupied bay
0.64 0.65 0.69 0.69 0.70 0.63 0.44 0.42 0.42 0.38 0.01 0.21 0.27 0.27 0.32
Table 8.3 Reverberation times estimates for both the unoccupied and occupied bay post ceiling change
Frequency (Hz)
250 500 1k 2k 4k
Post ceiling change measured RT20 (s) in unoccupied bay

0.51 0.42 0.29 0.36 0.38
Post ceiling change estimated MLE-RT20 (s) in occupied bay

0.39 0.33 0.27 0.29 0.27
Decrease (s)
0.12 0.09 0.02 0.07 0.11
It can be seen from Tables 8.2 and 8.3 that in the more reverberant room (pre ceiling change) the difference in reverberation times at 1 kHz between the unoccupied and occupied space at 500 Hz and above is greater than 0.2 s. However, with the ceiling changed, there is a negligible difference between the unoccupied and occupied reverberation times at those frequencies. As discussed in Section 8.4.2 the results at 250 Hz are unreliable and therefore this estimate should be ignored.
8.6. Conclusions
This chapter clearly illustrates the benefits of the installation of an acoustic ceiling, which results in consistently lower measured average noise levels and a decrease in reverberation times (from 0.7 s to 0.3 s at 1 kHz in the unoccupied room). Notable reductions in the numbers of high level noise events associated with visiting time, patient procedures, conversation, meal times and cleaning were also found after the ceiling change. This intervention study has also provided interesting data regarding the level of acoustic absorption provided by the occupants and soft furnishings in a 4-bed bay, with a reduction in the reverberation time of over 0.2 s at most frequencies in the bay before installation of an acoustic ceiling. After the addition of a large area of acoustic absorbency in the form of a ceiling, the effects on the room acoustics of the occupants and soft furnishings becomes negligible. The following chapter presents the results of the objective and subjective surveys at Addenbrookes Hospital, Cambridge while Chapter 10 gives details of the method used to estimate the occupied RTs and its validation using data collected from both Bedford and Addenbrookes Hospitals. 143
Acoustic Design for Inpatient Facilities in Hospitals Addenbrookes Hospital ___________________________________________________________________________________________________________________
9. Addenbrookes Hospital
9.1. Introduction
Three wards at Addenbrookes Hospital also formed part of the subject of the main study which took place over an eight month period from April to November 2010. To provide comparisons between buildings of different age, construction type and ward layout, three wards were identified for the study following liaison with the Estates team: a 1970s tower block of brick and concrete construction; a recently opened modular block of timber construction; and a privately funded building completed in 2006 and built to adhere to the latest acoustic design standards The study wards also provided useful comparisons in terms of the types of patient accommodation and care offered. Ward D8, situated in the 1970s tower block, is a trauma and orthopaedic ward. This ward is large and has a mixture of accommodation ranging from 3-bed bays through to a 12-bed bay. Care is offered to a diverse group of patients; some with severe injuries as a result of road accidents; some with injuries resulting from elective surgery; others suffering from both physical trauma and a level of dementia. Ward N3, situated in the modular block, is a respiratory ward for both acute and chronically sick patients and contains a specialist respiratory unit to care for patients who require noninvasive respiratory ventilation. Accommodation provided here is a mix of single rooms and 4-bed bays. Ward M4, situated in the privately funded Addenbrookes Treatment Centre, is a surgical ward specialising in urology. Accommodation provided here is again a mix of single rooms and 4-bed bays.
9.2. Background
Addenbrookes Hospital opened in 1766 and was one of the first provincial, voluntary hospitals in the UK. By the 1950s the hospital had started to outgrow its original site and in 1959 building began on a new 66-acre site south of Cambridge, with the first phase of the new hospital opening in May 1962.
Figure 9.1 Original building, Addenbrookes Hospital Now, nearly 250 years after its inception, the hospital provides emergency, surgical and medical services for people living in the Cambridge area and offers regional specialist services for organ 144
transplantation, cancer, neurosciences, paediatrics and genetics. Bed numbers currently stand at 1200, with 7000 staff employed and annual inpatient admissions at around 180,000. The site now includes a dedicated maternity and womens hospital, The Rosie, with deals with around 5800 births a year.
9.3. Ward D8 (surgical)

Ward D8 is situated on the eighth floor of the 1970s tower block, known as C & D block. This is a trauma and orthopaedic ward dealing with trauma as a result of an accident and performing elective therapeutic interventions, for example, for problems caused by wear and tear to the hips, joints and shoulders. The ward has total of 35 beds and is the second largest in terms of bed numbers on the Addenbrookes hospital site. This ward is divided between three single rooms, three 3-bed bays, a 4bed bay, a 7-bed bay and a 12-bed bay. Thirteen out of 35 patient beds on this ward are dedicated to care of elderly female patients. These beds, collectively known as the elderly trauma unit, consist of two 3-bed bays and a 7-bed bay and have the highest intervention of nursing on the ward, with four nurses to 13 patients. The unit treats those over 75 years of age who have been involved in an accident, or for example, have broken their hip as a result of a fall. Often these patients are suffering from a number of other complaints such as confusion, dementia or delirium, and consequently this section is considered to be noisy by the ward manager, with patients calling out and banging objects. Elderly male patients admitted to the ward are placed in the standard male accommodation.
9.3.1. Building design

C & D block is a 10 storey naturally ventilated tower block constructed in the early 1970s. Brick built with cavity walls and concrete floors, it is unknown whether it was built to comply with any particular acoustic standard. The building is a mirror image designed around a central lift shaft and stairwell, with C section on one side and D on the other. Internally, apart from general maintenance, the replacement of single glazed windows with double glazed units, and some redecoration, very little has changed since the building was completed. However the wards, which were originally built to house a total of 24 beds, now contain up to 35 beds. Bed spacing is considered to be very poor, leading to a compromise in privacy and dignity. To illustrate this lack of space further, the ward area of D8 is 850 m compared to 1250 m in ward M4 which is situated in the recently opened, privately funded treatment centre, and provides only 28 patient beds. The study ward, ward D8, has little in the way of acoustic absorbency at ceiling level. Most ceilings consist of metal pan tiles which work in conjunction with the heating system, by radiating heat from the hot water pipes running above them. These tiles are perforated, with a layer of insulation covering the water pipes. This may provide some level of acoustic absorbency at certain frequencies, but this 145
2 2
is unknown. An exception to this is the 4-bed female bay situated behind the nurse station, which has a suspended ceiling grid with solid plaster tiles. Other acoustic absorbency on the ward is provided by privacy curtaining which can be pulled fully around each bed; window curtaining; mattresses and bedding. All patient accommodation has heavy duty vinyl flooring and solid plaster walls.
9.3.2. Ward layout

Senior staff offices and the main ward reception area are positioned in the central lobby area which separates C and D wards. A card swipe system is in operation to allow access to ward D8, which is designed around a long interior corridor, with the majority of patient accommodation on one side, and a male 12-bed bay situated centrally at the end of the corridor. Healthcare utilities, a staff room, ward kitchen, day room and three single rooms are situated on the opposite side. There is also a service lift for patient transportation and ward deliveries. Ensuite toilets are provided in each bay, but single sex bathrooms are situated at each end of the corridor. A ward plan is shown in Figure 9.2 on page 150. The 12-bed and 3-bed bays at the end of the ward are male only. The male patients use the toilet and bathroom facilities at this end of the ward corridor to ensure same sex segregation. There is a small nurse station in the centre of this 12-bed bay, with a PC and telephone, which may be a source of some noise, especially at night. The main nurse station is situated directly outside the 4-bed female bay and opposite the three single rooms. Again, a potential source of noise, this area is often bustling with staff and is where the ward clerk is based during office hours. The remaining bays house the elderly trauma unit, which are opposite the healthcare utilities, staff room, and assisted bathroom and ward kitchen. Compared to wards in newer buildings, such as M4 and N3 (the other study wards on this site), this ward feels rather cluttered, with a general lack of storage space, and a narrow dark central corridor.
9.3.3. Ward specific information

The doors to all bays are always left open for observation purposes, with single rooms generally used for infectious patients. If barrier nursing is required, the doors to the single rooms are closed. The ward runs at a very high occupancy level, of around 99%. Staffing levels and shift patterns Staffing levels are highest from 07.15 until 15.15, with ten nursing staff present, seven days a week. This reduces to seven nursing staff for the afternoon shift and to six for the night shift, which lasts from 19.15 to 07.15. Other staff members include a ward clerk who works during office hours from Monday to Friday, and six therapists who work with the patients on weekdays and weekend mornings. There are 13 consultants and seven registrars serving this ward. The doctors rounds generally take 146
place in the morning and during this time there could be between eight and ten doctors on the ward. At weekends the number of doctors drops to three. Three domestic staff also work from 08.00 to 16.00, with an additional staff member working from 16.00 until 20.00. One of these staff members is dedicated to cleaning (moping, damp dusting, changing beds and picking up litter); the other two are housekeepers and are more involved in serving food, beverages and replenishing water jugs. The member of staff working the later shift plays a dual role. Ward routines The first patient visits by clinicians are around 06.00 for general observation, dispensing of drugs and theatre preparation. However, general activity on the ward does not begin until 06.50 when the main lights are switched on. Morning shift handover takes place between 07.15 and 07.45 in the day room, staff room and at the main nurse station in the corridor. After the handover the night staff accompany the day staff to their charges to discuss the patients progress and examine their notes. There can be as many as six people at the end of the patients bed at this time. From 08.00 onwards the ward is very busy with nurses attending to their patients and the arrival of surgeons, therapists, domestic staff and the ward clerk. The ward does not begin to calm down again until late morning when the nursing staff have a break from the patients and have a chance to catch up on phone calls and other administrative aspects of their jobs. From 14.00 until 15.00 there is a designated rest period on the ward which is primarily to help patients get some rest before visiting time begins. Drug rounds and patient observations continue throughout the day at noon, 17.00, and 21.00. A second shift handover takes place when the night staff arrive, between 19.15 and 19.45. As with the morning handover, the day and night staff visit their patients, check notes and discuss patient progress. The ward lights are finally dimmed around 23.00. During the night patients who are particularly unwell have further observations taken at 02.00. Meal times A cold breakfast of cereal and toast is prepared in the ward kitchen and served from a trolley from 08.15 to 09.30. This is followed by a hot drinks trolley at 10.00. A hot lunch arrives on the ward at midday and is plated up in the corridor. A selection of cold supper options is served at 17.00.
147
Visiting times The official hours for visiting time are 15.00 until 20.00. Sources of noise on the ward The following are thought to be the main sources of noise by the ward manager: Doorbell There is a buzzer but it is not felt that this is particularly loud Nurse Call This system has a changeable volume, but is never turned up very loud, even during the day. Telephones / PCs / fax / printers There are four telephones, three PCs, a fax and printer at the main nurse station, and other telephones and PCs on staff desks in the larger multi-bed bays. Patientline This TV, radio and telephone system is available here at a cost. Headphones are provided and any patient not using them would be asked to make use of them for watching TV or listening to the radio. Medical Equipment Alarms Plaster removal No treatment room is available on this ward and plaster casts are removed at the bedside. Deliveries Ward deliveries come up to the ward via the service lift. Large deliveries, such as linen, are delivered in wheeled cages and are generally left in the corridor. Smaller deliveries include pharmacy items.
9.3.4. Managing the study

The ward manager was very supportive of the study and made every effort to ensure that full cooperation was provided by the ward staff. There were no undue concerns regarding the cleaning of the noise measurement equipment and the ward manager spent several hours of his time identifying possible microphone positions that would be acceptable to staff and patients. For comparison purposes it was important that the microphone was placed in similar locations in each accommodation type. A ward plan showing the microphone positions can be seen in Figure 9.2. To avoid the microphone being knocked or contaminated, and in order for it to be as unobtrusive as possible, it was felt that suspending the microphone from the ceiling would be ideal. An identical 300
148
mm bracket to that used successfully at Bedford Hospital was used to suspend the microphone from the ceiling at each measurement position. Questionnaires were reviewed by the ward manager and it was decided that they would be distributed by the ward clerk to those patients who had been on the ward for over 24 hours and were felt to be fit enough to complete the survey. Staff questionnaires were to be left for staff to complete in the staff room. As with the pilot study, a number of laminated posters were displayed throughout the ward common areas. These posters were aimed at both staff and patients and explained in simple terms why and how the study was being undertaken. In addition to these posters the ward manager personally discussed the study with all his staff during staff meetings.
149
Figure 9.2 Detailed plan of ward D8 showing microphone positions
Microphone position
9.4. Overall acoustic survey results Ward D8

Noise level measurements were made at six different locations on the ward. Table 9.1 shows the locations and the time periods of all measurements made, and the patient genders where applicable. Unfortunately it was not possible to make measurements in the single rooms on this ward due to the infectious conditions of the patients. Table 9.1 Ward D8 - measurement locations, time periods and patient gender / types Position Nurse station 12-bed bay 7-bed bay 4-bed bay 3-bed bay A 3-bed bay B Length of measurement period 6 days 2 non-consecutive weeks (14 days) 6 days 8 days 7 days 7 days Patient gender / type N/A Male Elderly Female Male Elderly
Overall measurements of A-weighted equivalent sound pressure levels (LAeq) were averaged and are shown for 24 hours, day time and night time in Table 9.2. Table 9.2 Average LAeq measured for 24 hour, day and night time periods at each location.
Position in ward
Week average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Nurses Station 12-Bed Bay Week 1 12-Bed Bay Week 2 7-Bed Bay 4-Bed Bay 3-Bed Bay A 3-Bed Bay B
57.0 55.4 56.4 56.4 54.3 55.0 56.5
58.1 56.9 57.9 57.8 55.6 56.5 57.8
52.9 47.1 48.2 49.4 48.5 47.1 50.7
A summary of the day and night time average levels presented in Table 9.2 are presented graphically in Figure 9.3 for clarity. As with the wards at Bedford Hospital, levels in all patient accommodation exceed those suggested in the WHO guidelines without exception. Little variation can be seen within the day and night time average noise levels for the patient accommodation, despite the wide range of patient numbers (3 to 12), with day time levels ranging from 55.6 to 57.9 dB LAeq, 16hr and night time levels ranging from 47.1 to 50.7 dB LAeq, 8hr. Day and night levels at the nurse station are, as expected, higher than those measured in the patient accommodation. 151
In terms of the drop in level between day and night, only a relatively small drop of 5 dB was observed at the nurse station, which is comparable with nurse stations on the study wards at Bedford Hospital. This is unsurprising as the nurse station is staffed at all times. The male bays showed the greatest drop in level between day and night of around 10 dB, which is unexpected as one of the male bays measured was the largest bay in the study, containing 12 beds. The elderly patient bays and female bays showed a slightly lower day to night drop of around 7.7 dB, which could possibly be related to patients crying out.
65 Day 60 55 50 45 40 35 30 25 20 15 10 5 0 Nurses Station 12-Bed Bay 3-Bed Bay A 3-Bed Bay B 7-Bed Bay 4-Bed Bay Night
Figure 9.3 Average day and night LAeq levels measured at each location Detailed results of levels measured at the nurse station are discussed in the next section, with further results from the multi-bed bays shown in Section 9.4.2.

Figure 9.4 shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station. It can be seen that the lowest average ambient noise levels are recorded during the later part of the night, where they are around 50 dB LAeq. From 05.30, levels increase steadily and peak at 11.30 when the staff are free to catch up with administrative tasks and telephone calls and are therefore working at the nurse station. Levels remain fairly constant at around 59 dB LAeq during the middle part of the day, with a slight dip following lunch and during the patient rest period. Levels then begin to decrease around 16.00, peaking briefly again during the evening shift handover and then do not decrease substantially until 23.30. This pattern follows the ward routines described in Section 9.3.3. The night time background levels remain very constant at around 42 dB LA90, while day time background levels peak in the middle of the day at around 48 dB LA90. 152
50
40
30
20
Time (24h:00) Nurse station LAeq Nurse station LA90
Figure 9.4 Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station Viewing averaged noise levels over time, as in Figure 9.4 above, provides valuable information with regards to level consistency and overall day and night time variation patterns, but does not illustrate the fluctuating nature of noise in the short term. Figure 9.5 shows noise levels captured at the nurse station over a ten minute time interval at 04.30 with the microphone approximately 1.5 m away from the main desk area. Using the trigger files captured when LAmax exceeds 70 dB, certain high level noise events have been identified.
RINGER BINDERS
DESK DRAWERS SHUTTING
FURNITURE SRAPING
Figure 9.5
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a ten minute interval at the nurse station during the night
The sources of high level noise shown in Figure 9.5 are a good representation of types of high level noise captured at this nurse station during the night and are predominantly related to the
153
administrative tasks undertaken by the staff. Identified events consist mainly of the closing of ring binders, furniture scraping on the floor and desk drawers shutting. Day time high level noise at the nurse station was found to be mostly due to conversation, with some noise again associated with administrative tasks and some occurrences of the internal phone. Although in use on the ward, the nurse call system and doorbell were not captured as sources of high level noise. As discussed in Section 9.3.3, the ward manager confirmed that the volume of both these systems is deliberately turned down. To further illustrate the types and noise levels of typical high level events at the nurse station, examples are presented in Table 9.3. It should be noted that the levels shown are for individual events and may not be representative of every noise event of that type. Table 9.3 Examples of noise events at the nurse station
Noise event Furniture scraping Ring binder Desk draw shutting Internal telephone 9.4.2. Multi-bed bays
LAmax (dB) 78 85; 90 81 68
Figure 9.6 shows the averaged LAeq,1hr levels measured over 24 hours for two 3-bed, a 4-bed, a 7-bed and 12-bed bay. It is clear from the figure that although the numbers of beds in the bays vary considerably, there does not to appear to be a relationship between bed numbers and noise levels, with some of the highest levels measured in the elderly 3-bed bay. Interestingly, the largest bay (12 beds) actually shows one of the lowest averaged LAeq,1hr levels during the night.
154
70 Night time 60
Sound Pressure (dB LAeq, 1hr )
Day time
50
40
30
WHO GUIDELINES
20
Time (24h:00)
12-Bed Bay (Male) 7-Bed Bay (Elderly) 4-Bed Bay (Female) 3-Bed Bay (Male) 3-Bed Bay (Elderly)
Figure 9.6
Average LAeq,1hr levels over 24 hours for the multi-bed bays
The figure also shows that the WHO day / night division is not a particularly good fit. Noise levels increase steadily from around 05.30 rather than 07.00, and begin to decrease after the evening meal is served and then further decrease at 23.00. This suggests it might be appropriate to redefine the day and night time periods for hospital noise assessment and perhaps consider the addition of an evening period. The measured levels essentially follow some very general patterns, climbing steadily as morning activity increases on the ward; peaking as lunch is served and then decreasing slightly during the patient rest period. Levels increase again during the afternoon reflecting additional noise generated during visiting times, and then begin to decrease after the early evening meal. This ward is large, with a diverse group of patients; some with severe injuries as a result of road accidents; some with injuries resulting from elective surgery; others suffering from both physical trauma and a level of dementia. Due to this diversity very different levels and types of care are required on this ward. This in itself leads to less well defined routines of patient care and hence more variation in measured noise levels. This is not necessarily the case in other study wards which deal with more specific types of medical problems or surgical procedures. Further analysis of high level noise sources is carried out in the next section.

Average LAeq,1hr levels over 24 hours, as presented in the previous section, provide some general comparisons and show fluctuations related to ward routines. However, to build up a more detailed 155
picture of the sources of high level noise at each location, the numbers of occurrences of LAmax in 5 dB bands from 70 to 95 dB have been examined. Figures 9.7 and 9.8 show the average number of high level noise events captured during the day and night in different measurement locations. It can be seen in Figure 9.7 that apart from the nurse station, which has already been discussed in some detail in Section 9.4.1, the largest numbers of day time high level noise events are recorded in the 12-bed bay, closely followed by two bays in the elderly trauma unit. This is interesting as the average noise level in the 12-bed bay is very similar to that in the other bays, see Table 9.2 and Figure 9.6. This is a good illustration of why it is necessary to break the data down and fully understand the content of the noise, rather than relying on overall noise levels.
700
70 LAmax < 75 dB 75 LAmax < 80 dB

600
80 LAmax < 85dB 85 LAmax < 90dB
500
90 LAmax < 95 dB
400
300
200
100
Figure 9.7 Average number of high level noise events recorded at each location per day The 12-bed bay has a small desk situated in the centre of the ward with several PCs and a telephone. For logistical purposes the microphone was suspended above this area and thus many of the high level noise sources captured were related to nurse activity at or around this desk, including conversation; talking on the telephone; and the use of ring binders. This desk area was open to the ward and thus any high level sounds captured here would be heard by the patients on the bay. This area was busiest during the day, however high level noise attributable to administrative tasks was also captured during the night a potential disturbance to patients on the bay. Other high noise levels captured can be attributed to the furniture scraping on the floor; the drinks cup dispenser; the jangling of crockery and cutlery at meal times; general movement around the ward and conversation at visiting times. The corridor to the bathroom was situated behind the microphone, and doors banging were often captured at high levels. 156
The 7-bed and 3-bed bays in the elderly trauma unit had many more high level noise events associated with patients calling out and clinicians talking to patients loudly. This had been expected by the ward manager. Patients here are often very confused or distressed, and so to try and engage with, and subsequently comfort the patients, clinicians need to raise their voices. It can be seen in Figure 9.8 that apart from the nurse station, the largest numbers of night time high level noise events are recorded in the elderly trauma unit. As many of the patients are woken up before the designated start of day (07.00), many of the high level sources of noises are related to distress and confusion during this process, which occurs between 06.30 and 07.00. Although listed as night time noise, this gives a misleading indication that these bays are noisy throughout the night, which is not generally the case. The high noise levels during this time are further illustrated by the noise level trace shown in Figure 9.9, which clearly demonstrates high levels of noise from 06.41 to 07.00, much of which can be attributed to a confused patient crying out and the staff trying to calm them. The blue line represents the average LAeq,8hr measured in the bay during the night, further highlighting the high levels of noise shown here.
120

100
80
60
40
20
157
Figure 9.9
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 19 minute interval in the elderly trauma unit
To illustrate the types and noise levels of typical high level events found in the multi-bed bays, examples are presented in Table 9.4. It should be noted that the levels shown are for individual events and may not be representative of every noise event of that type. Table 9.4 Examples of noise events in the multi-bed bays
Noise event Bed rails Patients crying out Drinks cup dispenser Shoes squeaking on floor Trolley Medical equipment alarm
LAmax (dB) 84 85 - 91 78 74 73 71
It can be seen in Table 9.4 that patients crying out are measured at levels from 85 to 91 dB LAmax. Such high noise levels would undoubtedly cause disturbance to other patients on the bay.
9.4.4. Representative measurement interval

In Chapter 6 it was established that a representative measurement interval was one week in duration (5 days when the ward occupancy decreased at weekends). It was felt important that this measurement interval was validated during the main study when it was possible to measure two nonconsecutive weeks worth of noise level data in the same bay. Noise level measurements in the 12bed male bay were captured for two seven day periods, several weeks apart. Figure 9.10 shows the averaged LAeq,1hr levels measured during the two measurement periods. It can clearly be seen that the averaged levels are similar in both level and fluctuation. A goodness of fit test found no statistically significant difference between the two datasets at the 1% level. 158
2
70 Night time Sound Pressure (dB LAeq,1hr ) 60
50
40 Day time 30
20
Time (24h:00)
Week 1
Week 2
Figure 9.10
Average LAeq,1hr levels over 24 hours for two non-consecutive weeks in the 12-bed bay
It can therefore be said that a seven day measurement interval is representative interval for the 12bed bay, and reinforces the findings of the pilot study, that one weeks worth of data (5 days when the ward occupancy decreased at weekends) is a representative measurement interval for the study.
159
9.5. Ward N3 (medical)

Ward N3 is situated on the third floor of a new modular wing. It is a respiratory ward for both acute and chronically sick patients and contains a specialist respiratory unit to care for patients who require non-invasive respiratory ventilation. The ward has a total of 25 beds divided between nine single rooms; two 4-bed bays (one male and one female); and a respiratory care unit which contains two adjoining 4-bed bays (one male and one female). The ward generally runs at 100% occupancy. Weekends may be slightly quieter as there are no routine tests carried out, but occupancy levels remain the same.

Completed in 2009, this modular block is of mainly timber construction, double glazed with both mechanical and natural ventilation. Built to comply with acoustic standard HTM 2045, the suspended ceiling grid is fitted with Ecophon acoustic tiles with good absorption properties; the plasterboard stud walls are sound insulated with 50 mm of mineral wool in the cavity; and the floor is heavy duty vinyl on flooring grade ply, again well insulated in the void. Interestingly, staff working on this ward have found that the floor (which is timber and a little springy) has had a detrimental effect on their feet. The design of the ward, which is L shaped, also makes the staff cover quite large distances each day, potentially adding to this detrimental effect. Complaints regarding the floor were so widespread that remedial work has been carried out to stiffen the floor slightly since it was first constructed. Feedback from staff suggests that this is generally felt to be an improvement. The Estates team were particularly interested to see whether this particular building, which is of untypical construction, suffered from any unusual noise problems. It was not envisaged that these would be related to sound attenuation, but it was considered that the floor construction could potentially magnify certain sounds.
9.5.2. Ward Layout

As can be seen in Figure 9.11 on page , the ward is L shaped with patient accommodation, staff offices and healthcare utilities on both sides of a long central corridor. The entrance to the ward is at the top of the L, with the patient multi-bed bays situated on the longer section of corridor and the single rooms on the shorter section. The siting of the single rooms in this way potentially means less traffic flow in the corridor, as apart from a store at the very end of the ward, there is no reason to walk past the single rooms unless going to see a patient. All patient accommodation has full ensuite bathroom facilities. The nurse station is centrally located at the corridor intersection. It is opposite the 4-bed female bay, the clean and dirty utility rooms and storeroom, with a drug preparation room behind. It may be a 160
source of noise to the female patients in the opposite bay, as not only is the nurse station itself busy, but staff are constantly opening and closing the doors to the utility rooms and store. The other staff administrative areas are situated close to the ward entrance, with the ward clerks desk area, seminar room and general offices based here. The respiratory care unit which contains two adjoining single sex 4-bed bays has its own nurse station.

Doors Doors to the 4-bed bays and single rooms are kept open at all times, except if a patient is infectious and requires barrier nursing (this applies to single rooms only). The doors to the respiratory care unit are shut at night as this ward is perceived as quite noisy due to the amount of equipment in use. This unit has its own dedicated nurse station, which enables the doors to be closed. Staffing levels and shift patterns Staffing levels are highest on the day shift, from 07.15 to 19.15, with eight nursing staff present. This reduces to five nursing staff for the night shift which lasts from 19.15 to 07.15. Other staff members start work at 08.30, and include a ward clerk; four physiotherapists; two occupational therapists; and four other assistants who are either pharmacists or are involved in collecting bloods. There are eight doctors serving this ward, with the doctors rounds generally beginning around 10.00 and continuing into mid- afternoon. Four domestic staff also work on the ward, cleaning and serving meals and drinks During weekends there are slightly fewer staff with six staff nurses and one healthcare assistant. Ward routines The first patient visits by clinicians occur at around 06.00 for general observation; however, general activity on the ward does not begin until 07.15 when the main lights are switched on. The morning shift handover also takes place at this time and is carried out at the bedside, where there are generally three members of staff present. At 08.00 the first drug round begins and staff begin to get patients up ready for breakfast. The majority of patients are bed bound, so staff will sit them up, sorting out toileting and so on. Following breakfast, patients are washed and beds are changed. Beds are often moved around at this time in readiness for admissions and discharges. Doctors begin ward rounds at around 10.00; these continue throughout the morning and into mid afternoon. Drug rounds and patient observations continue throughout the day at noon, 17.30 and 21.00.
161
The night shift changeover takes place at 19.15, and, as with the morning shift, a handover takes place at the bedside. This is followed by patient observations and staff settling patients down for the night. The ward lights are finally dimmed at around 22.30. Meal times A cold breakfast is served by the ward domestic staff at 08.30, followed by a separate hot drinks trolley. Lunch is served at midday, with the food arriving on a heated trolley and subsequently plated up in the ward corridor, next to the nurse station. Meals are taken round to patients on a smaller trolley and are served by the ward domestic staff. This is followed by a separate hot drinks trolley. Afternoon hot drinks are served at 15.00, followed by the evening meal and hot drinks at 18.00. Visiting times Visiting times are strictly adhered to in this ward, except for terminally ill patients. Visiting time is split to allow patients to eat their evening meal without disturbance. The hours are from 14.30 until 17.00, and 19.00 until 20.30. Sources of noise on the ward The following were considered to be the main sources of noise: Daily facilitator meeting This meeting is held each day at 09.15 by the shift co-ordinator at the ward scheduling board next to the nurse station. All staff attend this meeting which generally lasts for 30 minutes. This is a potential source of noise to those female patients in the opposite 4-bed bay. Nurse station Noise levels at the nurse station are felt to be high, with one staff member commenting that there are often multiple conversations taking place and that it is sometimes difficult to hear what is being said on the telephone. Telephone There are two telephones at the nurse station. The ward manager feels that they ring incessantly, especially during the morning. Doorbell This is answered by the ward clerk during working hours; otherwise it is answered by the nursing staff at the nurse station.
162
Patientline This system is available here at a cost. Headphones are provided but are often not used. Mobile phones The ward manager considers the phone charges levied by Patientline to be exorbitant and is therefore lenient towards the use of mobile phones on the ward. Equipment Due to the nature of patient care on this ward, there are a number of pieces of medical equipment that are in use. Ventilators, pumps and vital signs equipment are all common here. Nurse call & emergency call These systems are set on the night time level at all times, and are felt to be ineffectual in terms of design. The ward manager feels that the nurse call system on the ward is poorly designed, as busy staff are often unaware that it has been activated. For example, if a patient in 4-bed bay A presses the nurse call bell, a tone is emitted in this bay, with the occurrence shown on a display panel in this bay and at the nurse station. A light is also displayed outside the bay in the corridor. However, this occurrence is not shown on the display panels in any other patient accommodation, so if nursing staff are all busy out on the ward, they are not necessarily aware that the nurse call has been pressed. The ward manager feels that all occurrences of nurse call should be shown on all screens in all bays. Each time the nurse call is activated, a tone is emitted in the ward managers office, which cannot be cancelled or changed in volume. This is extremely annoying to the ward manager who finds that the nurse call constantly sounding has a negative impact on her work. As the person responsible for overseeing the running of the entire ward, she does not consider it necessary to hear every occurrence of the nurse call, preferring instead to only be able to hear the emergency call bell, which signifies a more serious event. The ward manager also commented that the emergency call bell was ineffectual as it could not be heard throughout the ward and considered that a remote paging system or similar would be preferable to the current systems.

The ward manager was supportive of the study and spent time ensuring that ward routines and systems were understood. There were no undue concerns regarding the cleaning of the equipment and the proposed microphone positions were acceptable. For comparison purposes it was considered important that the microphone was placed in similar locations in each accommodation type. The microphone positions are shown on the ward plan in Figure 9.11.
163
As in previous locations, a 300 mm ceiling bracket was used to suspend the microphone from the ceiling grid. The environmental case housing the SLM was positioned so as to minimise the risk of theft and its impact on staff duties and patient care. In the patient bays the case was placed under a section of worktop, and in the single rooms it was positioned behind an easy chair. The distribution of patient questionnaires was not as straight forward as in previous study wards. Patients who were bed bound and suffering from respiratory conditions were not considered to be fit enough to complete the questionnaire. Only those patients who were able to get up and walk down to the day room would be approached by the ward clerk (whose desk was opposite this room). This meant that the number of patient questionnaires completed during the study period was low (n=13). Laminated posters were displayed throughout the ward common areas explaining the study, and it was hoped that staff would be informed of the research during staff meetings. Unfortunately, this was not the case, and many of the staff approached during the study period had no knowledge of the work being carried out. Questionnaires were left for completion in the staff room, however only a relatively small amount were completed by the staff (n=10).
164
Figure 9.11 Detailed plan of the ward N3 showing microphone positions
9.6. Overall acoustic survey results Ward N3

Noise level measurements were made at six different locations on the ward. Table 9.5 shows the locations and the time periods of all measurements made, and the patient genders where applicable. Table 9.5 Measurement location, time interval and patient type Position Nurse Station 4-bed bay A 4-bed bay B 4-bed bay C Single room J Single room K Length of measurement period 5 days 7 days 8 days 7 days 7 days 9 days Patient gender N/A Male Female (special care) Female Male / Female Male / Female
Overall measurements of A-weighted equivalent sound pressure levels (LAeq) were averaged and are shown for 24 hours, day time and night time in Table 9.6. Table 9.6 Average LAeq measured for 24 hour, day and night time periods at each location.
Position in ward
Week average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Nurse Station 4-Bed Bay A 4-Bed Bay B 4-Bed Bay C Single Room J Single Room K
51.7 50.9 52.1 48.8 49.3 51.1
53.4 52.3 53.3 50.0 50.6 52.5
47.4 44.6 47.2 44.3 43.6 45.1
A summary of the day and night time average levels presented in Table 9.6 are presented graphically in Figure 9.12 for clarity. As with the wards at Bedford Hospital, noise levels in all patient accommodation exceed those suggested in the WHO guidelines. It can also be seen that all average day and night time levels measured in the patient accommodation were very similar, within a 3.3 dB range from 50.0 to 53.3 dB LAeq, 16hr during the day and 3.6 dB range from 43.6 to 47.2 dB LAeq, 8hr at night. Interestingly, the nurse station on this ward has comparable noise levels to those shown in most of the patient accommodation.
166
The day to night drop in levels is fairly consistent, with an average drop of 7 dB in the patient accommodation and a smaller drop of 5 dB at the nurse station. This is comparable with nurse stations on the study wards at Bedford Hospital and Ward D8. The highest day and night time levels were measured in 4-bed bay B, the special respiratory care unit. This is unsurprising given the amount of respiratory equipment in use on this bay.
70 Day Night 60
50
40
30
20
10
0 Nurses Station 4-Bed Bay A 4-Bed Bay B 4-Bed Bay C Single Room J Single Room K
Figure 9.12 Average day and night LAeq levels measured at each location Detailed results of levels measured at the nurse station are discussed in the next section, with further results from the multi-bed bays shown in Section 9.6.2 and levels measured in the single rooms in Section 9.6.3.

Figure 9.13 shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours for the nurse station. The figure shows that noise levels are very steady during the night, from around 21.30 until 05.00, after which they climb to a temporary peak during the daily facilitator meeting at around 09.30. Other small peaks can be observed at lunch time, dinner time and during evening shift changeover. The measured LA90,1hr levels provide a good indication of the variation in background noise levels over time, with night time background levels remaining very constant at around 39 dB LA90, and day time background levels peaking at lunch time at around 45 dB LA90.
167
70 Night time 60
50
40 Day time 30
20
Time (24h:00)
Nurse station LAeq Nurse station LA90
Figure 9.13 Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse station To illustrate the fluctuating nature of noise in the short term, Figure 9.14 shows noise levels captured at the nurse station over a 13 minute time interval during the afternoon, with the microphone approximately 2 m away from the main desk area. Certain high level noise events with LAmax greater than 70 dB have been identified.
BANGING DOORS
RING BINDERS
BANGING DOOR
Figure 9.14
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 13 minute interval at the nurse station during the afternoon
A number of rooms are situated close to the nurse station, including the clean and dirty utility rooms, and a storeroom. All doors to these areas have security access via a key code pad. Staff often have their hands full when entering or leaving these rooms, and so the doors to these rooms are generally left on the latch, avoiding the need to input the security code and making the doors easier to push open. Unfortunately, this has the adverse effect that the door literally bounces when it shuts, causing a loud noise. This is a good example of the ward not being used as its designer intended and is 168
further illustrated by Figure 9.15, which shows the 1/3 octave band frequency spectrum of a typical door bang. Banging doors accounted for 48% of the total number trigger files captured during the measurement period in this location. The acoustic ceiling tiles installed throughout the ward are optimal at speech frequencies. The frequency content of the banging door shown in Figure 9.15 is clearly biased towards low frequencies. This is a good illustration of the type of high level noise for which current levels of absorbency may not be so effective.
Figure 9.15 Frequency content of door bang at the nurse station As in other study wards, the use of ring binders also caused a relatively high number of trigger files at the nurse station (8% of the total number captured during the measurement period). All patient records are kept in ring binders, along with other reference materials, and often are the main source of late night noise, when the staff are catching up with administrative work. To further illustrate the types and noise levels of typical high level events at the nurse station, examples are presented in Table 9.7. It should be noted that the levels shown are for individual events and may not be representative of every noise event of that type. Table 9.7 Examples of noise events at the nurse station
Noise event Doors banging Ring binder Trolley Footsteps
LAmax (dB) 77; 79 72 77 71
Unusually, footsteps are a source of high level noise at the nurse station. This is caused by groups of people walking past the microphone, and not a single person, which would not be sufficiently loud. However, the fact that footfall is responsible for noise levels exceeding 70 dB LAmax indicates a potential issue with the timber floor and its vinyl covering. This is confirmed by the questionnaire responses in which 40% of ward staff found the sound of footsteps annoying, and 20% of patients
169
cited footsteps as a cause of night time disturbance. These percentages are the highest found in relation to annoyance and disturbance from footsteps on any study ward. Unlike other study wards the internal telephone, nurse call and doorbell were not set at levels which were loud enough to create trigger files at the nurse station, however this did not prevent these systems from causing annoyance and interference to staff, as can be seen in Section 9.9, and suggests that noise annoyance is not related to noise level alone.

Figure 9.16 shows the averaged LAeq,1hr levels over 24 hours for the three 4-bed bays and the average background level (LA90,1hr) of all bays. It can be seen that the levels measured in the male and female bay are reasonably consistent, with the male bay appearing slightly noisier during the day and the female bay slightly noisier at night. Measured levels in the special respiratory care unit are
consistently higher than in the other two bays. The reasons for this are investigated further in Section 9.6.4. The average background level varies from around 38 dB LA90 during the night to 42 dB LA90 during the day; these levels are slightly lower than those measured at the nurse station. Levels appear to follow ward routines to some degree, with the levels rising steadily after the main lights are switched on just after 07.00, and temporarily peaking at the end of the staff daily meeting, when nursing staff and doctors return to the wards for the ward rounds. Levels peak again around lunch time. After lunch levels in the respiratory care unit remain fairly stable until around 23.00, but levels in bays A and C can be seen to fluctuate slightly more, with a peak around 18.30 in bay A. This may be related to the serving of the evening meal.
50
40
30
WHO GUIDELINES
20
Time (24h:00)
4 Bed Bay A (Male)
4 Bed Bay B (Special Respiratory Care)
4 Bed Bay C (Female)
Average LA90
Figure 9.16
Average LAeq,1hr for each multi-bed bay and combined average LA90,1hr level for all bays over 24 hours 170
Figure 9.16 also shows that the WHO day / night division is not a particularly good fit. Noise levels increase steadily from around 06.00 rather than 07.00, and begin to decrease after the evening meal is served. This suggests it might be appropriate to redefine the day and night time periods for hospital noise assessment.

Figure 9.17 shows the average LAeq,1hr levels over 24 hours in the two single rooms measured. The average level from the three multi-bed bays is also shown for comparison purposes. It can be seen that levels for single room J are consistently lower than the average of the multi-bed bays, and noise levels measured in single room K are similar to the multi-bed average. It is also noticeable that the effects of the ward routines on the measured levels appear to be much less pronounced in the single rooms.
50
40
30
WHO GUIDELINES
20
Time (24h:00)
Single Room J
Single Room K
Mean 4-Bed Bays
Figure 9.17
Average LAeq,1hr levels over 24 hours for the single rooms
The day time measured LAeq,16hr for single room J is lower than for any other single room measured in any of the study wards. This was in part due to the patient, who was elderly, unable to speak well due to his respiratory problems, slept a great deal and received few visitors. In this instance, patient procedures accounted for the highest percentage of high level noise events, including doctors visits, nurses observations, and patient washing and changing. This patient was bedridden and so required a high level of care. Medical equipment alarms, when they were activated, were typically no louder than 57 dB LAmax.
171
The occupant of single room K had a very bad cough, but was well enough to talk to staff and received more visitors than the patient in room J. High level coughing, clinical activity and conversation accounted for the higher noise levels in this single room.

To help build up a further picture of the sources of high level noise at each location, the numbers of occurrences of LAmax in 5dB bands from 70 to 95 dB have been examined. Figures 9.18 and 9.19 show the average number of high level noise events during the day and night in different measurement locations.
250

200
150
100
50
Figure 9.18 Average number of high level noise events recorded at each location per day It can be seen that 4-bed bays A and B show the largest numbers of high level noise events during the day, with 4-bed bay C registering much fewer. This highlights the limitations of simply viewing the average LAeq,1hr levels, as shown in Section 9.6.2, which indicates similar levels between bays A and C. The average number of day time high level noise events with an LAmax between 70 and 75 dB in bay A is 100 higher than in bay C, suggesting some differences in the noise climate. To explain some of the differences between bays in terms of high level noise, it was found that the nurse call was responsible for 20% of trigger files captured during the seven day measurement period in bay A, with each intermittent tone captured at an average level of 72 dB LAmax. However this was not the case in bays B and C as here the nurse call did not generate any trigger files, only registering at around 51 dB LAmax. Much of the high level noise generated in bay B fell into the unidentifiable category, suggesting more clinical activity in this bay. Conversation and administrative tasks at the integral nurse station also added to these noise sources. 172
It can also be seen in Figure 9.18 that, unusually, the nurse station has more high level noise events with LAmax between 75 and 80 dB, than those with LAmax between 70 and 75 dB. As discussed in Section 9.6.1, this is largely due to the banging doors, which are measured at consistent levels in this range. Figure 9.19 shows the night time average number of high level noise events recorded at each location. In this case it is the nurse station and bay B which show the highest numbers of events. As with the day time noise, the nurse station is affected by doors banging, and other sources of high level noise are mostly generated by administrative activity, especially the use of ring binders.
40
35
30

25
20
15
10
Figure 9.19 Average number of high level noise events recorded at each location per night As discussed previously, bay B also has its own integral nurse station, and it is this area where some of the high level noise was generated. Figure 9.20 shows a further breakdown of night time noise in bay B. It can be seen that patients coughing / sneezing / gurgling and groaning accounts for 45% percentage of trigger files captured during the night. This is unsurprising as patients in this bay have acute respiratory problems. Noise due to administrative tasks carried out at the integral nurse desk accounts for 35% of trigger files; which suggests that patients may be disturbed by activity in this area during the night.
Bathroom door Cupboard door / drawers Ring binder / admin at nurses' desk Furniture scraping Cough / sneeze / gurgling / groaning Unidentifiable
0 5 10 15 20 25 30 35 40 45 50
% of high level noise sources (LAmax > 70 dB)
Figure 9.20 Percentage break down of high level noise events by type in 4-bed bay B 173
9.7. Ward M4 (surgical)

The Addenbrookes Treatment Centre (ATC) was opened in November 2007 at a cost of 84 million. This building was built under the PFI (Private Finance Initiative) scheme and houses operating theatres and wards for emergency surgery, a new endoscopy suite, and facilities for all inpatient gynaecology, urology and breast surgery services. Ward M4 is situated on the 4 floor of the ATC and is a surgical ward specialising in urology. The ward has a total of 32 beds, divided between five 4-bed bays and 12 single rooms, and is predominantly male. Women admitted for treatment on this ward tend to be put into a single room. 60% of admissions are elective (through a GP) and 40% of admissions are emergencies. Ward admissions can be at any time during the day and night, with 60% of patients arriving on the ward on the day of their procedure.
th
9.7.1. Building construction

The construction of the ATC building is primarily concrete, double glazed and mechanically ventilated. Built to comply with acoustic standard HTM 2045, the suspended ceiling grid is fitted with acoustic tiles with good absorption properties; the plasterboard stud walls are sound insulated; and the 1 m thick reinforced concrete floor is covered with heavy duty vinyl.
9.7.2. Ward layout

A number of senior staff offices and meeting rooms are positioned just before the ward entrance, which requires card access. The ward is laid out on both sides of a long straight corridor, with the patient accommodation located on the outer wall of the building, making best use of the natural light and views over the countryside. All patient rooms have ensuite bathroom facilities. Healthcare utilities are situated in the centre of the building (on the left of the corridor when walking from the entrance). Some of these utilities, for example the clean and dirty utility rooms, larger storerooms, larger offices, and the lifts used for patient transport, are shared with another ward which runs parallel to M4, as shown in Figure 9.21 on page 179. There are three nurse stations on the ward, one at the entrance, one halfway down the ward and one at the end of the corridor. The ward manager felt that the central nurse station and the one closest to the entrance were likely to be the noisiest, and so for this reason these areas were chosen for measurement. The ward is designed so that single rooms are positioned in pairs, and due to the location of the ensuite facilities, the rooms are set back, allowing patients in these rooms to be further away from well trafficked corridor areas. This can be seen in Figures 9.21 and 9.22 on pages 179 and 180 respectively. The 4-bed bays open directly out onto the main corridor. The ward corridors and patient accommodation have a spacious and uncluttered feel, and as discussed in Section 9.3.1, the ward area is large (1250 m with 28 patients) compared to ward D8 174
2
(850 m2 with 35 patients). Single rooms have a floor area of approx 18 m2 and 4-bed bays have a floor area of approx 53 m .
2

Doors The ward manager prefers all doors to patient accommodation to be left open at all times to allow the nursing staff to monitor the patients. If a patient falls or gets into difficulties the nursing staff can either see or hear. Even in rooms where patient barrier nursing is required, doors are still left open. In general, the only exception is when a patient is nearly fully recovered and waiting to go home. MRSA All patients are swabbed for MRSA on admittance and if the results are positive they are generally placed in a single room. Prior to the swab result being received, the patients are placed in a room with others whose results are not yet known. Patients are often moved to different locations on the ward during their stay. Occupancy levels During the week the ward runs at 100% occupancy. During the weekend (from Friday evening until Sunday evening) five beds should be closed, however, it appears that this is a rare occurrence. Bay 23 is always fully occupied as this is closest to the nurse station and therefore is good from an observation point of view. Staffing levels and shift patterns Staffing levels from 07.15 until 19.15 are the highest, with seven staff nurses and three healthcare assistants. This reduces to six nursing staff during the night shift, from 19.15 until 07.15. Other staff include three cleaners, two ward assistants and a ward clerk. During weekends there are slightly fewer staff with six staff nurses and one healthcare assistant. Ward routines The first patient visits begin at 06.00 for general observations, including temperature, bloods and urine samples. New admissions begin to arrive on the ward at 07.00. Morning shift handover takes place in the seminar room between 07.15 and 07.45 and is immediately followed by a drug round. The first theatre admissions begin at 08.00, with two sets of porters collecting and returning patients to and from theatre. This continues throughout the day. At this time there is a great deal of noise at the nurse station with doctors checking the admissions board and bed managers arriving on the ward. Doctors begin their ward rounds at this time. 175
At 10.30 the phlebotomists arrive on the ward with two trolleys to collect bloods, this is followed by further drug rounds and general observations which are carried out at 12.30, 17.00 and 21.00. Doctors pay post-operative visits to patients at 18.00 and the night shift handover takes place following this, between 19.15 and 19.45 in the seminar room. The main ward lights are turned out at 22.30. Meal times Breakfast and coffee is served by ward assistants on trolleys between 07.45 and 08.30. This is followed by a mid morning tea round at 10.00. Lunch is served at 12.00, followed by a tea round, with food delivered to the ward in a large hot trolley. Meals are plated up and taken to the patients by hand by two ward assistants and a healthcare assistant. Unlike Bedford Hospital, mealtimes are not protected on this ward and patients may still be undergoing CT scans or being taken to and from surgery during the serving of meals. There is a further tea round at 15.00, followed by a light supper at 17.00. The final tea round is at 20.00. Deliveries Fresh linen is delivered to the ward on a daily basis, at 09.00. The linen is delivered in a large cage which is taken from the lift lobby to the central nurse station and swapped for the dirty linen trolley. Thursday is the main day for ward deliveries, with items brought up to the ward in large cages after lunch. Visiting times The official hours for visiting time are 14.00 until 20.00. Sources of noise on the ward The following are thought to be the main sources of noise: Doorbell The entrance to the ward is through a security door, which has a doorbell, the use of which peaks during visiting time. Cleaning Cleaning begins in the morning and continues throughout the day and comprises of dusting and mopping the floor. Full floor buffing takes place on a weekly basis. Bins are changed twice daily (more if required).
176
Doors Fire doors are heavy and tend to slam loudly. Nurse Call Telephone There are 3 telephones on the ward. The ward manager felt that during busy times the phones ring endlessly as there is no member of staff free to answer them. A cordless phone was tried as an alternative, but this was not entirely successful and was eventually lost. Patientline This system is available here at a cost. Headphones are provided but are not always used. Portable DVD players Patients sometimes bring these in to watch their own DVDs. Mobile Phones To avoid setting a precedent, even the doctors tend to go out into the hallway to use their mobile phones.

The ward manager was supportive of the study and spent time ensuring that ward routines were understood. Particular interest was expressed in measuring noise levels at two of the nurse stations for comparison purposes, as it was suspected that the central station was adversely affected by noise from the nearby lifts, and had become a congregation point for staff. Both areas were incorporated into the study. Two four bed bays and two single patient rooms were identified as suitable study locations with input from the ward manager. For each accommodation type, similar positions were found for the microphone, which was suspended from the ceiling grid using a 300 mm ceiling bracket. The environmental case housing the SLM was positioned so as to minimise the risk of theft and its impact on staff duties and patient care. In the patient bays the case was placed on a section of worktop, and in the single rooms it was positioned behind an easy chair. A ward plan showing these positions can be seen in Figures 9.21 and 9.22. Although the ward clerk was tasked with passing the questionnaires to patients who had been on the ward for over 24 hours and were deemed fit enough to complete the questionnaire, this was not as successful as in other study wards. The role of ward clerk rotated between a number of staff and so the questionnaires were often forgotten; with only 14 completed questionnaires collected.
177
Laminated posters were displayed throughout the ward common areas explaining the study, and it was hoped that staff would be informed of the research during staff meetings. Unfortunately, this was not the case, and many of the staff approached during the study period had no knowledge of the work being carried out. Questionnaires were left for completion at the nurse station, and a box was provided for completed questionnaires, however only a relatively small number (10) were completed by the staff.
178
Figure 9.21 Plan of Ward M4 detailing shared areas and microphone positions
Figure 9.22 Detailed plan of Ward M4 showing study locations and microphone positions
9.8. Overall acoustic survey results Ward M4

Noise level measurements were made at six different locations on the ward. Table 9.8 shows the locations and the time periods of all measurements made, and the patient genders where applicable. Table 9.8 Position Nurse station 1 Nurse station 2 4-bed bay A 4-bed bay B Single room A Single room B Measurement location, time interval and patient gender Length of measurement period 5 days 7 days 6 days 8 days 7 days 9 days Patient gender N/A N/A Male Male Female Male / Female
Overall measurements of A-weighted equivalent sound pressure levels (LAeq) were averaged and are shown for 24 hours, day time and night time in Table 9.9. Table 9.9 Average LAeq measured for 24 hour, day and night time periods at each location
Position in ward Week average of A-weighted equivalent sound pressure levels 24 hours LAeq, 24hr Day time LAeq, 16hr Night time LAeq, 8hr
Nurse Station 1 Nurse Station 2 4-Bed Bay A 4-Bed Bay B Single Room A Single Room B
56.4 52.2 50.6 52.4 50.9 53.2
55.1 53.6 52.7 54.0 52.3 53.5
48.1 46.7 43.0 43.4 42.1 49.1
A summary of the day and night time average levels presented in Table 9.9 are presented graphically in Figure 9.23 for clarity. As with the wards at Bedford Hospital, noise levels in all patient accommodation exceed those suggested in the WHO guidelines without exception. It can also be seen that all average day time levels measured in the patient accommodation were very similar, within a 1.7 dB range from 52.3 to 54.0 dB LAeq, 16hr. Night time levels were also very consistent with a 1.3 dB range from 42.1to 43.4 dB LAeq,
8hr,
with the exception of single room B which had a much
higher average of 49.1 dB LAeq, 8hr . Day to night drop was also fairly consistent with an average drop of 10.2 dB in patient accommodation, with the exception of single room B which had an average drop of just 4.4 dB.
181
As the ward manager suspected, noise levels were higher at the central nurse station (nurse station 1), however the differences between the nurse stations were very small, with a 1.5 dB difference in average levels during the day and 1.4 dB difference at night. Further analysis is necessary to understand the level fluctuations and sources of high level noise in these two cases, which are discussed in Sections 9.8.1 and 9.8.4. Results from the multi-bed bays and single rooms are presented in Sections 9.8.2 and 9.8.3 respectively.
70 Day Night 60
50
40
30
20
10
0 Nurse Station 1 Nurse Station 2 4-Bed Bay A 4-Bed Bay B Single Room A Single Room B
Figure 9.23 Average day and night LAeq levels measured at each location

Figure 9.24, shows the averaged LAeq,1hr and LA90,1hr levels over 24 hours for the two nurse stations. The figure shows that noise levels at Nurse Station 1 (NS1) are consistently higher than those measured at Nurse Station 2 (NS2) during the day and then converge at around 21.00 for four hours. Levels at NS2 are again consistently lower from 01.00 until 05.00, suggesting less activity in the area at this time. Average levels at NS1 show several peaks in line with the ward routines. Levels rise to an initial high at around 08.00, when the doctors and bed managers arrive at the nurse station to check the days schedules. Another peak can be seen around lunch time and then, as suspected by the ward manager there is a peak at the end of the day shift. Background levels during the night are around 37 LA90,8hr, and provide a good indication of the overall building services noise.
182
70
Night time 60
Day time
50
40
30
20
Time (24h:00)
LAeq Nurse Station 1
LAeq Nurse Station 2
LA90 Nurse Station 1
LA90 Nurse Station 2
Figure 9.24 Average LAeq,1hr and LA90,1hr levels over 24 hours at the nurse stations Figure 9.25 shows noise levels captured at NS2 over a 15 minute time interval at 05.30 in the morning, with the microphone approximately 2 m away from the main desk area. Using the trigger files captured when LAmax exceeds 70 dB, certain high level noise events have been identified.
CLOSING DOOR
BANGING DESK DRAWERS CLOSING DOORS
Figure 9.25
LAmax,2s (green trace) and LAeq,2s (red trace) fluctuating over a 15 minute interval at the nurse station at 05.30
The door to the dirty utility room is situated behind NS2 and each time the door closes it generates a noise with an LAmax of 72 dB. This room is used a great deal and is responsible for many trigger files, both day and night. Other sources of night time noise at this nurse station are generally related to administrative tasks and include ring binders, and desk drawers banging shut. High level noise events during the day are similar, with more high level conversation and noise from corridor traffic, such as trolleys.
183
High level noise events captured at NS1 are generated in the main by high levels of conversation. This nurse station is larger than NS2 and many more staff congregate here. As with NS2, administrative tasks are also responsible for some high level noise events, with a ring binder captured at 83 dB LAmax. Interestingly, although the nurse call, doorbell and internal telephone are all cited by staff in this ward as annoying and interfering with their ability to carry out their jobs effectively (see Section 9.9.1), none of these systems is loud enough to generate high level noise events, that is LAmax greater than 70 dB.

Figure 9.26 shows the averaged LAeq,1hr and background levels (LA90,1hr) levels over 24 hours for the two 4-bed bays. It can be seen that the levels measured in bays A and B are reasonably consistent, with bay B appearing slightly noisier during the main part of the day. The average background level in bay B is consistently lower at night, at around 32 dB LA90, with bay A around 35 dB LA90. An explanation for these differing levels could be the proximity of the main nurse station. Bay A is opposite the ward lifts and close to NS1, whereas bay B is opposite an aided bathroom and store, which would be little used at night. As discussed previously doors to the multi-bed bays are left open at all times, and thus noise from the ward lifts and NS1 may affect background levels.
70
Night time 60
Day time
50
40
30
WHO GUIDELINES
20
Time (24h:00)
LAeq 4-Bed Bay A LAeq 4-Bed Bay B LA90 4-Bed Bay A LA90 4-Bed Bay B
Figure 9.26
Average LAeq,1hr and LA90,1hr level for each multi-bed bay over 24 hours
Levels appear to follow ward routines to some degree, with the levels rising steadily at around 06.00, when the first observation round begins. At around 08.00, when the doctors arrive on the ward and theatre admissions begin, levels increase sharply. A small dip can be seen just before the start of visiting time, with levels decreasing slowly after the evening meal is served. 184
Figure 9.26 also shows that the WHO day / night division is not a particularly good fit. As discussed, noise levels increase steadily from around 06.00 rather than 07.00, and begin to decrease after the evening meal is served. This suggests it might be appropriate to redefine the day and night time periods for hospital noise assessment. The microphone situated in the multi-bed bays was suspended from the ceiling in the corner of the room close to the entrance doors. In both bays this was over a small worktop area, where there was also a rubbish bin, plastic apron and glove dispenser, sink and some hospital wheeled equipment stored. To avoid any coloration of results due to the use of the sink and bin, this location was chosen in both bays, to allow for comparisons. Further discussion of high level noise can be found in Section 9.8.4.

Figure 9.27 shows the average LAeq,1hr levels over 24 hours in the two single rooms measured. The average level from the two multi-bed bays is also shown for comparison purposes. It can be seen that at night, levels for single room A are consistently lower than the average level of the multi bed bays and levels for single room B are consistently higher than this average. During the day all levels are comparable, following similar patterns of fluctuation
50
40
30
20
Time (24h:00)
LAeq Single Room A LAeq Single Room B Mean LAeq 4-Bed Bays LA90 Single Room A LA90 Single Room B
Figure 9.27
Average LAeq,1hr and LA90,1hr levels over 24 hours for the single rooms
Single room A is situated opposite the main nurse station; the internal phone ringing and staff talking loudly and laughing can often be heard in the background when reviewing trigger files of high level noise events inside this room. However, it is assumed that the door to this room is shut at night,
185
resulting in the extremely low background noise levels measured, at around 24 dB LA90. These are the lowest background levels measured in any study ward. During the measurement period there were three different patients occupying single room A. All were female, and relatively quiet. Sources of high level noise were routinely caused by room cleaning, bin bag changing and use of the rubbish bin, which was captured at a level of 81 dB LAmax. Visits by clinicians and the serving of meals also accounted for a percentage of high level noise, with domestic staff talking to the patients unnecessarily loudly at times. The patients in single room B during the measurement period were male, with sources of high level noise found to be similar to those in room A. Additional high level events were the closing of the room door, measured at around 78 dB LAmax, and persistent coughing of the second patient staying in the room at levels up to 79 dB LAmax. Night time LAeq levels in this room were on average 7 dB higher than those measured in single room A, with background levels around 13 dB higher. It is thought this was due to the constant use of a portable cooling fan and some low level alarms of the monitoring equipment in this room.

To help build up a wider picture of the sources of high level noise at each location, the numbers of occurrences of LAmax in 5 dB bands from 70 to 95 dB have been examined. Figures 9.28 and 9.29 show the average number of high level noise events during the day and night in different measurement locations. It can be seen that the main nurse station (NS1) accounts for the majority of high level noise events, with an average of over 450 events during the day between 70 and 75 dB LAmax. As already discussed, much of this is due to high level conversation between staff, and some administrative tasks. Much of the high level noise at NS2 is due to the banging shut of the door to the dirty utility room; the effect of this which may be to build up an artificially high picture of noise levels in this area. 4-bed bay B shows the largest numbers of high level noise events measured in a multi-bed bay during the day, with 4-bed bay A registering much fewer. Here, the specific difference appears to be primarily due to a patient with a loud and persistent cough in bay B. All other high level events are similar, with the use of the rubbish bin, plastic apron and glove dispenser and sink noticeable in both bays. A loud bang is consistently captured as the lid of the rubbish bin lid closes, measured at an average level of 83 dB LAmax. With the nearest patient in a bed only a short distance away from this bin, this may cause annoyance. Although not loud enough to create a trigger file, music can sometimes be heard in the background of trigger files captured in the bays. This is substantiated by a patient comment made regarding the use of radios and TVs without enforcing the use of headphones.
186
500
70 LAmax < 75 dB
450
75 LAmax < 80 dB 80 LAmax < 85dB
400
85 LAmax < 90dB 90 LAmax < 95 dB
350
300
250
200
150
100
50
Figure 9.28 Average number of high level noise events recorded at each location per day
Figure 9.29 shows the night time average number of high level noise events recorded at each location. In this case it is NS2, NS1 and single room B respectively which show the highest numbers of events. As with the day time noise, NS2 is affected by the door of the dirty utility room banging, whereas the majority of high level noise at NS1 is caused by high levels of conversation and administrative tasks. As discussed previously, there was a patient in single room B with a loud, persistent cough that accounts for the numbers of high level noise sources captured during the night.
50
70 LAmax < 75 dB 75 LAmax < 80 dB 80 LAmax < 85dB 85 LAmax < 90dB
40
30
20
10
187

The design of the questionnaire surveys was discussed in detail in Section 5.4. Staff response was good in Ward D8 with 21 questionnaires completed, but response rate in the other wards was low, with only 10 staff completing the survey in wards N3 and M4. The following sections discuss results from the staff questionnaires and examine the differences between perceptions on the three wards.

To establish certain attributes about the staff, the first section posed a number of basic questions. Out of all the respondents, 73% were female and 27% male, with similar percentages in all three wards. Figure 9.30 shows the ages of the respondents. It can be seen that in all wards the respondents were generally younger than 50, with a higher percentage of young staff members completing the questionnaire in wards D8 and N3.
60+
Age band (years)
51-60 41-50 31-40 20-30 Less than 20 0 20 40 60 80 100 D8 (n=21) N3 (n=10) M4 (n=10)
Percentage of respondents
Figure 9.30 Age of respondents by band The length of time worked both on the wards and at the hospital are shown in Figures 9.31 and 9.32 respectively, which suggest that many of the staff have worked at the hospital for longer than they have worked on their specific ward.
188
Time worked on the ward
5+ years D8 (n=21) 4-5 years 3-4 years 2 - 3 years 1- 2 years < 1 year 0 10 20 30 40 50 60 70 N3 (n=10) M4 (n=10)
Time worked at the hospital
5+ years 4-5 years D8 (n=21) 3-4 years 2 - 3 years 1- 2 years < 1 year 0 10 20 30 40 50 N3 (n=10) M4 (n=10)
Figure 9.31 Time worked on the ward
Figure 9.32 Time worked at the hospital
9.9.2. Noise annoyance

General feelings of noise annoyance were investigated by asking staff to what extent they were annoyed by noise. Figure 9.33 shows that the highest percentages of staff on all wards were slightly annoyed by noise. Ward D8 shows the most diverse response, with some staff indicating not at all annoyed and a small proportion choosing extremely annoyed.
Staff perception of noise annoyance
Extremely Very much Moderately Slightly Not at all 0 20 40 60
D8 (n=21) N3 (n=10) M4 (n=10)
80
100
Figure 9.33 Staff perception of noise in terms of annoyance
Staff were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Figure 9.34 shows the percentages of staff who rated a noise event with a 2, 3 or 4, and so could be said to be more than a little annoyed by the event. It can be seen that the four events rated by a high percentage of staff in Ward M4 were the doorbell (80%), internal telephone (70%), medical equipment alarms (60%), and the nurse call (60%). Staff in wards N3 and D8 also rated these events, but the percentage of those annoyed by the internal telephone, nurse call and medical equipment alarms was 10 - 20% lower. The doorbell was not a major source of annoyance on wards N3 and D8. 189
Other events cited by 30% or more staff on the wards were people talking, and staff talking on the telephone. Visiting time; the talking on and ringing of mobile phones; and TV / radio use were all rated more highly in wards D8 and N3 than M4. This maybe an indication of a more lenient approach by ward staff. Some anomalies can be seen in Figure 9.34. Trolleys, footsteps, rubbish bins and external noise were all cited as annoying by 40% of respondents in N3, but much less so (or not at all) in the other wards. The timber floor construction of N3 is thought to increase the noise of trolleys and footsteps in this ward, as discussed in Section 9.5.1. External noise annoyance may have been exacerbated as the study was carried out during warmer weather in this ward and so more windows may have been open at the time of the survey.
Visiting time Talking on mobile phones Mobile phones ringing TV / radio Meal times Trolleys Rubbish bins Cleaning People talking Medical Equipment Footsteps Doorbell Nurse call Staff talking on the telephone Internal telephone Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
D8 (n=21) N3 (n=10) M4 (n=10)
Figure 9.34 The percentage of staff rating an annoyance noise event with a 2, 3 or 4
9.9.3. Interference with work

Respondents were asked to what extent noise interfered with their ability to work effectively. As can be seen in Figure 9.35, opinion of the respondents in all the wards was very split, with staff in ward D8 seemingly slightly more adversely affected by noise than in the other wards, with 33% choosing moderately.
190
Staff perception of noise interference
Extremely D8 (n=21) Very much N3 (n=10) M4 (n=10) Moderately
Slightly
Not at all
20
40
60
80
100
Figure 9.35 Staff perception of the extent to which noise interferes with work
Staff were also asked to rate how much each noise event interfered with their ability to carry out their job effectively (again the rating scale of 0 to 4 was used). Figure 9.36 shows the percentages of staff who rated a noise event with a 2, 3 or 4, and so it could be said that this noise event interfered to some extent with their ability to carry out their job effectively.
Visiting time Talking on mobile phones Mobile phones ringing TV / radio Meal times Trolleys Rubbish bins Cleaning People talking Medical Equipment Footsteps Doorbell Nurse call Staff talking on the telephone
D8 (n=21)
Internal telephone
N3 (n=10)
Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
M4 (n=10)
Figure 9.36 The percentages of staff rating an interference noise event with a 2, 3 or 4
191
It can be seen that the numbers of staff rating events as interfering with their work are generally fewer than for annoyance. However, the internal telephone is still consistently rated by 50% or more staff, with visiting time and the nurse call both still rated by 30% or more respondents in each ward. Medical equipment, which was rated as annoying by 50% or more, is rated less severely in the case of interference, presumably as staff feel that they need to hear these alarms to make necessary decisions. There are several anomalies worth noting. TV / radio usage, trolleys, rubbish bins, footsteps, doors banging and external noise are all rated by more staff in Ward N3. As mentioned in relation to annoyance, some of these events may be exacerbated by the construction of the timber floor. Banging doors has also shown to be a problem specifically in the area surrounding the nurse station in this ward (see Section 9.6.1).
9.9.4. Important sounds

To aid understanding of which sounds were felt by staff to be important to be heard in order to carry out their jobs effectively, staff were asked to rate different noise events on a scale of 0 to 4, where 0 indicated not at all important and 4 indicated extremely important. Figure 9.37 shows the mean ratings for each noise event. It can be seen that patients calling out were considered by staff in all wards to be the most important noise event. However, the average ratings were consistently high in all cases suggesting that all of these events are important for staff.
4
D8 (n=20) N3 (n=11) M4 (n=10)
0 Nurse call Conversations with colleagues Conversations with patients Medical equipment alarms Patients calling Patient activity out
Figure 9.37
Mean importance rating of certain noise events 192
9.10. Results of the patient questionnaires

With the help of the ward clerks, questionnaires were distributed to those patients who had been on the ward for over 24 hours and were judged to be physically and mentally fit enough to complete the survey. In total 74 patients completed the questionnaire: 47 in Ward D8; 13 in Ward N3; and 14 in Ward M4. The following sections discuss results from the patient questionnaires and examine the differences between perceptions on the three wards.
9.10.1. Patient profiles

As with the staff questionnaire, the first section aimed to establish certain attributes about the patients, beginning with the question of gender. As discussed previously, wards D8 and N3 are split male / female wards and ward M4 is predominantly male. This is shown clearly in Figure 9.38 below:
D8 (n=47) Female N3 (n=13) M4 (n=14)
Male
20
40
60
80
100
Percentage (%)
Figure 9.38 Gender split by ward type Respondents were asked for their age range, and as shown in Figure 9.39, a high percentage of patients were older, with 60% in the respiratory ward (N3), and 50% in wards D8 and M4 aged 60 years or above.
60+ 51-60
Age range
41-50 31-40 20-30 < 20 0 10 20 30 40 50
D8 (n=47) N3 (n=13) M4 (n=14)
60
70
Percentage (%)
Figure 9.39 Patients age by band
193
Respondents were asked how long they had been on the ward, and it can be seen in Figure 9.40 that the majority had been on the ward for less than one week. Ward D8 shows the longest stay patients, which is unsurprising due to the variety of conditions treated on this ward, with some patients admitted for elective surgery and others on the ward as a result of a serious accident, where recovery may be substantially longer. Length of stay in Ward N3 is more likely to be more variable, with some patients suffering from very serious respiratory conditions remaining on the ward for several weeks. Ward M4 is a surgical ward where the procedures utilised are of a more standard nature. As such patients on this ward tend to be discharged more quickly.
3+ weeks
Length of stay
D8 (n=47) N3 (n=13)
2 - 3 weeks
M4 (n=14)
1- 2 weeks
< 1 week
20
40
60
80
100
Percentage (%)
Figure 9.40 Length of patient stay when completing the questionnaire
The presence of a hearing impairment was also explored, with 39% of respondents on Ward D8, 31% on N3 and 14% on Ward M4 indicating that they did suffer to some degree. The high incidence of hearing impairment on wards D8 and N3 probably reflects the age profile on these wards. However it is not clear why the incidence is lower on ward M4 which has a similar age distribution; it may be due to the fact that the majority of patients on this ward are male and may be more reluctant to admit to a hearing problem. The bed number of the respondent was noted on the front of the questionnaire by the ward clerk. This number provided useful location information which is considered when investigating relationships between bed positioning and patient accommodation type and noise annoyance and disturbance, which are explored in Chapter 11. In terms of the single room / multi bed bay split, 99% of respondents in Ward D8 were staying in multi-bed bays, with 85% in each of the wards N3 and M4.
9.10.2. Noise annoyance and disturbance

The questionnaire sought to identify the sources of noise that may annoy or disturb patients. Respondents were given two lists of noises and were asked to rate the day time annoyance and night time disturbance on a scale of 0 to 4 (where 0 indicated no annoyance / disturbance and 4 indicated a
194
great deal). Several lines were left blank at the bottom of the lists for patients to add and rate additional noise sources. Patients were first asked how they perceived the day time noise environment on the ward. Figure 9.41 details the responses, which shows over 60% of patients in Ward D8 found the ward a little noisy, with similar percentages finding Wards N3 and M4 to be quiet. Interestingly, when asked whether they were annoyed by noise, a relatively low percentage (26%) of patients in Ward D8 felt annoyed, with even lower percentages of 15% and 21% of patients in Wards N3 and M4 respectively.
Patient perception of the daytime ward environment
Extremely noisy Very noisy A little noisy Quiet M4 (n=14) Very quiet 0 20 40 Percentage (%) 60 80
D8 (n=47) N3 (n=13)
Figure 9.41 Patient perception of the day time ward noise environment
The patients who had indicated that they were annoyed by noise during the day, were then asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. With relatively small patient samples in wards N3 and M4, the number of people annoyed by day time noise was too low for any meaningful analysis (n=2 and n=3 respectively). However in ward D8, which had a larger sample, 12 patients rated day time noise as annoying, and consequently their ratings are worth examining further. Figure 9.42 shows the percentage of patients within this sample who rated a noise event with a 2, 3 or 4, and as such could be said to be more than a little annoyed by the event.
195
Patients crying out Talking on mobile phones Mobile phones ringing TV / radio Meal times Visiting time Trolleys Rubbish bins Cleaning People talking Medical Equipment Footsteps Nurse call Staff talking on the telephone Internal telephone Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
D8 day time annoyance (n=12)
% of patients who rated each event 2 or above in terms of day time noise annoyance
Figure 9.42
The percentages of patients on Ward D8 rating an annoyance noise event with a 2, 3 or 4
It can be seen that patients crying out and the internal telephone are the most highly rated noise events, with nearly 60% of patients in the sample annoyed by each source. Ward D8 has a female only elderly trauma unit of 13 beds, with any elderly male patients admitted on the ward sharing the same bay accommodation as the other male patients. Many of the elderly patients suffer from a degree of confusion or dementia and are likely to cry out, but of course patients who are in a great deal of discomfort will also vocalise their pain. There are four internal telephones at the main nurse station in Ward D8, and in some of the other patient bays there are staff desks with a telephone. If these phones are left unanswered, or take some time to divert, this could explain this level of annoyance. Visiting time, medical equipment alarms and people talking are also cited as annoying by over 30% of patients in this ward. Two patients in Ward D8 added an additional noise event that they themselves found to be annoying during the day. The events were: Staff in a performing mood External building work Patients were asked how they perceived the night time noise environment on the ward. Figure 9.43 details the responses, which can be seen to be a little more split than for daytime noise annoyance, with several patients having more extreme perceptions of night time noise.
196
When asked whether they were disturbed by noise at night, 51% of patients in Ward D8; 33% of patients in Ward N3; and 57% of patients in Ward M4 felt they were. Thus overall approximately 50% of patients were disturbed by noise at night; this figure is similar to that in Bedford Hospital.
Patient perception of the night time ward environment
Extremely noisy
Very noisy
A little noisy
D8 (n=47) N3 (n=13)
Quiet
M4 (n=14)
Very quiet 0 10 20 30 40 50
Percentage (%)
Figure 9.43 Patient perception of the night time ward noise environment Patients who had indicated that they were disturbed by noise during the night were asked to rate the annoyance of various noise sources on a scale of 0 to 4, with 0 indicating not at all annoying and 4 indicating a great deal. Sample sets were higher than for the day time annoyance with n=23 for Ward D8, but were small for wards N3 and M4 (n=5 and n=7 respectively). Again it is possible that in these cases, only those patients who had felt they had something specific to say about noise may have chosen to take part in the survey, therefore skewing the results to a degree. Figure 9.44 shows the percentages of patients within these samples who rated a noise event with a 2, 3 or 4, and so could be said to be more than a little disturbed by the event. It can be seen that of those patients who were disturbed by night time noise in Ward D8, around 40% of respondents found that patients crying out, people talking, medical equipment alarms and the internal telephone were disturbing. This was a very different split to the other wards. In Ward M4 it was the internal telephone that was rated as disturbing by the most respondents over 70%. This far exceeded the ratings for any other sources of disturbance on Ward M4, with doors banging, staff talking on the telephone and patients crying out, rated by around 30%. Ward N3 again showed differences, with the nurse call system cited as disturbing by the highest percentage of patients (60%), and trolleys, people talking, medical equipment and doors banging cited by 40% of respondents.
197
Patients crying out
D8 night time annoyance (n=23)

Talking on mobile phones Mobile phones ringing TV / radio Trolleys Rubbish bins People talking Medical Equipment Footsteps Nurse call Staff talking on the telephone Internal telephone Doors banging External noise 0 10 20 30 40 50 60 70 80 90 100
N3 night time annoyance (n=5) M4 night time annoyance (n=7)
% of patients rating night time disturbance event of 2 or above
Figure 9.44 The percentages of patients rating a disturbance noise event with a 2, 3 or 4
It can be seen that mobile phones ringing, talking on mobile phones and rubbish bin noise are cited as a disturbance only on Ward D8. It is possible that, with regards to the use of mobile phones on this ward, the policy may be more lenient than on wards N3 and M4. Also, the rubbish bins in this ward are possibly older metal bins, rather than bins of a newer, quieter design installed in some of the more recently built wards. Four patients added an additional noise event that they themselves found to be disturbing at night. The events were: Snoring (D8) Generally noisy bed neighbours (D8) Vibration of the floor (N3) Nurses talking (N3) Interestingly, the timber floor construction in Ward N3 does appear to add to the disturbance in some cases, with a patient specifically citing vibration of the floor, and higher numbers of patients on this ward disturbed by trolleys (40%) and footsteps (20%) than on Wards D8 and M4.
9.10.3. Positive sounds

Looking at sound in a positive rather than in a negative light, patients were asked if there were any sounds that they actually found comforting. Most patients left the answer blank (70% in ward D8; 77% in ward N3; and 86% in M4), however, there were nineteen completed responses which were similar 198
in content to those from Bedford Hospital. Responses included listening to music on the radio, knowing that the nursing staff were nearby to provide care, the tea trolley, droning sounds (such as the hoover and floor cleaner), and maintaining some connection with the outside world. The full responses can be seen in Appendix B. Respondents were also asked if they felt that there was ever too little sound in a room. Only three patients in total said that they did. Unfortunately, two of these patients did not complete the details regarding their accommodation, but the other respondent was in a single room.
9.10.4. Ease of hearing and privacy

Patients were asked whether high levels of background noise may at times make it difficult to hear doctors and nurses who talk to them. 42% of respondents on Ward D8 felt that this was the case, with this percentage made up of 12 hearing impaired patients and six patients who did not indicate any hearing problem. On Wards N3 and M4 however, only one respondent from each felt that high levels of background noise made it difficult to hear, and both these patients were hearing impaired. Conversational privacy was investigated by asking whether the patient felt that they could have a private conversation at their bedside. The lowest percentage of patients who felt they could speak privately was in Ward D8 (59%), which is unsurprising given the bay sizes and bed spacing in this ward. Out of those who felt they could speak privately, 56% of patients on this ward felt that they would need to lower their voice. On wards N3 and M4, 69% and 79% of patients respectively felt that they could speak privately at their bedside. Out of those who felt they could speak privately, 71% on Ward N3 and 73% on Ward M4 felt that they would have to lower their voice. All respondents in single patient rooms on these wards (n=4) were happy with conversation privacy.
9.11. Questionnaire comments

Staff and patients were invited to make additional comments at the end of the questionnaire if they wished. Very few staff made comments, but many patients did leave some feedback which was very varied. Several patients cited noise from the wearing of high heeled shoes; crying out and shouting of other patients was disturbing or distressing; and noise attributable to visiting times and the lack of enforcement of visiting hours by staff was mentioned by several patients; A detailed list of these comments is shown in Appendix B.
9.12. Summary
This section summarises the main findings from the study of the three wards at Addenbrookes Hospital: The nurse station in Ward D8, had the highest average noise levels of all the nurse stations measured in the three study wards, with day time levels of around 58 dB LAeq and night time levels of around 53 dB LAeq. Typical sources of noise here included high levels of conversation, 199
administrative tasks, furniture scraping on the floor and the closing of desk drawers. The nurse station in ward N3 had the lowest measured levels, with day time levels of around 53 dB LAeq and night time levels of around 47 dB LAeq. 48% of high level noise events captured here were due to the banging of the clean and dirty utility room doors, and so levels could be lowered still further, with different opening mechanisms installed on these doors. The smaller nurse station in Ward M4, also suffered from noise due to the banging of the dirty utility room door. Other sources of high level noise captured at the nurse stations in this ward were primarily due to high levels of conversation, and corridor traffic. Noise level measurements made in the multi-bed bays in Ward D8 were consistently higher both during the day and night than on the other wards. This ward provided a mixture of patient accommodation with 3-bed, 4-bed, 7-bed and 12-bed bays. Noise levels were very consistent throughout , and did not appear to be affected by the number of patients occupying the bays, with some of the highest levels measured in a 3-bed bay for elderly patients. Many of the high level noise events identified were related to activity at the nurses desk in the 12-bed bay, and confused elderly patients crying out in the bays in the elderly trauma unit. Noise levels measured in the multi-bed bays in Wards N3 and M4 were similar, with day time levels of around 53 dB LAeq and night time levels of around 44 dB LAeq, with little variation between the bays. However, subsequent investigation of the numbers of high level noise events recorded in each bay indicated differences in the noise climate. Single rooms were found to have less consistent patterns of noise levels which in some cases were found to be higher than those measured in multi-bed bays. Staff activity, and patient and visitor behaviour was shown to be the main reason for this. All measured levels in the patient accommodation were above those suggested by the WHO guidelines and the day / night division specified by the WHO did not appear to be realistic. Over 50% of staff in all three wards rated medical equipment alarms and the internal telephone as annoying noise events, with people talking and staff talking on the telephone rated by 30% of respondents in each ward. However, opinion on the annoyance of other noise sources was more split. High percentages of staff on Ward M4 rated both the nurse call and ward doorbell as annoying; whereas more staff in wards D8 and N3 rated visiting time, the use of mobile phones and of TV / radio. In the modular ward, Ward N3, higher percentages of respondents rated the noise from trolleys, footsteps and external noise as annoying than on the other two wards. Noise of footsteps and trolleys may be magnified on this ward by the construction of the timber floor. 62% of patients in Ward D8 found the ward to be a little noisy during the day, with patients crying out, the internal telephone ringing and visiting time cited as the most annoying events, whereas on wards N3 and M4, over 60% of patients found the wards quiet or very quiet during the day.
200
Of those questioned, 51% of patients in Ward D8; 33% of patients in Ward N3; and 57% of patients in Ward M4 felt that they were disturbed by noise at night. Opinion was split with patients crying out, people talking, medical equipment alarms and the internal telephone disturbing patients in Ward D8; the internal telephone far outranking any other source of noise disturbance in Ward M4; and the nurse call system cited as disturbing by the highest percentage of patients in Ward N3, followed by medical equipment alarms, people talking and doors banging.
9.13. Conclusions
As in Bedford Hospital, noise level measurements and questionnaire surveys have confirmed that noise is a problem in both medical and surgical wards. Staff responses indicate that they are annoyed by noise, and a significant number of patients questioned felt that they were disturbed by noise during the night, a time when they should be able to rest and recuperate. Noise levels did not appear to be related to occupancy levels, with similar levels measured in both four and six bed bays, and higher levels measured in single patient rooms than in the multi-bed bays on occasions. Much of the high level noise identified could be reduced with changes to behaviour, correct enforcement of hospital policies, simple improvements to design and maintenance of equipment. This is discussed further in Chapter 12. The following chapter investigates the use of the Maximum Likelihood Estimation method to estimate reverberation times in occupied wards using the data captured during the study.
201
Acoustic Design for Inpatient Facilities in Hospitals Blind estimation of reverberation time ___________________________________________________________________________________________________________________
10. Blind estimation of reverberation time

10.1. Introduction
Reverberation time (RT) is generally used as an indicator of the acoustic comfort of a space and is an important measurement in the field of room acoustics. There are several standard methods used to calculate the RT of a room, however these rely on generating high level noise, which would be unacceptable in an occupied hospital ward. All of the wards taking part in the main study were at capacity at all times and therefore it was not possible to make physical RT measurements in these wards. An alternative estimation technique was identified that could possibly be used with some of the noise measurement data collected in the study wards. The method was developed at Salford University (Kendrick, Cox and Li, 2007; Kendrick et al, 2011) and is known as the Maximum Likelihood Estimation (MLE) method. The technique had been successfully used to estimate a number of acoustic parameters blind (including RT), by using sounds already present in a room such as speech or music. As discussed in previous chapters, the noise measurement data collected from the wards contained numerous discrete sound files or trigger files, created when LAmax exceeded 70 dB. These files were between six and ten seconds in length and contained sounds such as speech; impulsive noises, for example, rubbish bins, doors banging and dropped objects; bed rails; and other sounds associated with general movement in a space. It was thought that these trigger files may be suitable for use with the MLE algorithms, and therefore could potentially be of use in estimating the RT20 (MLE-RT20) of the occupied wards. However, extensive validation would be required to prove that accurate estimates could be obtained. This chapter begins by describing the validation of the MLE-RT20 method using measurement data from the surveys of Bedford and Addenbrookes Hospital and from a simulated hospital environment. Following the discussion of the validation, the maximum likelihood method is used to estimate reverberation times for a number of occupied study wards.
10.2. Initial validation

The MLE-RT method works by identifying short periods of reverberant decay within a noise dataset where the dynamic range is greater than 25 dB. Data selection techniques are used to ensure correct recognition of suitable decay phases, and the mean RT is estimated from multiple values. To ensure estimates are accurate, the method relies on a large amount of suitable data. More details of the method are provided by Kendrick (2009; 2011).
202
As the MLE-RT method was initially developed for use with speech or music, it was unknown whether it would work with the trigger files collected. It was thought that if the files were of a reasonable sound quality, and enough suitable decays could be identified, then the discrete nature of the files would not necessarily prevent the method from working. Consequently, trigger files collected during noise measurements in a 4-bed bay during the pilot study were sent for initial processing. The sound level meter (SLM) had a number of settings which could be adjusted to change the format and quality of the audio file recordings. These settings included three different word lengths of 8, 16 and 24-bit; two sampling frequencies at 12 kHz and 48 kHz; and the recording gain. The word length and sampling frequencies affect the size of the file, and so with limited storage available on the meter, these had been initially set to minimum values for the pilot study. The gain setting increases the loudness of the recorded sound file, and had been set fairly high, to ensure playback on a laptop would be easily audible during analysis. Following the initial processing of the trigger files, a number of issues with the data were found. The low quality of the audio recording caused a number of dropouts, and the high gain setting caused a certain amount of clipping. Both these issues needed to be resolved before the data would work with the MLE algorithms. It was felt however, that the type of sounds recorded would be suitable, and if the audio issues were resolved, the data would probably be capable of yielding some reasonable results. The audio quality settings were subsequently changed to 24-bit sampling at a frequency of 12 kHz with a gain of 24 dB. These were considered to be the optimal settings in terms of both audio quality and file size, and would allow estimations up to 4 kHz to be computed. With confidence in the discrete data files established, the next step was to perform a simulation experiment to compare the accuracy of estimated MLE-RT values with RTs measured in a laboratory space. This is discussed in the next section.
10.3. Validation using real and simulated measurements

To test the accuracy of the estimated RT values against those measured, a scenario was developed that would allow both real time RT measurements to be made along with the recording of noise data consisting of the type of sounds that would be found in an occupied hospital ward. A new building had recently opened at London South Bank University, which was used in part for nurse training. Several clinical skills laboratories were laid out as hospital wards to provide students with clinical practice space. Each laboratory was equipped with standard hospital furniture including beds with rails, bed tables, dustbins, sinks, privacy curtains and wheeled equipment. It was thought that within this environment it would be easy to create a noise climate similar to that of an occupied ward. This would allow enough trigger files to be captured to provide data for the MLE-RT method. Real time RT measurements could also be made using an impulsive noise source and the results
203
compared to the estimates generated. Two validation experiments were carried out and are discussed in the following sections.
10.3.1. Validation 1
The clinical skills laboratory used for the first validation, shown in Figure 10.1, had a volume of 171 m . The ceiling was exposed concrete soffit; the floor was concrete with a heavy duty vinyl covering; and the walls were plasterboard. As can be seen from Figure 10.1, the room was fully furnished, and included dummy patients in the beds.
3
Figure 10.1 Clinical skills laboratory used for validation 1
RT measurements were made using thick latex balloons as an impulsive noise source. Six measurements were made; with three source and two receiver positions. Figure 10.2 shows the spatially averaged RT20 values over third octave bands from 250 Hz to 4 kHz as stipulated in BS EN ISO 3382-2 (2008). At 500 Hz and above the 95% confidence limits for each RT20 value are within 0.1 s, suggesting good accuracy at these frequencies. However, accuracy at 250 to 400 Hz is lower, with slightly more measurement variation at these frequencies.
1.0
0.9
0.8
RT20 (s)
0.7
0.6
0.5
0.4
0.3 250 315 400 500 630 800 1k 1.25 k 1.6 k 2k 2.5 k 3.15 k 4k
Frequency (Hz)
Figure 10.2 Average RT20 measurements with 95% confidence limits (Impulse Response Method)
204
A 60 minute noise measurement was made within the room during which noise was created that would be comparable to that found in an occupied hospital ward. Noise included conversation; moving of furniture and bed rails; use of rubbish bins and sinks; dropping objects; and opening and closing of doors. During this measurement period 149 trigger files (sound files where LAfmax exceeded 70 dB) were created, totalling approximately 20 minutes in length. This data was sent to the University of Salford for processing. Table 10.1 shows both the measured and estimated RT values and the difference between them. Due to the processor intensive nature of the algorithm used in the MLE-RT method, only octave band RT values are estimated. Table 10.1 Comparisons between measured and MLE-RT20 values
Frequency (Hz) 250 500 1000 2000 4000
Measured T20 (s) 0.712 0.712 0.706 0.722 0.707
MLE-RT20 (s)
0.710 0.828 0.771 0.702 0.709
Difference (s) -0.002 0.116 0.065 -0.020 0.002
It can be seen that the results show reasonable accuracy with less than 0.1 s difference in the majority of cases (differences greater than 0.1 s are highlighted in green). The value 0.1 s is of particular significance when estimating RT values because of the subjective difference limens. Kendrick (2009; 2011) discusses that in order to judge the performance of a measurement method it must be compared against the ability of the human ear to detect subtle changes in acoustic conditions. Subjective difference limens are the smallest change in a parameter value that can be detected and are determined using just noticeable differences. Bork (2000) shows that in a room with an RT value of 2 s or less, the subjective difference limen is 0.1 s, and hence any change in the RT that is less than 0.1 s would be inaudible to the listener.
10.3.2. Validation 2
Following the positive results obtained during the first validation, it was felt that further simulations in different acoustic conditions would be required to reinforce the initial outcome. Another clinical laboratory of different dimensions (153 m3) was available for use and it was decided to perform several sets of tests within this room. Apart from the room volume, the layout and finishes were identical to those of the first room used. To ensure as much data as possible was collected, the acoustic room conditions were varied. This was achieved by fully drawing the privacy curtains around the beds during one test scenario and then opening all the curtains for a second scenario. Two identical Norsonic 140 sound level meters (SLM1 205
and SLM2) were used to make noise level measurements of the simulated hospital sounds. The meters were situated on different sides of the room and hence captured slightly different levels, thus providing further data for validation purposes. Two sets of real time RT measurements were made using the same impulse response method as in the previous validation; one set with the privacy curtains open; the other with the curtains drawn. As before, six measurements were made in each case; with three source and two receiver positions. The measured results were found to be consistent above 500 Hz, with 95% confidence limits for each mean RT20 value within 0.1 s, but with some inconsistencies found at 500 Hz and below, as shown by Figure 10.3. It can be seen that as a result of drawing the curtains, RT20 values were reduced by between 0.1 s and 0.3 s.
1.0
0.9
0.8
RT20 (s)
0.7
0.6
0.5 Curtians drawn 0.4 Curtains open
0.3 250 315 400 500 630 800 1k 1.25 k 1.6 k 2k 2.5 k 3.15 k 4k
Frequency (Hz)
Figure 10.3 Average RT20 measurements with 95% confidence limits (Impulse Response Method) Two 60 minute noise measurements were made within the room on each SLM, with curtains open and curtains drawn. Again noise was created that would be comparable to that found in an occupied hospital ward. Table 10.2 shows the number of trigger files recorded during each scenario by each SLM. Table 10.2 Numbers of triggers recorded during the simulations Number of trigger files recorded SLM 1 SLM 2 Curtains open 188 233 Curtains drawn 164 196
The trigger files recorded were used to provide MLE-RT20 estimates for the four scenarios. The estimates were compared with the actual measured values and the results can be seen in Tables 10.3 to 10.6. Differences greater than 0.1 s are highlighted in green. 206
Table 10.3 SLM 1 with curtains open

SLM 1 - Curtains Open Measured RT20 (s) Estimated RT20 (s) 0.775 0.812 0.861 0.910 0.921 0.752 0.949 1.004 0.904 0.963
Table 10.4 SLM 1 with curtains drawn

SLM 1 - Curtains Drawn Measured RT20 (s) Estimated RT20 (s) 0.684 0.704 0.645 0.646 0.659 0.660 0.628 0.640 0.734 0.619
Frequency (Hz) 250 500 1000 2000 4000
Difference (s) -0.023 0.137 0.143 -0.006 0.042
Frequency (Hz) 250 500 1000 2000 4000
Difference (s) -0.024 -0.076 -0.005 0.088 -0.040
Table 10.5 SLM 2 with curtains open

SLM 2 - Curtains Open Measured RT20 (s) Estimated RT20 (s) 0.775 0.812 0.861 0.910 0.921 0.794 0.928 0.940 1.067 1.014
Table 10.6 SLM 2 with curtains drawn

SLM 2 - Curtains Drawn Measured RT20 (s) Estimated RT20 (s) 0.684 0.704 0.645 0.646 0.659 0.652 0.778 0.672 0.666 0.692
Frequency (Hz) 250 500 1000 2000 4000
Difference (s) 0.019 0.116 0.079 0.156 0.093
Frequency (Hz) 250 500 1000 2000 4000
Difference (s) -0.032 0.074 0.027 0.020 0.033
The results show good accuracy with the curtains drawn, but slightly less so with the curtains open, with two octave frequency bands showing a difference slightly above 0.1 s in each case, which is just outside the difference limen for RT. It is thought that this is primarily due to the number of suitable decay phases with a sufficient signal to noise ratio available for analysis at these frequencies. Figure 10.4 further illustrates the accuracy of the MLE-RT20 estimates, by plotting the estimated values against the measured values. The 0.1 s difference limens are represented by dashed lines.
1.4 1.3 1.2 1.1 1.0 0.9
MLE-RT20 (s)
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Measured T20 (s)

SLM1 - Curtains open SLM1 - Curtains closed SLM 2 - Curtains open SLM2 - Curtains closed
Figure 10.4 Accuracy of RT20 estimations in relation to actual measured values
207
10.3.3. Validation study conclusions

The validation study demonstrated that with the change of audio settings, the trigger file data could be successfully analysed using the MLE-RT20 method. The accuracy of the estimates was reasonable overall, but some small inconsistencies were found for the different room conditions. This was primarily due to insufficient data with suitable regions of decay. Given that the data captured in the simulation was captured for a relatively short time period (60 minutes) and that measurements made during the main study were over a seven day period, it was considered likely that sufficient data would be available to provide reliable estimates of reverberation times in occupied hospital wards.
10.4. Estimation of RT in occupied hospital wards

The validation study described in the previous section yielded on average 1.4 samples of data with suitable free reverberant decay per minute. Examination of the trigger files from the occupied wards showed the density of suitable data was approximately eight times less than in the validation study. Previous testing of the MLE-RT20 method had established a minimum amount of suitable data needed to produce accurate estimates, and when this was applied to the hospital data it was calculated that at least forty hours of measurement data would be required to provide enough suitable decays. As discussed previously, the measurement data collected from the occupied hospital wards was over a period of approximately seven days, and as such should provide ample data to produce an accurate estimate of RT. Table 10.7 shows the data available for use with the MLE-RT20 method, which consisted of data from 17 different spaces. Data was not available from all wards due to poor quality audio files. Table 10.7 Locations with data available for MLE-RT20 estimation
Hospital
Addenbrookes Addenbrookes Bedford Bedford
Ward
D8 N3 Surgical Medical
Areas for which MLE-RT data is available

2 x 3-bed bays, 1 x 4-bed bay, 1 x 7-bed bay, 1 x 12-bed bay 3 x 4-bed bays, 2 x single rooms 2 x 4-bed bays, 1 x 6-bed bay, 2 x single rooms, nurse station Pre and post ceiling tile change in 4-bed bay
10.4.1. Methodology
For each room, the trigger files captured over the entire measurement period (generally seven days), were segmented into two groups: from 06.00 to 18.00 (day); and from 18.00 to 06.00 (night). The day and night split chosen was not the same as was used in rest of the study, as too little data would exist for the period 23.00 to 07.00 for estimation purposes. Data was grouped in this way for two reasons: (i) to reduce the possibility of compromising day time estimates with potentially less accurate night 208
time estimates (due to the lower amount of suitable data available at night); (ii) to allow comparisons to be made between day and night time estimates (given that night time data was found to be sufficiently accurate). This comparison could yield interesting information regarding the effects of ward conditions on the MLE-RT20 estimates, mostly in relation to occupancy levels (no visitors and less clinical and domestic activity at night and hence less acoustic absorbency). Estimates were computed for octave frequency bands 250 to 4000 Hz, for the two groups of data. Lower frequencies were discarded, as previous experience of this method with other data sets showed inaccuracies at frequencies less than 250 Hz. For each data group, initial estimates were calculated from suitable decays captured during four hour windows of data. An example of this can be seen in Table 10.8 for the day time data. The final MLERT20 estimate for each day and night group was computed by calculating the mean of all the estimates over the measurement interval. Table 10.8 Day time data shown in 4 hour windows; overall mean estimate with 95% confidence intervals
Frequency (Hz) 250 500 1000 2000 4000 Day 1 06.00 - 18.00 Day 2 06.00 - 18.00 Day 3 06.00 - 18.00 06.00 - 10.00 10.00 - 14.00 14.00 - 18.00 06.00 - 10.00 10.00 - 14.00 14.00 - 18.00 06.00 - 10.00 10.00 - 14.00 14.00 - 18.00 0.500 0.514 0.541 0.434 0.436 0.520 0.720 0.604 0.693 0.530 0.637 0.500 0.611 0.486 0.416 0.332 0.483 0.494 0.571 0.605 0.365 0.526 0.512 0.652 0.451 0.483 0.461 0.596 0.621 0.510 0.445 0.406 0.455 0.590 0.586 0.591 0.509 0.663 0.649 0.573 0.585 0.568 0.614 0.697 0.621 Mean RT 95% CL (7 days) 0.057 0.047 0.034 0.041 0.037
0.539 0.559 0.579 0.632 0.554
It has been shown (Kendrick 2009), that by calculating the standard error of these estimates, and by only accepting measurements where the 95% confidence limits are within 0.1 s, very good accuracy is obtained.
10.4.2. MLE-RT20 estimates from day time data

This section graphically presents MLE-RT20 estimates from three different study wards; examines the accuracy of the estimates; and discusses the effects of the different levels of acoustic absorbency on the wards. Figure 10.5 shows the MLE-RT20 day time estimates for five multi-bed bays on Ward D8 at Addenbrookes Hospital. It can be seen that the estimates are between 0.4 s and 0.57 s at 1 kHz. The majority of estimates have low 95% confidence intervals of around 0.05s, suggesting that the estimated RTs are accurate. All 95% confidence intervals calculated for the estimates are within the 0.1 s difference limen. This particular ward has little in the way of acoustic absorbency at ceiling level. As explained in Chapter 9, most ceilings in this ward consist of metal pan tiles which work in conjunction with the heating system, by radiating heat from the hot water pipes running above them. These tiles are perforated, and have a layer of insulation covering the water pipes which may provide a level of acoustic absorbency at some frequencies. The 4-bed bay which is situated behind the nurse station 209
has a suspended ceiling grid with solid plaster tiles. It can be seen that it is the larger volume multibed bays with more occupants and hence more absorption, which have the lowest MLE-RT20 estimates, as would be expected.
0.8 0.7
0.6 0.5 0.4 0.3 0.2 0.1 0

3-Bed Bay A 3-Bed Bay B 4-Bed Bay 7-Bed Bay 12-Bed Bay
250
500
1000
2000
4000
Octave band (Hz)
Figure 10.5
MLE-RT20 estimates for five multi-bed bays in Ward D8, Addenbrookes Hospital (day time data) with 95% confidence limits
Figure 10.6 shows MLE-RT20 estimates for six locations in Ward N3 at Addenbrookes Hospital. It can be clearly seen, that the estimates for 4-bed bay B and single room J are higher than for the other locations and also have larger confidence intervals, particularly at 250 Hz, where they are greater than 0.1 s. As this data falls outside the stipulated 0.1 s confidence limits, it must be assumed to be inaccurate and therefore should be ignored. However, the majority of estimates have low 95% confidence intervals with a mean of 0.04s, which suggests that the estimated values are accurate.
0.8 0.7
0.6 0.5 0.4 0.3 0.2 0.1 0

4-Bed Bay A 4-Bed Bay B 4-Bed Bay C Single Room J Single Room K
250
500
1000
2000
4000
Octave band (Hz)
Figure 10.6
MLE-RT20 estimates for six locations in Ward N3, Addenbrookes Hospital (day time data) with 95% confidence limits 210
The MLE-RT20 estimates at 1 kHz show much lower RTs than in Ward D8, with all estimated values of around 0.25 to 0.3 s. All these areas have suspended ceilings with good quality acoustic ceiling tiles. Figure 10.7 shows MLE-RT20 estimates for seven locations in the surgical ward at Bedford Hospital. The estimate for 4-bed bay 1 shows confidence intervals which are greater than 0.1 s at 250 Hz and therefore this data should be ignored. However, this is the only estimate for the ward where the 0.1 s confidence limits were exceeded, with most confidence limits generally low, with values less than 0.04 s. The MLE-RT20 estimates at 1 kHz for the nurse station and 4-bed bay 1 show the lowest estimated values of around 0.25 to 0.3 s. These areas both have a suspended ceiling with acoustic ceiling tiles and it can be seen that the values are similar to those in Ward N3 at Addenbrookes Hospital (see Figure 10.6), which also had acoustic ceiling tiles. The single rooms and 6-bed bay 3 are similar with a low value of around 0.35 s. These rooms have either solid plaster ceilings (6-bed bay 3 and single room 1) or a non acoustic suspended ceiling (single room 3). 4-bed bay 4 has noticeably higher estimates at all frequencies, with an MLE-RT20 estimate of 0.55 s at 1 kHz. This bay has been refurbished more recently and has reflective, plaster ceiling tiles throughout.
0.9 0.8
4-Bed Bay 1 4-Bed Bay 4 6-Bed Bay 3 Single Room 1 Single Room 3 Nurse Station
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 250 500 1000 2000 Octave band (Hz) 4000
Figure 10.7
MLE-RT20 estimates for seven locations in the surgical ward, Bedford Hospital (day time data) with 95% confidence limits
It appears that the results are consistent with the amount of absorbency provided by the ceiling, with longer RTs estimated in those rooms with solid plaster ceilings or those with reflective ceiling tiles.
211
10.5. Comparison of day and night time MLE-RT20 estimates

As discussed in Section 10.4.1, the data was segmented into day and night groups to allow for comparisons to be made between times when acoustic conditions on the ward could potentially be quite different. Day times are busy with visitors; there is a great deal of clinical and domestic activity; and privacy curtains are often drawn around beds. All these events provide additional absorbency which could affect the estimated MLE-RT20 values. Night time is a relatively quiet time, with minimal disturbance by clinical staff, much less ward activity and hence lower additional absorbency. It could therefore be assumed that the MLE-RT20 estimates would be longer at night than those during the day, but further analysis of the results is required to establish this. As expected, due to lower amounts of suitable data, the estimates calculated from the night time data were much more variable. Many of the 95% confidence limits were found to be in excess of 0.1 s, therefore the estimates were unusable and no comparisons could be made. However, there were three instances where enough night time data with suitable decay phases existed. Figures 10.8, 10.9 and 10.10 show the comparisons between the day and night MLE-RT20 estimates in three different patient bays in Addenbrookes and Bedford Hospitals.
0.7 0.6
Day time estimate 7-Bed Bay Night time estimate 7-Bed Bay
0.5 0.4 0.3 0.2 0.1 0 250 500 1000 Octave Band (Hz) 2000 4000
Figure 10.8 Comparison of day and night time estimates, 7-bed bay, Ward D8, Addenbrookes Hospital
212
0.7 0.6 Estimated MLE-RT20 (s) 0.5 0.4 0.3 0.2 0.1 0 250 500 1000 Octave Band (Hz)
Day time estimate 12-Bed Bay Night time estimate 12-Bed Bay
2000
4000
Figure 10.9 Comparison of day and night time estimates, 12-bed bay, Ward D8, Addenbrookes Hospital
0.6
Day time estimate 4-bed bay
0.5
Night time estimate 4-bed bay
0.4 0.3 0.2 0.1 0 250 500 1000 2000 4000
Octave Band (Hz)

Figure 10.10 Comparison of day and night time estimates, 4-bed bay, medical ward, Bedford Hospital
It can be seen from each figure, that the night time estimates are slightly higher than the day time, as would be expected. However, the differences between the day and night time estimates for the seven and 12-bed bays are very small, as little as 0.01 s at some frequencies. In the four bed bay, where more data was available (12 days rather than seven), the MLE-RT20 estimate can be seen to be at
213
least 0.05 s lower during the day in each octave band (with slightly larger differences found at lower frequencies). During a 24 hour period the acoustic conditions on a hospital ward are constantly changing due to occupancy levels, opening and closing of privacy curtains, opening windows and many other aspects. There is no easy way to record all of the conditions in the ward at a given time. It can therefore be said that MLE-RT20 estimates are representative of snapshots of the acoustic conditions on the ward when they are at their least reverberant. To explain this further, the MLE-RT20 method works by searching the dataset for the fastest decaying region over a given four hour time window and hence each estimate will be consistent with the highest occupancy / highest absorption on the ward. Where night time estimates were available for comparison, they were found not to differ greatly from the day time estimates, which indicates that the acoustic conditions on the wards were fairly stable at all times.
10.6. Summary
The application of the MLE-RT20 method was investigated for use with discrete sound or trigger files collected in occupied hospital wards. After rectifying initial audio quality issues, a number of validation studies were carried out in simulated hospital wards. In each study, high trigger files (generated when LAMAX exceeded 70 dB) were collected over a 60 minute period and a reverberation time estimate (MLE-RT20) was calculated and compared with actual RT20 measurements made in the rooms using an Impulse Response Method. Initial findings from the validation studies were positive, but highlighted the need for sufficient data with uninterrupted reverberant decays, meaning significantly longer recordings would be required. The MLE-RT20 method was applied to data from noise measurements made in a number of ward locations, at Bedford and Addenbrookes Hospitals. Each measurement interval was at least seven days in length and the trigger files captured were segmented into day and night time groups. For each data group, a day and night mean estimate was calculated for the entire measurement period. For the day time MLE-RT20 estimations, in the octave bands 500 to 4000 Hz, the worst case 95% confidence limits indicated a maximum error of 0.08 s and as such the MLE-RT20 estimation method can be said to demonstrate similar accuracy to standard measurement methods such as the Impulse Response Method. Night time MLE-RT20 estimations were found to be less accurate due to the availability of suitable data, but it is felt that this could be further improved by longer measurement periods. Comparisons of the night and day estimates provide an indication of the variation of acoustic conditions on the wards, and were found not to differ greatly. This suggests that the acoustic conditions on the wards were fairly stable at all times despite increased activity and occupancy levels.
214
10.7. Conclusions
An estimation method which provides information regarding reverberation times in occupied spaces has been trialled and has been shown to demonstrate similar accuracy to standard measurement methods such as the Impulse Response Method. The data provided by this method can usually only be generated using complex and time consuming modelling techniques, and as such the MLE-RT20 method could be used to provide reverberation time estimates in occupied areas where real time measurements are not practical or possible. The following chapter presents some overall results from the main study; looking for general trends and relationships within the objective and subjective data collected.
215
Acoustic Design for Inpatient Facilities in Hospitals Analysis of objective and subjective data ___________________________________________________________________________________________________________________
11. Analysis of objective and subjective data

11.1. Introduction
Analysis of the noise level data and questionnaire responses for each of the main study wards has been carried out and presented in Chapters 7 and 9. This chapter aims to explore the overall results; looking for general trends and relationships within the objective and subjective data collected from the main study. Due to the differences in the care group, results from the pilot study are not considered in this chapter. The chapter begins by summarising the objective and subjective data collected in the study, as shown in Table 11.1, which lists factors such as building age; ceiling types; ward configuration; measurement durations and numbers of questionnaire responses. The chapter continues by examining the objective data to investigate factors which might affect the acoustic environment of a hospital ward. The physical aspects of the room design are considered and relationships between noise levels and factors such as bay size are explored. Patient perceptions are studied, with the effects of gender, age, hearing, length of stay and bed position on noise annoyance and disturbance investigated. Aspects of privacy and speech intelligibility are also examined. The final part of this chapter considers staff perceptions of noise annoyance and interference with work, and examines factors such as gender, age, and length of service.
11.2. Factors affecting noise levels

11.2.1. Effect of bay size
There is a general assumption among the acoustic, medical and architectural professions that the larger the size of bay (that is the higher the number of patients in a bay), the greater the potential noise level. In their 2004 paper on the role of the physical environment in the hospital of the 21st Century, Ulrich et al (2004) state that a clear-cut finding in the literature is that noise levels are much lower in single-bed than multi-bed rooms. The data was therefore examined to see if this was found to be the case in the current study. Figure 11.1 shows the day time noise levels (LAeq,16hr) for each bay size, with the different study wards represented by different bar colours. Interestingly, many of the noise levels measured in the single rooms and three bed bays can be seen to be as high as those measured in the seven and twelve bed bays. In fact the highest measured day time level is 60.6 dB LAeq,16hr which was measured in a single room.
216
Table 11.1 Summary of the objective and subjective data collected during the study
Details Hospital Building Type Age Ward Name Ward Type Ward capacity Ward Config GOSH PFI < 10 years Sky Surgical 18 2 x nurse stations; 3 x 4-bed bays; 6 x single rooms Addenbrooke's PFI < 10 years M4 Surgical 32 2 x nurse stations; 5 x 4-bed bays; 12 x single rooms Addenbrooke's 1970's tower 50 years D8 Trauma / Orthopaedics 35 1 x nurse station; 1 x 12-bed bays; 1 x 7-bed bay; 1 x 4-bed bay; 3 x 3-bed bays; 3 x single rooms Metal pan with some insulation Addenbrooke's Modular build < 10 years N3 Respiratory care 25 1 x nurse stations; 4 x 4-bed bays; 9 x single rooms Bedford Early 1980's build 30 years Elizabeth Medical 30 1 x nurse station; 3 x 6-bed bays; 2 x 4-bed bays; 4 x single rooms Bedford Early 1980's build 30 years Howard Surgical 26 1 x nurse station; 1 x 6 bed bays; 4 x 4 bed bays; 4 x single rooms
Acoustic in patient acc / plaster in common areas Measurement: Location and length Multi-bed bays 4 bed bay A - 10 days; 4 bed bay B - 5 days;
Ceiling tile type
Acoustic throughout
Acoustic throughout
Acoustic throughout
Acoustic throughout
4 bed bay A - 6 days; 4 bed bay B - 8 days
Single rooms
Single room A - 5 days; Single room B - 5 days
Single room A - 7 days; Single room B - 9 days
Nurse Stations
Nurse station 1 - 10 days; Nurse station 1 5 days; Nurse station 2 - 5 days Nurse station 2 - 7 days
4-bed bay 1 (refurbished) -12 days (over 2 non consecutive weeks); 4-bed bay 2 - 7 days; 6-bed bay 3 - 6 days; 6-bed bay 4 - 8 days Single room J - 7 days; Single room A Single room K - 9 days (refurbished) - 5.5 days; Single room B - 7 days Nurse station - 6 days Nurse station - 5 days Nurse station - 7 days; Ward clerk's desk area 7 days 3 bed bay A - 7 days; 3 bed bay B - 7 days; 4 bed bay - 8 days; 7 bed bay - 6 days; 12 bed bay - 2 non consecutive 7 day periods 4-bed bay A - 7 days; 4-bed bay B - 8 days; 4-bed bay C - 7 days
4-bed bay 1 - 6 days; 6-bed bay 3 - 8 days; 4-bed bay 4 - 7 days
Single room A (refurbished) - 7 days; Single room B - 7 days Nurse station - 7 days
Questionnaires: Numbers of responses Staff Quest reponse Patient Quest response 12 31 10 14 20 47 11 13 18 40 7 42
70 60 50
LAeq,16hr (dB)
40 30 20 10 0 1 3 4 6 7 12
Bay size (number of patients)
Ward M4, Addenbrookes, Ward N3, Addenbrookes, Ward D8, Addenbrookes Medical ward, Bedford, Surgical ward, Bedford
Figure 11.1 Average day time levels by bay size for all main study wards The relationship between the average day time noise level for each bay and bay size was investigated, but this was not statistically significant (=.038, ns). Figure 11.2 shows average night time noise levels (LAeq,8hr) in each bay, grouped according to bay size with the different study wards represented by different bar colours. It can be seen that some of the highest average night time noise levels were measured in single rooms and three bed bays.
60 50 40 30 20 10 0 1 3 4 6 7 12
LAeq,8hr (dB)
Bay size (number of patients)
Ward M4, Addenbrookes, Ward N3, Addenbrookes, Ward D8, Addenbrookes Medical ward, Bedford, Surgical ward, Bedford
Figure 11.2 Average night time levels by bay size for all main study wards 218
As with day time noise levels, the relationship between average night time noise levels for each bay and bay size was not statistically significant (= .014, ns).
11.2.2. Surgical and medical wards

The study has enabled the effect of ward type on noise levels to be examined. At Bedford Hospital, two inpatient wards of similar layout were chosen in the main five storey ward block. Both wards were subject to the same general hospital routines and regulations, but one ward was a surgical ward, the other medical. This provided an opportunity for a comparison to be drawn between the type of care provided and any differences this may have on the noise environment and on patient and staff perceptions. More detailed discussion can be found in Chapter 7. Overall noise levels at the nurse stations were found to be very similar between the wards, but the content of the noise varied, with more frequent use of the nurse call and higher levels of staff conversation captured at the surgical ward nurse station. Overall noise levels in the patient accommodation were also found to be very similar between the wards, except for the bay directly opposite the main nurse station in the medical ward, where instances of patients crying out and increased clinical activities increased noise levels. This was specifically related to the numbers of acutely ill elderly patients in this bay, many of whom were suffering from confusion or dementia. Surgical and medical wards are very different in the type of services they provide. Surgical wards are very busy with constant admissions for day or even half day procedures. Operations are booked in advance and efficiency and timing are imperative. Medical wards are slower paced, with fewer admissions and longer patient stays. These differences may in part explain the differences in perceptions which are discussed in the following paragraphs. Response to the questionnaire survey by staff in the surgical ward was poor, with only seven completed questionnaires. It is possible that only those staff who felt strongly about noise made the time to express their opinion. If this is the case, the results may be not wholly representative. However, it is still worth comparing the staff perceptions between the wards. Surgical ward staff rated visiting time, internal telephone, meal times, the nurse call, and trolleys as interfering with their job much more highly than the staff on the medical ward. As this ward is more fast paced, staff may find the presence of visitors, the constant ringing of the ward telephones and the serving of meals impacts their efficiency further and hence causes more interference. The increased use of the nurse call is reinforced by the results of the objective survey, but it is unknown why this is used more on this ward. Trolleys are known to be used more extensively in the surgical ward with patients constantly being wheeled through the ward to go to and from surgery and X-ray. The major difference between patient perceptions was found to be the disturbance caused by patients crying out at night in the medical ward. This was mainly due to the numbers of elderly patients on the 219
ward, many of whom were suffering from confusion or dementia. As with the staff, patients also rated trolleys and the nurse call more highly on the surgical ward. Other differences in perceptions were due to maintenance issues on the ward or the lack of enforcement of hospital policy, but as such were not an indication of the impact of the differences of care provided on the noise environment. The three study wards at Addenbrookes Hospital were also a combination of surgical and medical wards. However, these wards were not considered suitable for comparison due to the mixture of different building design variables and care groups.
11.2.3. Impact of high level noise events on overall noise levels

Throughout this study, average day and night noise levels have been calculated for each bay, with average hourly noise levels plotted over 24 hour intervals. Levels presented in these ways do provide a general indication of the daily patterns of noise and overall levels measured in each bay, but fail to illustrate the continuously fluctuating nature of the noise or provide any understanding of its content. To help build up a more detailed picture of the noise climate, high level noise events (over 70 dB LAmax) have been investigated and are reported in Chapters 7 and 9. This section examines the impact of the numbers of high level noise events on the overall noise levels. Figure 11.3 shows the average day time noise levels for each bay measured plotted against the average number of day time high level noise events. It can be seen that the data points are a good fit to the trend line suggesting a strong relationship between the average day time noise levels and average numbers of high level noise events, confirmed by the statistically significant correlation coefficient of =.924 (p<0.01). The gradient of the trend line shows that for every 100 high level noise events occurring during the day time, there is an increase in the average LAeq,16hr of 1 dB.
70 60
LAeq,16hr (dB)
50 y = 0.0101x + 49.88 40 30 20 10 0 0 200 400 600 800 1000 1200
Average number of high level noise events per day
Figure 11.3 Average day time noise levels and average number of day time high level noise events for each bay 220
Night time noise levels and high level noise events have been also examined to see if a similar relationship exists. Figure 11.4 shows the average night time noise levels for each bay plotted against the average number of night time high level noise events.
60 50
LAeq,8hr (dB)
40 30 20 10 0 0 50 100
y = 0.0555x + 43.764
150
200
250
Average number of high level noise events per night
Figure 11.4 Average night time noise levels and average number of night time high level noise events for each bay Although there is again a statistically significant relationship (=.737, p<0.01) between noise levels and the numbers of high level noise events, it is weaker for the night time than for the day time noise. However, the gradient of the trend line is steeper than for the day time data and shows that for every 21 high level noise events occurring during the night time, there is an increase in the average LAeq,8hr of 1 dB. This indicates that the effect of high level noise events on the overall noise levels (LAeq,8hr) is greater at night than during the day and may suggest that these high level noise events cause greater disturbance during the night. This is confirmed by the subjective responses of patients, where 52% were disturbed by noise at night in contrast to 21% during the day (see Section 11.3.1).
11.2.4. Noise levels and reverberation times

As discussed in Chapter 10, reverberation times (RT) were estimated using the maximum likelihood estimation method (MLE-RT20) for those bays with suitable data available. The relationship between the daytime LAeq, 16hr and day time MLE-RT20 estimates has been investigated and is shown in Figure 11.5. There is a statistically significant correlation between the LAeq, 16hr and the RT, (=.521, p<0.05), with lower noise levels related to lower reverberation times as would be expected.
The gradient of the trend line shown in Figure 11.5 provides a relationship between the average day time LAeq,16hr and RT, showing that for every 0.1 s decrease in the RT, there is a decrease in the average day time noise level of 1.2 dB LAeq,16hr. This relationship is close to that found in the ceiling intervention study discussed in Chapter 8, where the addition of an acoustically absorbent ceiling was 221
found to decrease both RTs and overall noise levels. In this case, a 0.1 s decrease in RT corresponded to a 1.8 dB reduction in noise level.
70 60
LAeq,16hr (dB)
50 40 30 20 10 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 y = 12.471x + 50.15
MLE-RT20(secs)
Figure 11.5 Average day time noise levels and estimated reverberation times in each bay
11.3. Factors affecting patient perceptions of noise

Sources of noise annoyance and disturbance to patients were examined in detail in Chapters 7 and 9 in relation to each study ward. These sources were found to be ward specific and it would be inappropriate to look for general trends in this data. This section aims to look at more general relationships including whether gender, age, length of stay or bed position exacerbate feelings of noise annoyance. Privacy, speech intelligibility and the effects of hearing impairment are also examined.
11.3.1. Overall
Figure 11.6 shows the perception of day and night time noise for all 154 respondents. It can be seen that the majority of respondents perceive the wards to be either quiet or a little noisy, both during the day and at night. The more intense perceptions of very noisy or extremely noisy are chosen by very few respondents.
222
80 70 60
Day time perception Night time perception
Number of respondents
50 40 30 20 10 0 Very quiet Quiet A little noisy Noisy Extremely noisy
Patient percepption of the noise climate
Figure 11.6 Overall patient perception of the noise climate
Figure 11.7 shows the percentages of all 154 respondents who indicated that they were annoyed by day time noise and disturbed by noise at night. A relatively low percentage (21%) of patients felt annoyed by noise during the day, but over half those patients questioned (52%) were disturbed by noise at night.
60 Day time
Percentage of patients (%)
50 40 30 20 10 0
Night time
Annoyance
Disturbance
Figure 11.7 Overall percentages of patient annoyed / disturbed by noise
Patients were also asked if they ever found it too quiet in the ward. Out of 154 respondents, only 9 patients said this was the case. Room data was only available for six of these respondents, with two in a single room and four in either four or six bed bays.
223
11.3.2. Patient gender

Figure 11.8 shows the mean perceptions of day and night time noise by male and female patients on a scale of 1 to 5, where 1 indicates very quiet and 5 indicates extremely noisy. It can be seen that the average perception of day time noise is identical for both male and female patients (mean value of 2.5), however, male patients appear to perceive the wards to be noisier at night, whereas female patients perceive the levels to be similar to those during the day. It should be noted that apart from the average male perception of 3.5 at night (slightly noisy to noisy), the average ratings are around 2.5 (quiet to slightly noisy).
5 Day time Night time
Mean perception of noise level (1=very quiet / 5= extremely noisy)
1 Male (n=92) Patient gender Female (n=62)
Figure 11.8 Mean patient perception rating of noise by gender
Figure 11.9 shows the percentages of patients by gender that are annoyed or disturbed by day and night time noise. It can be clearly seen that similar percentages of both male and female patients are annoyed by noise during the day, which agrees with the perception of the noise climate during the day as shown in Figure 11.8. However, it is interesting to note that during the night, slightly more women (55%) than men (51%) are disturbed by noise. This is in contrast to the night time perception, as shown in Figure 11.8, which shows that women perceive the ward to be quieter at night than men.
100 Day time annoyance
Percentages of patients annoyed / disturbed (%)
Night time disturbance 80
60
40
20
0 Male (n=92) Patient gender Female (n=62)
Figure 11.9 Percentages of patients annoyed / disturbed by noise by gender 224
11.3.3. Age
This section examines the effects of age on patient perceptions of noise, and on annoyance and disturbance. Figure 11.10 shows that perceptions of day time noise are found to be very consistent across all age ranges, with mean values between 2.4 and 2.6, suggesting that patients of all ages gauge the noise climate to be in the range quiet to slightly noisy during the day. However, the perception of noise at night appears to be more variable, with two age groups perceiving slightly noisier conditions: the under 20s (mean value 2.8); and those in the 41-50 age group (mean value of 3.1). Patients in the age groups 31-40, 51-60 and 60+ appeared to perceive the conditions similarly, with those in the 20-30 age bracket perceiving the wards at night to be the quietest (mean value of 1.9). It can also be seen that the perception of noise on the ward during the night is generally lower than it is during the day, except in age groups under 20 and 41-50 years.
(n=19)
3
(n=5)
(n=8) (n=16)
(n=22) (n=84)
1 <20 20-30 31-40 41-50 51-60 60+ Age group (years)
Figure 11.10 Mean rating of patient perceptions of day and night noise and age
Noise annoyance / disturbance and age group is shown in Figure 11.11. It can be clearly seen that patients in the age groups 31-40 and above are more disturbed by noise at night than during the day; which is in contrast to many of the perceptions shown in Figure 11.10. Patients in the 41-50 age group appear to be more disturbed by night time noise than those in any other age group (74%), followed by the 51-60 age group (61%). This could be due to independent lives led by the individuals in these groups and levels of control they are used to having within their lives.
225
Percentages of patients annoyed / disturbed (%)
80
(n=19) (n=5)
60
(n=22) (n=84) (n=16)
40
20
(n=8)
0 <20 20-30 31-40 41-50 51-60 60+ Age group (years)
Figure 11.11 Percentages of patients annoyed / disturbed and age
11.3.4. Hearing impairment

26% of respondents indicated that they suffered from a hearing impairment of some kind. This could include many different conditions such as hearing loss; increased sensitivity at particular frequencies and tinnitus. Figure 11.12 shows the breakdown of hearing impairment by patient age. It can be seen that out of those suffering from some type of hearing impairment, 75% are over years 60 years of age. This may account for the lower perceptions of the noise climate by this age group (shown in Figure 11.10)
80
70
Percentage of hearing impaired (%)
60
50
40
30
20
10
0 <20 20-30 31-40 41-50 51-60 60+
Age group (years)
Figure 11.12 Percentage of hearing impaired by age group 226
To see if hearing impaired patients were any less annoyed or disturbed by noise, percentages of annoyance and disturbance are plotted against hearing impaired / non hearing impaired patients, as shown in Figure 11.13. In terms of annoyance 7% fewer hearing impaired patients indicated they were annoyed by day time noise than the non hearing impaired, with 3% fewer patients indicating that they were disturbed at night.
100 Day time annoyance
Percentage of patients annoyed / disturbed (%)
60
40
20
0 Hearing impaired No impairment
Figure 11.13 Percentage of patients annoyed / disturbed with hearing impairment
Speech intelligibility and perception of speech privacy were also examined for differences between the hearing impaired patients and non hearing impaired. In terms of speech intelligibility, 35% of hearing impaired patients felt they could not always hear when staff spoke to them, in contrast to 82% of patients without any impairment. Little difference was found in terms of conversational privacy, with 69% of patients with a hearing impairment feeling that they could carry out a private conversation on the ward, and 65% of those without a hearing impairment feeling the same.
11.3.5. Length of stay

Length of patient stay has been investigated to see whether there is a difference in perceptions of noise and annoyance and disturbance between short term and longer term patients. Figure 11.14 shows mean perceptions of day and night time noise for patients who have been in hospital for less than 1 week, between 1 and 3 weeks and for more than three weeks. Responses are very consistent for patients who have been on the ward for up to three weeks. Longer term patients, however, appear to perceive day time noise levels very slightly higher and night time noise levels slightly lower. Perhaps this in an indication that longer term patients manage to acclimatise to the noise of the ward at night. However, it should be noted that the sample size for patients who had been on the ward for over three weeks was comparatively small (n=14).
227

(n = 14) (n = 96) (n = 44)
1 < 1 week 1 - 3 weeks Length of patient stay > 3 weeks
Figure 11.14 Mean rating of patient perceptions of day and night noise and length of stay
Figure 11.15 shows the percentages of patients annoyed / disturbed by noise in relation to length of stay. It can be seen that 20% of patients who had been on the ward for less than one week, and 19% of patients whose stay was one to three weeks indicated they were annoyed by day time noise. However, this percentage more than doubled for longer term patients, with 43% annoyed by noise during the day. This trend was reversed for night time disturbance, with only 36% percent of long term patients expressing night time disturbance in comparison to those patients who had been on the ward for less time. Again, as with the noise perception, this response suggests a certain amount of
acclimatisation to the night time noise climate by longer term patients, but more sensitivity to noise during the day.
100 Day time annoyance Percentage of patients annoyed / disturbed (%) Night time disturbance 80
60
40
20
0 < 1 week 1 - 3 weeks Length of patient stay > 3 weeks
Figure 11.15 Percentages of patients annoyed / disturbed and length of stay 228
11.3.6. Bed position

In general beds are only specifically allocated to either infectious patients who require barrier nursing (these patients may be placed in a single room); on the basis of gender (if no same sex accommodation is available patients may be placed in a single room); or if the patient is very unwell and requires observation (the patient may be placed in the bay nearest to the nurse station). This section investigates whether there are any relationships between noise annoyance / disturbance and bed position, with positions defined as next to the ward corridor, in the centre of the bay, by the window, or in a single patient room. Figure 11.16 shows the percentages of patients annoyed / disturbed in each type of bed location. It can be seen that surprisingly, the highest percentages of patients reporting daytime annoyance and night time disturbance are those staying in single rooms, with 50% and 58% reporting annoyance and disturbance respectively. Percentages of patients in multi-bed bays reporting daytime annoyance are much lower, between 17% and 24%, with those by the window reporting slightly lower levels of annoyance. Percentages of patients reporting night time disturbance are similar in all cases except for those patients in beds by the window, where 10% fewer respondents report disturbance. It is interesting that the lowest levels of annoyance / disturbance are reported by patients in beds located by windows. This is a similar finding to that by Yildirim et al (2007), who observed that for office workers, being positioned by a window compensated for some of the negative perceptions of open plan offices, such as low levels of visual or acoustic privacy.
100 Daytime annoyance
Percentage of patients annoyed / disturbed (%)
60
(n=9) (n=54)
(n=18)
(n=12)
40
20
0 Corridor Window Middle Single room Bed position on ward
Figure 11.16 Percentages of patients annoyed / disturbed and bed position 229
11.3.7. Speech intelligibility and privacy

Speech intelligibility on the wards is extremely important to enable clinical staff and patients to be able to discuss issues surrounding the patients condition and their treatment without the loss of important information. Speech intelligibility was not measured objectively during this study, but was gauged in part by asking patients if they could always hear clearly when the doctors and nurses spoke to them, which overall 69% of respondents said that they could. The relationship between speech intelligibility and the size of the bay was investigated by averaging the number of people who said they could always hear clearly. No relationship between was found, with a low correlation coefficient (=.333, ns). The perceptions of speech intelligibility among hearing impaired patients was also investigated and discussed in Section 11.3.4. In recent times there has been much publicity surrounding poor conversational privacy on hospital wards. To investigate this, patients were asked if they felt able to hold a private conversation with clinicians at their bedside. The responses in relation to the bay size are shown in Figure 11.17. It can be seen that 100% of patients in single rooms felt that they could hold private conversations. This is unsurprising as these patients have the option to close the door to their room if necessary. Between 60% and 68% of patients on wards with four or more beds felt that they could hold private conversations, with only 40% of patients in three bed bays feeling they could. The relationship between conversational privacy and the size of the bay was investigated by averaging the number of people who found conversational privacy to be acceptable. There was not a significant correlation found between ward size and conversational privacy (= -.322, ns).
100
Percentage of patients who felt they could hold a private converstaion (%)
80
60
40
20
0 1 3 4 6 7 12 Bay size (number of beds)
Figure 11.17 Patient privacy and bay size 230
11.4. Factors affecting staff perceptions of noise

This section considers the responses of staff on the main study wards in Bedford and Addenbrookes Hospital. Sources of noise annoyance and interference to staff were examined in detail in Chapters 7 and 9 in relation to each study ward. As with patient perception, these sources were found to be ward specific and so to look for general trends in this data would not be appropriate. This section looks further at more general relationships including whether gender, age, and length of service increase feelings of noise annoyance and noise interference in relation to work.
11.4.1. Overall
Figure 11.18 shows the perceptions of day and night time noise for all 66 staff respondents. It can be seen that the highest percentage of staff (25%) cite levels of annoyance and interference as slight, with 18% moderately annoyed by noise. Higher levels of annoyance are cited by only nine staff in total (13%), with eight choosing very noisy and one selecting extremely.
30
Annoyance Interference
25
Percentage of staff (%)
20
15
10
0 Not at all Slightly Moderately Very much Extremely
Staff annoyance and interfernce levels
Figure 11.18 Staff levels of annoyance and interference
11.4.2. Gender
Figure 11.19 shows the average feeling of noise annoyance and noise interference with work by staff gender. It can be seen that the average level of annoyance and interference on the ward is higher for male staff, however all ratings are fairly low in the range slightly to moderately for both men and women. Noise annoyance is rated consistently higher than interference with work by both male and female staff.
231
5 Noise annoyance Noise interference

Level of noise annoyance / interference (1=not at all / 5= extremely)
(n=13) (n=54)
1 Male Staff gender Female
Figure 11.19 Level of noise annoyance / interference by staff gender
11.4.3. Age
The next section considers whether the age of the staff member is related to the level of noise annoyance / interference they perceive in their working environment.

(n=3)
3
(n=34)
(n=11)
(n=16)
2
(n=3)
1 <20 20-30 31-40 Age group (years) 41-50 51-60
Figure 11.20 Level of noise annoyance / interference by staff age
It can be seen in Figure 11.20 that it is the younger members of the staff and the older staff who appear to have more extreme views. Staff under 20 years old appear to be untroubled by noise while 232
those in the 51-60 age group rate noise annoyance and interference more highly than any other age group (with a mean of 3.0 for both annoyance and interference). However, it should be noted that the sample sizes for both these age groups are very low (n=3). Statistically, no relationship was found between levels of annoyance and age (= -.262, ns), or levels of interference and age (= -.16, ns).
11.4.4. Time worked on the ward

The relationship between length of time worked on the ward and level of noise annoyance / noise interference is illustrated in Figure 11.21. It can be seen that in all cases the feelings of annoyance and interference are fairly low, predominantly in the range slightly to moderately, with both new staff and staff in the 4-5 year bracket reporting very low levels. Although ratings increase for the first three years, this is not a constant trend overall. One interesting point to note is that staff who have worked on the ward for over five years rate interference with work more highly than annoyance, in contrast to all the other time brackets.
Levelof noise annoyance / interference (1=not at all / 5= extremely)
(n=8)
(n=9)
(n=4)
(n=11)
(n=30)
2
(n=3)
1 <1 year 1-2 years 2-3 years 3-4 years 4-5 years 5+ years Time worked on the ward
Figure 11.21 Level of noise annoyance / interference by time worked on the ward
No statistically significant relationship exists between time worked on the ward and annoyance ratings (=.16, ns), and time worked on the ward and interference ratings (=.29, ns).
11.4.5. Time worked at the hospital

The relationship between length of time worked at the hospital and levels of noise annoyance / noise interference is illustrated in Figure 11.22. It can be seen that in most cases the feelings of annoyance and interference are fairly low, predominantly in the range slightly to moderately, with those new staff and staff in the 3-4 and 4-5 year brackets reporting very low levels. Although levels increase for the first three years, this is not a constant trend overall. As with time worked on the ward, staff who 233
have worked in the hospital for over five years rate interference with work more highly than annoyance; in this case in the moderately to very much range. Statistically, there is no significant relationship between time worked at the hospital and annoyance and interference ratings (= .16, ns and = .29, ns respectively).
5 Noise annoyance
Noise interference 4
(n=24)
3
(n=7)
(n=9) (n=21)
2
(n=3)
(n=2)
1 <1 year 1-2 years 2-3 years 3-4 years 4-5 years 5+ years Time worked at the hospital
Figure 11.22 Level of noise annoyance / interference by time worked at the hospital
11.4.6. Relationship between noise annoyance and noise interference

Considering the overall data set of 66 responses, the relationship between noise annoyance and noise interference was investigated. A statistically significant correlation between annoyance and interference ratings was found (= .730, p=0.01). This suggests that staff rating noise annoyance highly will also rate noise interference highly. If this is the case, it is possible that posing a single question on noise annoyance / interference would be sufficient to provide the information required.
11.5. Discussion
Analysis of the overall dataset on noise and patient and staff perceptions has yielded some interesting and surprising results. It has been shown that the size of a bay is not related to the noise levels in the bay during the day or at night, with some of the highest levels measured in single rooms and in three and four bed bays. Overall measured noise levels in a medical and surgical ward at the same hospital were found to be similar; however the content of the noise and the perceptions of the staff and patients differed. In the busier surgical ward, visiting time, the internal telephone, meal times, the nurse call, and trolleys were 234
all found to cause more interference to the work of the staff. Patients on the medical ward found the crying out of other patients caused a great deal of annoyance and disturbance, whereas patients on the surgical ward rated trolleys and the nurse call more highly. The average number of high level noise events (greater than 70 dB LAmax) were shown to be strongly correlated to overall measured noise levels both day and night. Analysis suggests that the impact of high level noise events on night time levels is considerably greater than during the day, as the increase of day time ambient levels provide a certain amount of masking. This suggests that patients would experience greater disturbance by high level noise events during the night. Noise levels have been shown to be directly related to RT, with lower noise levels corresponding to lower RTs, as expected. Analysis has found that for every 0.1 s decrease in the RT, there is a decrease in the average day time noise levels of 1.2 dB LAeq,16hr . This confirms the relationship that is illustrated by the ceiling intervention study in Chapter 8, where the addition of an acoustically absorbent ceiling was found to decrease both RTs and overall noise levels. The perception of the noise climate by male and female patients was found to differ, with male patients perceiving a noisier environment during the night than female patients. However, annoyance and disturbance ratings were the same for both genders. Patients whose stay on the ward was less than three weeks perceived the noise climate on the ward to be similar both day and night, however they were more disturbed by night time noise than longer term patients (> 3 weeks). However, longer term patients appeared to be more annoyed by day time noise than shorter term patients but were less sensitive to noise during the night. This suggests a certain amount of acclimatisation to noise on the ward at night. Differences in noise perception and annoyance were found between patient age groups. In terms of perception, day and night levels were generally perceived similarly and were fairly low, in the quiet to slightly noisy category, with patients in the 41-50 age group at night time having the worst perceptions of the noise environment, with an average rating of 3.2. In terms of annoyance and disturbance it was found that patients in the age groups 31-40 and above were much more disturbed by noise at night than during the day, with the highest percentage again in the 41-50 age group (74%). Perhaps this is an indication of the higher expectations of those patients in this particular age bracket. Patients staying in a single room were found to be more annoyed by noise than those in multi-bed bays, with 50% and 58% of single room patients reporting day time annoyance and night time disturbance respectively. Day and night time levels reported by patients in a multi-bed bay were found to be considerably lower during the day (20%) and slightly lower at night (53%). It was found that patients in beds situated by the window reported lower rates of day time annoyance and night time disturbance (17% and 47% respectively) than patients in other bed locations. 235
Satisfactory conversational privacy was reported by all patients in single rooms, and by similar percentages, 63% on average, of those in bays of four to twelve beds. Interestingly only 40% of patients in three bed bays felt that they could speak privately. Little difference was found in terms of noise annoyance and disturbance and conversational privacy between those patients with a hearing impairment and those without. Levels of noise annoyance and interference rated by staff were generally found to be low, in the most part in the slightly to moderately range. In terms of staff gender, male staff rated noise annoyance and interference slightly higher than female staff and in relation to age, the youngest members of the staff and the older staff had more extreme views. Staff under 20 years old appeared to be untroubled by noise, with those in the 51-60 age group rating noise annoyance and interference more highly than any other age group. However, even in this case it must be stressed that the mean rating of 3.0 is not particularly severe. No significant relationship was found between staff attitudes to noise and either the length of time they had worked on the ward or the length of time they had worked at the hospital. The main difference was in the 5+ years bracket in both categories, where staff rated noise interference with work more highly than noise annoyance. Perhaps this is related to the changes that these staff have seen over many years in relation to additional noise sources, for example, the more prevalent use of ward equipment with alarms, and it may be an indication of the worsening noise climate in hospitals.
11.6. Conclusions
The main conclusion of this chapter is that, contrary to current thinking, single bed rooms are not quieter than multi-bed bays. This has been shown both objectively, from the measured noise level data, and subjectively from patient perceptions of noise. This is an important finding given the current thinking and the preference for providing more single rooms in hospital wards. The results also highlight the need to reduce RTs in hospital wards in order to decrease the overall noise levels. Other points to note are that there is some evidence that longer term patients acclimatise to noise at night, but become more annoyed by noise during the day time; and that consideration needs to be given to the issue of speech privacy and how it may be improved, particularly in 3-bed bays. The following chapter looks at ways of controlling hospital noise using a multi-faceted approach. The validity of relevant standards is discussed and acoustic reporting metrics are also explored.
236
Acoustic Design for Inpatient Facilities in Hospitals Noise control in inpatient care ___________________________________________________________________________________________________________________
12. Noise control in inpatient care

12.1. Introduction
This study has shown that the problem of hospital noise is very complex in nature, with many different factors affecting the noise climate. This was also found to be the case in the previous research on this topic. It is felt that if any significant improvement is to be achieved, a multi-faceted approach is required, which should be centred on three main areas: Optimising the acoustic design of the ward Minimising the disturbance caused by equipment in use on the ward Modifying the behaviour of those on the ward This chapter discusses each of these areas in detail in relation to the findings from the current study. Validity of relevant standards and reporting metrics are also explored.
12.2. Optimising the acoustic design of the ward

This study focuses on a number of areas of acoustic design, including whether the design for infection control purposes has compromised ward acoustics; the effects of adding acoustic absorbency; and the impact of the ward design and building construction on the noise climate. These areas are considered and discussed further in the following section.
12.2.1. Design for infection control

One of the aims of the study was to investigate whether the design of a hospital building for infection control purposes was detrimental to the acoustic design. Three of the study wards which were built in the last decade, a time when concerns over infection control were being considered carefully in terms of building finishes. The study found that in each of the wards, suspended ceilings with good quality acoustic tiles were installed throughout all patient accommodation, and in some instances in the common areas. Reverberation times in all the areas with acoustic tiles were found to be very low. The use of acoustic tiles in these new buildings is largely due to the efforts of the acoustic product manufacturers, who have responded positively to infection control concerns by developing a number of different ceiling tiles specifically for use in healthcare buildings, and thus ensuring that the acoustic design of the wards is not compromised. Tiles are now available which can withstand cleaning using strong disinfectants, or even steam cleaning. Specific tiles have even been produced that are treated with antimicrobial agents, preventing any bacteria from growing on their surface.
237
12.2.2. The effects of adding acoustic absorbency

The latest acoustic design guidance, HTM 08-01: Acoustics (The Stationary Office, 2008), suggests that the most appropriate area for acoustically absorbent material should be a ceiling, with the minimum absorption area equivalent to 80%. The beneficial effects of installing an acoustic ceiling are illustrated by the ceiling intervention study, which was discussed in Chapter 8, and showed a reduction of 0.15 s in reverberation time (RT) and 2.4 dB in noise level. The relationship between RT and noise levels is discussed further in Section 11.2.3, which shows a statistically significant relationship between day time RT estimates and daytime LAeq,16hr (=.498, p<0.05). It is shown that for every 0.1 s decrease in the RT, there is a decrease in the average day time noise levels of 1.2 dB LAeq,16hr. This is reasonably consistent with the findings of the ceiling intervention study. These findings stress the importance of installing an acoustically absorbent ceiling in hospital wards.
12.2.3. Ward design

The five main study wards were all designed around a long central corridor, with patient accommodation generally situated on one side, and healthcare resources on the other. The pilot study ward was based on a racetrack design with patient accommodation situated on the outside of the building, and healthcare resources in the centre. Objective and subjective data from the study suggests that the most important aspect to be considered in terms of ward layout is not the overall design, but the positioning of patient accommodation in relation to the healthcare utilities. Careful thought must be given to ensure that there are no direct sound paths from potentially noisy areas to patient accommodation. The two study wards in Bedford Hospital provide several examples of where poor positioning of patient accommodation leads to annoyance and disturbance of patients, and are discussed below. The kitchen, ward clerks desk and staff room are all situated directly opposite a multi-bed bay in the two study wards. All these areas are potentially noisy, with bangs and crashes often heard from the kitchen area, and conversational noise heard from the staff room and at the ward clerks desk. Objective and subjective results both indicate that noise from these areas has a detrimental effect in terms of noise levels and patient annoyance and disturbance in the bay opposite. Two single rooms are positioned directly behind the main nurse station in both wards. Objective and subjective results again indicate that patients in these rooms are adversely affected by noise from activity at the nurse station: with staff conversation; the nurse call; and general activity clearly audible. Not only is thought required in the siting of patient accommodation, but care must also be taken in ensuring that areas that require easy access by staff are designed as such. One particular example of 238
impractical design is found in Ward N3 at Addenbrookes Hospital. Here, a number of rooms are situated close to the nurse station, including the clean and dirty utility rooms, and a general storeroom. None of these rooms contain anything requiring secure storage, but for some reason all doors to these rooms have been fitted with security access via a key code pad. Staff entering these rooms often have their hands full, and so to overcome the need to input the security code, the doors to these rooms are all left on the latch. Unfortunately, this has a negative impact on the noise environment as the doors literally bounce when shutting, causing a loud bang. In fact this noise accounted for 48% of the total number of trigger files captured during the measurement period in this location, with levels consistently measured at 79 dB LAmax. One issue highlighted by the study is that of open door nursing, which is still used by clinicians in UK hospitals today. The study found that the staff have been trained to carry out their nursing duties with doors to the patient bays and single rooms left open at all times for observation purposes. The only exception to this is if barrier nursing is required, and then doors to a single room may be closed. Several sections in the latest acoustic healthcare design guidance HTM 08-01 (The Stationary Office, 2008) advise on levels of noise attenuation between rooms, and of the type and properties of suitable acoustic doors to be installed on the wards. It is considered extremely important that an understanding of the way the building will be used when occupied is considered when specifying the acoustic design. Considerable expense may be incurred in relation to the installation of acoustic doors and sound attenuating material, when in reality they will be of no actual benefit, as the staff will leave the doors to the patient accommodation open.
12.2.4. Building construction

The majority of the ward buildings in this study are primarily of concrete construction, with concrete floors, suspended ceilings, and stud or block partitions. As this study is concerned with occupied buildings, it has not been possible to view the building construction as a singular entity. However, subjective responses have provided clues as to areas within the building construction that may cause annoyance to staff and patients. These areas are discussed below. The mechanical ventilation system was cited by ward staff in Sky Ward, GOSH. The installed system is centrally controlled, but air flow was found to be noticeably different in some areas of the ward, with some patient accommodation and staff offices adversely affected by noise and heat from the system. At Bedford Hospital, where wards are naturally ventilated, annoyance and disturbance to patients due to external noise increased during the summer months, when the windows were open, with 37% of patients citing external noise as a night time disturbance. This appeared to be a less of problem in the cooler months, with lower percentages of patients (13%) disturbed.
239
Single glazing also accounted for additional disturbance by external noise, as discussed in Section 7.9.2. However, it is interesting to note that patient opinions were split, with some finding the lack of attenuation by the single glazed unit to be annoying, but with others finding the connection with the outside world beneficial. Careful consideration needs to be given to the choice of mechanical and / or natural ventilation, as both have their own inherent problems. One study ward is of a very different construction and this is worth discussing in further detail. Ward N3 is a modular ward at Addenbrookes Hospital and is of mainly timber construction. This block was opened in 2009 and was built to comply with acoustic standard HTM 2045. It has a good quality acoustic ceiling and all room partitions are sound insulated, but it is the floor that is of interest. When the ward first opened, the springy nature of this floor generated many complaints from staff, who found it detrimental to their feet. Remedial work was carried out to further stiffen the floor, and although improved, the floor does appear to have a negative effect on the noise climate, by magnifying certain sounds. Groups of people walking past the nurse station were found to generate noise levels exceeding 70 dB LAmax, a footfall noise level not found on any other study wards. This was confirmed by questionnaire responses, where 40% of ward staff claimed they found the sound of footsteps annoying, and 20% of patients cited footsteps as a night time disturbance. These percentages are the highest found in relation to annoyance and disturbance from footsteps on any study ward. The noise of trolleys also appears to be exacerbated by the timber floor construction, with 40% of patients citing trolleys as a source of annoyance. Care should be taken when choosing the type of floor construction for use in hospital wards, as the use of a timber floor can exacerbate noise levels and is found by staff cause discomfort.
12.2.5. Building age and overall noise levels

The age of the hospital buildings included in the study varied from less than ten years to around 40 years, with some new build (within the past ten years) and others built in the 1970s and 1980s. The 1970s ward lacked space, having a larger number of beds than it was originally designed for (as discussed in Chapter 9.3.1) and the largest bay sizes of all the study wards. Occupied RTs were found to be slightly higher than the newer wards with acoustic ceilings, but these values were still relatively low (ranging from 0.4 s to 0.6 s at 1 kHz). The 1980s wards were single glazed, which accounted for additional disturbance by external noise, as discussed in the previous section. The positioning of patient accommodation in relation to the kitchen, ward clerks desk and nurse station was also found to be detrimental to noise levels. Buildings constructed in the last ten years were found to have good quality acoustic ceilings and very low reverberation times (for example, less than 0.3 s at 1 kHz in Ward N3). Wards were generally found to be much more spacious, with wider corridors, plenty of equipment storage and improved bed spacing.
240
Figure 12.1 shows the day time noise levels (LAeq,16hr) measured in the patient accommodation grouped by building age. It can be seen that with only one exception, overall measured LAeq,16hr values lie within a 10 dB band, between 50 and 60 dB LAeq,16hr. A statistically significant correlation was found between building age and overall day time noise levels (= -.589, p=0.01), suggesting that newer buildings are quieter.
70 60
LAeq, 16hr (dB)
50 40 30 20 10 0 1970's 1980's 2000's
Decade in which building comissioned
Figure 12.1 Average day time levels by building age for all patient accommodation Figure 12.2 shows the night time noise levels (LAeq,8hr) measured in the patient accommodation grouped by building age. It can be seen that all overall measured LAeq,8hr values lie within a 10 dB band, between 41 and 51 dB LAeq,8hr. In the case of night time noise levels and building age no significant relationship was found (= -.348, ns). However, the negative correlation coefficient indicates a trend for the overall noise levels to decrease in the newer buildings as with the daytime levels.
70 60
LAeq, 8hr (dB)
50 40 30 20 10 0 1970's 1980's 2000's
Decade in which building commissioned
Figure 12.2 Average night time levels by building age for all patient accommodation
241
12.3. Ward equipment

The study shows that the cacophony of alarms and tones emitted by the electronic equipment in use on the ward is becoming a major source of noise. Items such as nurse call systems, internal telephones and medical equipment occur frequently and were found to generate high noise levels, with some alarms left for some time before they are reset. These types of systems are continually cited by staff and patients as sources of annoyance and disturbance. The poor design and maintenance of other types of ward equipment is also found to contribute to high levels of noise, and is discussed further in this section, with possible noise control measures suggested where appropriate. Following the pilot study carried out at GOSH, follow up meetings were held with staff to discuss the results of the study and possible improvements that could be made to the noise environment. Where appropriate the feedback from these meetings is incorporated into the following sections.
12.3.1. Nurse call systems

In preliminary discussion with ward managers, the nurse call was cited as a source of annoyance, and this was supported by the questionnaire responses from staff on all the study wards. In some instances annoyance was related to correspondingly high noise levels, for example a level of 72.7 dB LAmax was measured at the nurse station in the surgical ward at Bedford Hospital. Some of the nurse call annoyance was due to the level of the noise emitted by the system. In several cases it was found that the day and night volume settings, which were meant to change the volume level, appeared to be faulty, with no difference found in the volume of the tone emitted. In other cases the volume was simply too loud and not adjustable. Unfortunately, although annoying this did not get logged as no clear reporting structure appeared to be available for issues of this nature. Commissioning of the system also appeared to be at fault on occasion, with staff finding that the system was not performing as they would wish in terms of functionality. Staff in one particular study ward commented that they were not able to hear the nurse call if they were attending to a patient in a different bay, and this made it hard to carry out their nursing duties effectively.
12.3.2. Internal telephones

As with the nurse call system, the internal telephone was cited as a source of annoyance and interference by staff, and as a disturbance by patients. Comments made suggest that one of the main issues is not about the volume of the ring tone, but the length of time the phone rings before being answered. In all study wards there are a number of wired, static phones on desks. In some cases these are used alongside personal paging systems. Given that the majority of people own mobile telephones which
242
have silent, vibrating functionality, surely the ringing desk phone is an unnecessary object in any hospital ward? In a post study meeting with staff from the pilot study ward at GOSH, a number of interesting points were made regarding telephone use on the ward. These are detailed below. The responsibility of the ward clerk on this ward is to act as a first point of contact for external telephone calls. When away from her desk, which is a regular occurrence, unanswered telephone calls are redirected through to the ward to be answered by a member of the clinical staff. This of course wastes the clinicians time, as they then have to take a message back to the ward clerks desk. It was suggested that by simply installing a voice mail system many of these unnecessary calls could be dealt with automatically. This idea was positively received by all staff and has been implemented. Some newer staff members commented that they were not aware of how to turn down volume levels of the telephone ring tones. This suggested a possible issue with the training of new staff. A new CISCO system is currently being trialled in another ward at GOSH which uses internet technologies. The system uses wireless telephones which are given to each member of staff and could potentially replace all existing telephones and paging systems, and the nurse call. Staff felt that if implemented, this system would be beneficial.
12.3.3. Medical equipment alarms

Medical equipment alarms were also one of the major sources of annoyance to both staff and patients, with the alarms emitted often loud, and on occasion seemingly unnecessary. One example captured during the study was that of a patient monitor which bleeped at a level of 75 dB LAmax every 30 seconds for two and a half hours in a single room. This would surely have disturbed the patient, who would have been trying to rest, and must surely warrant the question: if this did not require intervention, then why was there a tone emitted at all? It is possible that the setup of many pieces of medical equipment is so complex that once they are commissioned, the settings are never changed. For example, one particular ECG machine identified had a total of 16 different settings. Again this suggests that there should be some collaboration between the staff who use these pieces of equipment and the installer and / or manufacturer. Continued staff training may also be an issue. It is possible that when first installed, staff on the ward are provided with the necessary training on how to tailor a piece of medical equipment to their own requirements. However, with the extremely high turnover of staff in healthcare and the regular use of temporary agency staff, this knowledge may be lost relatively quickly. Perhaps follow on training should be provided periodically, or staff should be issued with necessary information on induction.
243
12.3.4. Doorbell
The pilot study ward at GOSH and each study ward at Addenbrookes Hospital could only be accessed using a security pass. For visitors to announce their arrival at the ward entrance, a doorbell was installed. These doorbells cause a great deal of annoyance and interference to staff not only due to the volume, which has been measured at levels as high as 80.6 dB LAmax in one case, but with visitors constantly pushing the buzzer until the entrance door is opened. Staff felt that visitors did not appreciate that they may be busy on the ward and could not simply drop everything to answer the bell. In the post study meeting with staff at GOSH it was felt that limiting the number of times the doorbell rang to once every 30 seconds would help a great deal in alleviating this annoyance. Staff also felt that somehow setting ward visitors expectations regarding the length of time it may take to answer the door might also help, perhaps using something as simple as a clear written notice on the entrance door.
12.3.5. Rubbish bins

Rubbish bins on several of the study wards were found to cause annoyance and disturbance to patients, and high levels of noise. Although many of the bins in use have been specifically designed to have quiet closing lids, they are often positioned too close to a sink or wall. On opening, the lid hits the nearby surface or object, causing a loud bang and entirely defeating the point of the quiet closing mechanism. The body of the bin is also a problem, as it is generally solid and undamped; hence discarding a heavy object also causes noise. Design improvements have been made to rubbish bins, with the addition of quiet closing mechanisms, however the correct siting of the rubbish bin is imperative; wall spacers and some additional damping in the body of the bin would alleviate much of this unwanted noise.
12.3.6. Ward furniture

Furniture scraping on the floor was found to be a source of high level noise in several study wards. This could be simply and cheaply controlled by fitting rubber feet or wheels to the chairs and other furniture used.
12.3.7. Wheeled equipment

Much of the equipment used on the wards has wheels to allow for portability. This includes patient beds, ward furniture, medical equipment, as well as the standard trolleys used to deliver meals, drinks, medication, linen and other supplies. Noise from trolleys has been measured at levels as high as 85 dB LAmax, which is due to in part to inherent flaws in their design. Many of these pieces of equipment are metal; have no damping; and have ill-adjusted wheels and unsuitable tyres. It is interesting to note that the Ministry of Health Hospital Design Note 4: Noise Control (Her Majestys Stationery Office, 1966) recognised trolleys as a major source of noise, recommending that 244
good quality rubber tyres be fitted. It is very sad to think that 45 years later, these inherent design flaws have not been improved.
12.3.8. Ring binders

In all the study wards, patient notes and other reading materials are stored in ring binders. Without exception the clicking shut of these ring binders was captured at high levels and on numerous occasions during the study, with measured levels as high as 85 to 90 dB LAmax. Often these levels were captured at a nurse station during the night, when nursing staff were catching up on administrative tasks. It is thought that other types of folder could easily be sourced which do not make use of the ring binder spring loaded mechanism and therefore do not cause unnecessary disturbance to those patients close by, who may be asleep or trying to sleep.
12.3.9. Doors
Banging doors were cited as disturbing in a number of the study wards and this could be simply remedied by fitting or adjusting quiet door closers, or in the case of metal cupboards, installing some damping. One occasion of poor workmanship was discovered in relation to a heavy fire door. Although the quiet closer was working correctly, the door frame was too large and on closing the door would rattle loudly, the sound carrying down the main ward corridor. Whether due to poor maintenance, lack of a quiet door closer or poor workmanship, all these problems are relatively cheap to fix. However, what appears to be unclear is the reporting mechanism for staff to log a fault of this nature. An example of this is illustrated in the pilot study ward at GOSH. Staff here became so irritated with doors banging that they took matters into their own hands and draped towels over the doors to prevent the noise. This was soon stopped by the ward manager on infection control grounds. It is unknown if the problem was rectified.
12.4. Human behaviour

Human behaviour and attitudes are inherently difficult to change in the long term. Studies have shown that short term noise level improvements can be gained if individuals are made aware of the impact of their actions in relation to noise. However, this is difficult to implement successfully for any length of time without a great deal of support from ward management to ensure all policies are strictly adhered to. There are many aspects of behaviour cited by both staff and patients in the subjective responses to the noise environment. Noise level measurements also captured certain activities causing high levels of noise. Cleaning and the changing of rubbish bins are very necessary activities, but ones that a number of patients felt could be carried out more quietly. Noise level measurements also found this to be the case. Loud staff telephone conversations were also mentioned, especially when they were of a 245
personal nature and on a mobile phone. Domestic staff talking loudly or shouting at each other across the ward was found to be annoying by the patients. Visiting time was listed by a high percentage of staff as annoying and causing interference to their work. Patients also found large numbers of visitors round a bed disturbing, and some felt that visiting hours should be more strictly adhered to. Ward managers had very differing views regarding leniency around visiting times, with some managers feeling that the more contact patients had with their friends and family, the better. This is obviously a very subjective area, but it is important that visitors are respectful of other patients on the ward and should also be aware of staff and the duties that they must perform. If visitors are being loud and impolite it is surely up to the senior members of staff to control the situation and let it be known that their behaviour is unacceptable and will not be tolerated. Noise levels in single rooms were often found to be higher and less consistent than those measured in multi-bed bays. In some instances this was due to the behaviour of those visiting the patients. High noise levels due to conversation were often captured for long periods of time, with visitors staying on for an hour or more after the end of designated visiting hours. On occasions staff appeared to turn a blind eye as the doors to the rooms could be closed and so this would not cause a disturbance to the other patients on the ward. Of course all patients are recovering and need their rest, and so could these longer, louder visiting times be potentially detrimental to the patients recovery? Patients crying out in confusion or pain were found to be annoying and disturbing by other patients, particularly on the medical wards, where there are a higher percentage of elderly patients being cared for. Patients suffering from confusion or dementia can often be very vocal and it can be both distressing and disturbing to the other patients in the bay. This is a particularly difficult area to control, short of segregating these patients, which is not practical in a medical ward running at nearly 100% occupancy at all times. Every patient on the study wards was provided with entertainment in the form of a telephone, TV and radio console, often provided by Patientline. In most cases patients were provided with headphones to watch TV or listen to the radio without causing disturbance to others. However, ward managers admitted that the use of these headphones was not always enforced, resulting in many patients listing TV / radio use as annoying and disturbing. Perhaps if staff were more aware of the level of annoyance / disturbance this causes others, they would be more inclined to enforce headphone use. Questionnaire responses indicated that on some of the study wards mobile phone use caused annoyance and disturbance to patients. The use of mobile phones in hospital wards is a contentious issue with each hospital having their own discretionary policy. The policy at Bedford Hospital, for example, prohibits the use of mobile phones on the wards, suggesting calls are made in the lobby areas directly outside the ward. Questionnaire responses from one of the study wards at this hospital indicated that this policy was not being enforced. Ward managers themselves also have their own opinions on the use of mobile phones. One of the study ward managers felt so passionately that the telephone charges incurred as a result of the Patientline system were unacceptable, that she allowed patients to use their mobile phones on the ward. 246
12.5. World Health Organisation guidelines

As discussed in Chapter 2, the most recent edition of the World Health Organisation (WHO) Guidelines for Community Noise was published in 1999 (Berglund et al, 1999). In relation to noise in hospitals, the guidelines state that the critical effects of noise are on sleep disturbance, annoyance and the communication interface, including interference with warning signals. A summary of the guidelines is shown in Table 12.1. Table 12.1 World Health Organisation guidelines for hospital wards and treatment rooms Specific Environment Hospital, ward rooms, indoors Hospital, ward rooms, indoors Hospital, treatment rooms, indoors Critical Health Effects Sleep disturbance Sleep disturbance Interference with rest and recovery LAeq (dB) 30 30 As low as possible Time Base (Hours) Night time (8 hours) Day time and evenings (16 hours) LAmax (dB) 40 -
The LAeq value stipulated by the guidelines for both day and night time on a ward is 30 dB LAeq. Interestingly, no noise levels measured in the study wards were found to comply with these levels, which is in accordance with the findings by Busch-Vishinac et al in 2005. In their comprehensive study the authors compiled data from all comparable studies post 1960 which listed LAeq noise measurement values. Not one single study showed a hospital which complied with the WHO guidelines for hospital noise. This raises the question of the validity of these guidelines. The WHO guidelines also stipulate a time base in terms of day and night, with day time beginning at 07.00 and ending at 23.00. Noise levels collected in each study ward suggest that this division is not realistic for hospital wards, with ward activity generally beginning earlier than 07.00 and decreasing before 23.00. It was also noted that noise levels were generally found to diminish after the evening meal had been served at around 18.30, suggesting that an evening period may also be applicable and more realistically reflect activity levels on the wards. In Chapter 6, the fact that the noise levels decreased earlier during the evening was assumed to be because this was a childrens ward. However, further analysis of adult wards suggests that this is the case for all wards.
12.6. Acoustic parameters

As discussed in Section 11.2.3, noise measurements have been presented throughout this study as average day and night LAeq levels for each bay, with average hourly noise levels (LAeq,1hr) plotted over 24 hour intervals. Levels presented in these ways do provide a general indication of the daily patterns of noise and the overall levels measured in each bay, but are generally found to be very similar with
247
relatively small amounts of variation. Busch-Vishniac et al (2005) also encountered this lack of variation, which was surprising given that their data was gathered from widely differing sources. To illustrate the fluctuating nature of the measured noise, and help build up a more detailed picture of its content, high level noise events (over 70 dB LAmax) have been investigated throughout this study and these are reported in some detail in Chapters 7 and 9. The use of the trigger files captured has provided a means of identification of all high level sources of noise, and it is felt that by looking at the types and also the numbers of high level noise sources over a measurement interval, this study has gone some way in providing data to describe the realities of the noise climate on the wards.
12.7. Conclusions
The discussion in this chapter has shown that it is important to consider noise control at the design stage of a hospital building. However, much of the impact of noise is related to the patient group, the activity and behaviour of staff and patients, and the supporting equipment. The next chapter presents the overall conclusions to the study and recommendations for further work.
248
Acoustic Design for Inpatient Facilities in Hospitals Conclusions ___________________________________________________________________________________________________________________
13. Conclusions
13.1. Introduction
This study has investigated, through objective and subjective surveys, the noise climate and acoustic design within general inpatient facilities in the UK, and their influence on the acoustic comfort of patients and staff. Noise and acoustic surveys have been carried out in six inpatient wards in three major UK hospitals, with corresponding questionnaire surveys of staff and patients. It has been shown that high levels of noise are not confined to ICU and operating theatres, but are found to be significant throughout inpatient wards in UK hospitals. Overall conclusions from the study are presented below, together with recommendations arising from the study and suggestions for further work.
13.2. Overall conclusions

The study has highlighted the need to understand the way the building will be used when occupied when specifying the acoustic design of a hospital building. Without due consideration expense may be incurred in relation to the installation of acoustic doors and sound attenuating material, when in reality they may be of no actual benefit. This has been illustrated by the open door nursing policy in use in UK hospitals, where staff are trained to nurse with doors open at all times for observational purposes. Careful thought must be also given to the ward layout to ensure that there are no direct sound paths from potentially noisy areas, such as kitchens, healthcare utilities and nurse stations, to patient accommodation. It has been shown that lower reverberation times result in lower noise levels. This stresses the need for sufficient acoustic absorbency on the wards in order to reduce noise, which can be attained by using good quality acoustic ceiling tiles. A new range of ceiling tiles that withstand stronger cleaning agents are now generally available for use in hospitals following infection control concerns. This proactive response by the manufacturers has ensured that in future the acoustic design of the wards need not be compromised by Control of Infection policies. The study found some evidence to suggest that there was a slight downward trend in noise levels and building age. Measured day time noise levels were found to decrease between buildings built in the 1970s, 1980s and in the last ten years suggesting the more recent use of acoustic design guidelines may be having some effect.
249
13.2.2. Patient accommodation

The study has found that contrary to general assumption, noise levels are unrelated to bay size and numbers of beds, with some of the highest noise levels measured in single bed rooms. Furthermore, patients in single rooms were found to be more annoyed by day time noise and more disturbed by night time noise than those in multi-bed bays. Conversational privacy was rated most poorly by patients in 3-bed bays, with improvements found in four to twelve bed bays. 100% of patients in single rooms were found to be satisfied with speech privacy. Measured noise levels in a medical and surgical ward at the same hospital were found to be similar, but the sources of the noise and the staff and patient perceptions differed.
13.2.3. Staff and patient perceptions

Ratings of noise annoyance and interference by staff at the main study hospitals were generally low; mostly in the slightly to moderately range. However, there were noticeable differences found with longer term staff (> 5 years service) rating noise interference more highly than noise annoyance, and more highly than staff who had worked on the ward or in the hospital for a shorter time. Analysis of the patient questionnaire responses from the main study sites indicated a shift in the attitude amongst longer term patients (> 3 weeks stay) to the noise climate, with increased sensitivity during the day but decreased sensitivity at night. This may suggest a certain amount of acclimatisation to night time noise. It was also found that patients in the 41-50 age group were more highly annoyed and disturbed by noise than any other age group.
13.2.4. Ward equipment

The study shows that the cacophony of alarms and tones emitted by the electronic equipment in use on the ward are a major source of noise, with the nurse call, internal telephone, ward doorbell and medical equipment continually cited as causes of annoyance throughout the study. Ward staff found much of this equipment to be either inflexible, faulty, over complicated or not fit for purpose. Training in the use of electronic equipment appeared to be lacking. Much of the noise attributed to ward furniture could be rectified by simple noise control measures, such as fitting rubber feet or wheels to furniture; fitting and maintaining quiet door closers; a common sense approach to the positioning of rubbish bins; and the replacement of ring binders. Simple ward maintenance was found to be an issue, with no clear fault logging process available to sort out problems such as an ineffectual door closer.
250
13.2.5. Human behaviour

Although human behaviour and attitudes are inherently difficult to change in the long term, it is felt that if staff, patients and visitors were made aware of the effects of their actions on the noise climate, they might think more carefully in the future. Thus, education of users of a hospital, accompanied by support and further enforcement by senior management may be needed to change the current culture to a culture of quiet.
13.2.6. Guidelines
Most of the current acoustic design guidelines were not found to be applicable to an occupied building; only the WHO guidelines for healthcare were concerned with noise levels in an occupied ward. However, these guidelines have been shown to be unrealistic and a review is needed of both noise levels and day and night divisions.
13.3. Recommendations
The study has highlighted the need for adequate consultation with the buildings users before the acoustic design criteria is specified for a new or existing building. This will allow a realistic set of acoustic requirements to be established that will positively support the buildings users and ensure that money is not spent on un-necessary acoustic treatment. Post occupancy surveys should be carried out some months after a new hospital building is first commissioned to ensure that the building is functioning as it was designed. Staff should be involved in this process, providing feedback regarding any problems they may have in terms of their environment and the ward systems. Feedback mechanisms should be put in place that encourage problems to be reported, no matter how small. Due to a lack of funding for new hospital buildings, refurbishment of the existing hospital building stock will be carried out over the next decade more extensively than ever. To ensure that the limited funding available targets the correct areas, it is important that adequate consideration is given to the staff and patient experience. Some of the differences in opinion identified during the study highlight the need to focus on areas of the greatest impact. For example, fitting expensive double glazed windows to an entire building may not have as much effect as changing the suspended ceilings, or making changes to some of the more problematic ward systems. Ongoing ward maintenance needs a proper reporting structure to be put in place. If this does not exist, busy staff will simply put up with issues that cause them annoyance and interference, assuming that nothing will be done. Regular reviews either by the estates teams or by ward management, who subsequent report to the estates teams, should be carried out.
251
Adequate training on the use of ward systems should be provided for new and existing staff and this should be an ongoing process. It is imperative that manufacturers take the noise impact into consideration when developing any type of new system or piece of equipment to be used on a hospital ward. Collaboration with staff is strongly recommended to ensure that their requirements are incorporated. Simplicity of use should also be a priority. It is felt that with properly implemented and maintained systems and equipment; an effective feedback system; a proactive approach to the management of noise issues by senior staff; and a common sense approach to noise control, many of the issues highlighted by the study could be improved and even eradicated completely.
13.4. Further work

The following areas would benefit from further work: Raising the awareness of hospital ward equipment and systems manufacturers as to the potential impact of their equipment on hospital noise, and on staff and patients. Design of medical equipment alarms to ensure that tones are only emitted when necessary, and that the alarms themselves are suitable for the event type. Development of more realistic guidelines for noise levels in occupied wards, including a more suitable day, evening and night division. Investigation of the feasibility and the trialling of replacement technology which incorporates wireless and silent technology that could replace existing systems such as the internal telephone, doorbell and nurse call.
252
Acoustic Design for Inpatient Facilities in Hospitals References ___________________________________________________________________________________________________________________
References
Aaron, J.N., Carlisle, C.C., Carskadon, M.A., Meyer, T.J., Hill, N.S., & Millman, R.P. (1996). Environmental noise as a cause of sleep disruption in an intermediate respiratory care unit. Sleep, 19(9), 707-710. Adam, K., Oswald, I. (1977). Sleep is for tissue restoration. Journal of the Royal College of Physicians London, 11, 376-388 Allaouchiche, B., Duflo, F., Debon, R., Bergeret, A., & Chassard, D. (2002). Noise in the post anaesthesia care unit. British Journal of Anaesthesia, 88(3), 369-373 Bailey. E., Timmons, S. (2005). Noise levels in PICU: An evaluation study. Paediatric Nursing, 17(10), 22-26. Barlas, D. Sama, A.E., Ward, M., Lesser, M.L. (2001). Comparison of the auditory and visual privacy of emergency department treatment areas with curtains versus those with solid walls. Annals of Emergency Medicine, 38(2), 135-139. Berg, S. (2001). Impact of reduced reverberation time on sound-induced arousals during sleep. Sleep, 24(3), 289-292. Bergmark, K., Janssen, M.R. (2008). Developing Acoustical Policies around in EU countries. Proc. Acoustics 08, Paris, 1319-1324 Berglund, B., Lindvall, T., & Schwela, D. H. (1995). Guidelines for community noise. World Health Organisation: Protection of the Human Environment. Berger, R.J., Phillips, N.H. (1995). Energy conservation and sleep. Behavioural Brain Research, 69, 65-73. Blomkvist, V., Eriksen, C. A., Theorell, T., Ulrich, R. S., & Rasmanis, G. (2005). Acoustics and psychosocial environment in coronary intensive care. Occupational and Environmental Medicine, 62, 1-8. Bork, I. (2000). A comparison of room simulation software the 2nd round robin on room acoustical computer software. Acta Acustica, 86, 943-956. Boulter, N. (2007). The role of acoustic design in comfort, health and wellbeing. Proc. Institute of Acoustics, 29, Part 2. British Standard BS EN ISO 3382-2 (2008). Acoustics Measurement of room acoustic parameters part 2 reverberation times in ordinary rooms Buelow, M. (2001). Noise measurements in four Phoenix emergency departments. Journal of Emergency Nursing, 27, 23-26. Busch-Vishniac, I., West, J., Barnhill, C., Hunter, T., Chivulula, R. (2005). Noise Levels in John Hopkins Hospital. Journal of the Acoustical Society of America, 118(6), 3629-45. 253
Elander, G., Hellstom, G. (1995). Reduction of noise levels in intensive care units for infants: Evaluation of an intervention program. Heart Lung, 24(5), 376-79. Fife, D., Rappaport, E. (1976). Noise and hospital stay. American Journal of Public Health, 66(7), 68081. Freedman, N. S., Gazendam, J., Levan, L., Pack, A. I., Schwab, R. J., (2001). Abnormal sleep/wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit. American Journal of Respiratory Critical Care Medicine, 163(2), 451-57. Gabor, J. Y., Cooper, A. B., Crombach, S. A., Lee, B., Kadikar, N., Bettger, H. E., et al. (2003). Contribution of the intensive care unit environment to sleep disruption in mechanically ventilated patients and healthy subjects. American Journal of Respiratory and Critical Care Medicine, 167(5), 708-715. Hagerman, I., Theorell, T., Ulrich, R.S., Blomkvist, V., Eriksen, C.A., & Rasmanis, G. (2005). Influence of coronary intensive care acoustics on the physiological states and quality of care of patients. International Journal of Cardiology. 98(2), 267-70. HFN 30 Infection Control in the built environment design and planning (2002). NHS Estates Health Technical Memorandum 08-01 Acoustics (2008). The Stationery Office Health Technical Memorandum 56 Partitions (1997). NHS Estates Health Technical Memorandum 60 Ceilings (2005). NHS Estates Health Technical Memorandum 2045 Acoustics: Design Considerations (1996). NHS Estates Hodge, B., & Thompson, J.F. (1990). Noise pollution in the operating theatre. Lancet, 335(8694), 891894. Holmberg, S.K., Coon, S. (1999). Ambient sound levels in a state psychiatric hospital. Archives of the Psychiatric Nursing, 13(3), 117-26. Johnson, P., Thornhill, L., (2006). Noise reduction in a hospital setting. Journal of Nursing Care Quality, 21(4), 295-97. Kahn, D. M., Cook, T. E., Carlisle, C. C., Nelson, D. L., Kramer, N. R., Millman, R. P. (1998). Identification and modification of environmental noise in an ICU setting. Chest, 114(2), 535-40. Kendrick, P., Li, F. F., Cox, T. J., Zhang, Y., Chambers, J. A. (2007). Blind estimation of reverberation parameters for non-diffuse rooms. Acta Acustica united with Acustica. 93, 76-770 Kendrick, P. (2009). Blind estimation of room acoustic parameters from speech and music signals. PhD Thesis, University of Salford. Kendrick, P., Shiers, N., Conetta, R., Cox, T. J., Shield, B. M., Mydlarz, C. (2011). Blind estimation of reverberation time in classrooms and hospital wards. Applied Acoustics, accepted for publication
254
Kracht, J., Busch-Vishniac, I., West, J. (2007). Noise in the operating rooms of Johns Hopkins Hospital. Journal of the Acoustical Society of America, 121(5), 2673-2680. MacLeod, M., Dunn, J., Busch-Vishniac, I., West, J. (2007). Quieting Weinberg 5C: A case study in hospital noise control. Journal of the Acoustical Society of America, 121(6), 3501-3508 McLaughlin, A., McLaughlin, B., Elliott, J., & Campalani, G. (1996). Noise levels in a cardiac surgical intensive care unit: A preliminary study conducted in secret. Intensive Critical Care Nursing, 12(4), 226-230. Ministry of Health Hospital Design Note 4: Noise Control (1966) Her Majestys Stationery Office Moorthy, K., Munz, S., Undre, S., Darzi, A., (2004). Objective evaluation of the effect of noise on the performance of a complex laparoscopic task. Surgery (St.Louis), 136, 25-30. Murthy, V. S., Malhotra, S. K., Bala, I., & Raghunathan, M. (1995). Detrimental effects of noise on anaesthetists. Canadian Journal of Anaesthesia, 42, 608-11. National Standards of Cleanliness for the NHS (2001). NHS Estates Orellana, D, Busch-Vishniac, I., West, J. (2007). Noise in the adult emergency department of John Hopkins Hospital. Journal of the Acoustical Society of America, 121(4), 1996-1999. Popplewell, A. (2008). The future of UK hospital design. Proc. Acoustics 08, Paris, 1325-1330 Pugh, R.J., Jones, C., Griffiths, R.D. (2007). The impact of noise in the intensive care unit. Intensive Care Annual Update 2007, 942-949. Raymond, I., Nielsen, T.A., Lavigne, G., Manzini, C., Choiniere, M. (2001) Quality of sleep and its daily relationship to pain in adult burn patients. Pain, 92, 381-388 Topf, M., & Dillon, E. (1988). Noise-induced stress as a predictor of burnout in critical care nurses. Heart & Lung, 17(5), 567-574. Tsiou, C., Eftymiatos, G., Katostaras, T. (2008). Noise in the operating rooms of Greek Hospitals. Journal of the Acoustical Society of America, 123(2), 757-765. Ulrich, R., Zimring, C., Quan, X., Joseph, A., Choudhary, R., (2004). The Role of the Physical Environment in the Hospital of the 21st Century: A Once-in-a-lifetime Opportunity. Report to the Center for Health Design for the Designing the 21st Century Hospital Project. Van de Glind, I., de Roode, S., Goossensen, A. (2007), Do patients in hospitals benefit from single rooms? A literature review. Health Policy 84, 153-161 Yildirim, K., Akalin-Baskaya, A., Celebi, M. (2007). The effects of window proximity, partition height, and gender on perceptions of open-plan offices. Journal of Environmental Psychology 27, 154-165
255
Acoustic Design for Inpatient Facilities in Hospitals Appendices ___________________________________________________________________________________________________________________
Appendix A
1. 2. 3. 4. 5. 6. Study publicity poster Staff information sheet Staff questionnaire Patient questionnaire Response from the National Research Ethics Service (NRES) Ethical approval received from London South Bank University
256
1.
Is it too noisy?
Research into the Acoustic Environment
Research is currently being carried out by the Estates Team here at XXXX Hospital together with London South Bank University. The purpose of this research is to understand the impact of noise on patients and staff in the wards. We are also keen to find out whether the design and materials used in our buildings help to make the buildings less or more noisy for everyone who uses them. Research Steps The first part of the research will involve taking some sound level measurements. This will help us to build up an understanding about the noise levels and sources of high level noise in the ward environments. To enable these measurements to be made, a portable sound level meter will be used. This meter measures sound levels in Decibels. The meter is small and unobtrusive and please be assured that it will be sterilised prior to use.
Sound Level Meter
The second part of the research is about understanding your feelings about noise - the patients and staff who spend time in the ward environments. A short questionnaire of no more than 10 minutes in length will allow us to build up an understanding of how the building is viewed in terms of noise. This will be on a completely voluntary basis and all information collected will be anonymous. Your involvement will be greatly appreciated to help positively influence the environment for the future.
For further information or to register your interest in participating in the study please contact XXXX in the Estates Team. Alternatively email Nicky Shiers at shiersn@lsbu.ac.uk
257
2.
Is it too noisy?
You are being invited to take part in a research study. Before you decide it is important for you to understand why the research is being carried out and what it will involve. Please take time to read the following information carefully. Talk to others about the study if you wish. Please ask us if there is anything that is not clear or if you would like more information. Take time to decide whether or not you wish to take part. The aim of this research is to investigate the impact of noise on the staff such as yourself, in the clinical environment in which you work. The results will enable us to understand whether the design and the materials used in the hospital buildings have a positive or negative impact on your acoustic comfort. This study is being completed as part of a PhD at London South Bank University and is being run in conjunction with the Estates Team here at XXXX Hospital. The initial part of the study will involve the researcher making some sound level measurements. This will help to build up an appreciation of the actual noise levels and sources of high level noise in the ward environments. The second part of the research will involve the completion of a questionnaire to explore your perceptions of the ward environment in relation to noise. The questionnaire should take between 5 and 10 minutes to complete. Of course, it is up to you to decide whether or not to take part. If you do so, you will be given this information sheet to keep. You are still free to withdraw at any time and without giving a reason. No personal information will be asked of you in the questionnaire and all information you do provide will be handled in a confidential manner and stored in a locked filing cabinet and on a password protected computer in an environment locked when not occupied. Only the researcher and supervisor will have direct access to the information. Any reference to you will be coded. This information will be held until the end 2012. If you have a concern about any aspect of this study, you should ask to speak with the researcher who will do their best to answer your questions. The contact details of the researcher are shown at the end of this sheet. If you would like any further
258
information regarding this study or have any complaints about the way you have been dealt with during the study or other concerns you can contact: Rosemary Glanville, Head of the Medical Architecture Research Unit on 0207 815 8329, who is the Academic Supervisor for this study. Finally, if you remain unhappy and wish to complain formally, you can do this through the Universitys Complaints Procedure. Details can be obtained from the university website: http://www.lsbu.ac.uk/research Researchers Contact Details: Mrs Nicola Shiers Medical Architecture Research Unit Faculty of Engineering, Science and the Built Environment London South Bank University 103 Borough Road London SE1 0AA T: 0207 815 8395 E: shiersn@lsbu.ac.uk
259
3.
Sound IN YOUR HOSPITAL

This questionnaire forms part of a study being carried out by the Estates and Facilities Team here at XXXX Hospital together with London South Bank University. The purpose of research is to understand the impact of noise on staff and patients in the wards. The questionnaire should take between 5 and 10 minutes to complete and is of course completely voluntary. By completing this questionnaire you are consenting to take part in this study. The responses which you give will be completely anonymous. The questions concern how annoyed you are by noise and how much noise interferes with your ability to do your work.
About You
1. Are you? Male Female 2. What age are you? Less than 20 41-50 years 20-30 years 51-60 years 31-40 years 60+ years
3. What staff grade are you? 4. How long have you worked on this ward? Less than 1 year 3-4 years 1-2 years 4-5 years 2-3 years 5+ years
5. How long have you worked at this hospital? Less than 1 year 3-4 years 1-2 years 4-5 years
260
2-3 years 5+ years
About Your Environment

1. Do you ever feel annoyed by noise while you are at work? Not at all Slightly Moderately Very much Extremely
Please indicate on a scale from 0 to 4 how much you are annoyed by each of the following noises (0 indicating not at all and 4 indicating a great deal) 0 External noise (traffic, aircraft etc) Doors banging Internal telephones ringing Staff talking on the telephone Nurse call Door bell Footsteps Medical equipment alarms People talking Cleaning Rubbish Bins Trolleys Meal times Television / radio Mobile phones ringing People talking on mobile phones Visiting time Other (please specify and rate) .. .. .. .. 1 2 3 4
261
2. Overall, how much do you feel that noise interferes with your ability to do your work? Not at all Slightly Moderately Very much Extremely
Please indicate on a scale from 0 to 4 how much each of the following noises interferes with your ability to do your work (0 indicating not at all and 4 indicating a great deal). 0 External noise (traffic, aircraft etc) Doors banging Internal telephones ringing Staff talking on the telephone Nurse call Door bell Footsteps Medical equipment alarms General conversation Cleaning Rubbish Bins Trolleys Meal times Television / radio Mobile phones ringing People talking on mobile phones Visiting time Other (please specify and rate) .. .. .. .. 1 2 3 4
262
3. It is important for you to be able to hear some sounds in order for you to carry out your job effectively. Please rate the importance of each of the following sounds on a scale from 0 to 4, where 0 indicates not at all important and 4 extremely important. 0 Nurse call Conversations with colleagues Conversations with patients Equipment alarms Patients calling out Patient activity Other (please specify and rate) .. .. 4. Sometimes high levels of background noise can make it difficult to hear important sounds. Please indicate on a scale from 0 to 4 how difficult it is to hear in the following locations:
I can always hear It is very difficult to hear
0 Nursing Station Hallway 4 bed bay Single patient room Treatment Room
If you have any further comments regarding noise, please write them in the space below.

THANK YOU FOR TAKING THE TIME TO COMPLETE THE QUESTIONNAIRE

263
4.
Sound IN YOUR HOSPITAL

This questionnaire forms part of a study being carried out by the Estates and Facilities Team here at XXXX Hospital together with London South Bank University. The purpose of research is to understand the impact of noise on patients and staff in the wards. We are also keen to find out whether the design and materials used in our buildings help to make the buildings less or more noisy for everyone who uses them.
The questionnaire should take between 5 and 10 minutes to complete and is of course completely voluntary. By completing this questionnaire you are consenting to take part in this study. The responses which you give will be completely anonymous.
About you
1. Are you? Male Female
2. What age are you? Less than 20 20-30 years 31-40 years 41-50 years 51-60 years 60+ years
3. How many days have you been at hospital this time? (Please enter number in boxes below)
days
Bed Number
264
About Your Environment

1. During the daytime in this ward, would you say it is: Very quiet Quiet A little noisy Very Noisy Extremely Noisy 2. Are you ever annoyed by noise during the daytime? YES NO If you answered YES, please indicate on a scale from 0 to 4 how much you are annoyed by each of the following noises (0 indicating not at all and 4 indicating a great deal) 0 External noise (traffic, aircraft etc) Doors banging Internal telephones ringing Staff talking on the telephone Nurse call Door bell Footsteps Medical equipment alarms People talking Cleaning Rubbish Bins Trolleys Visiting time Meal times Television / radio Mobile phones ringing People talking on mobile phones Other patients crying out Other (please specify and rate) .. .. .. .. 265 1 2 3 4
3. After lights are turned out and you are trying to sleep, would you say the ward is: Very quiet Quiet A little noisy Very Noisy Extremely Noisy Dont know 4. Are you ever disturbed by noise after lights out? YES NO If you answered YES, please indicate on a scale from 0 to 4 how much you are disturbed by each of the following noises (0 indicating not at all and 4 indicating a great deal) 0 External noise (traffic, aircraft etc) Doors banging Internal telephones ringing Staff talking on the telephone Nurse call Door bell Footsteps Medical equipment alarms People talking Rubbish bins Trolleys Television / radio Mobile phones ringing People talking on mobile phones Other patients crying out Other (please specify and rate) .. .. .. .. 1 2 3 4
266
5. Are there any noises that you actually finding comforting? Please specify .. ..................
6. When doctors and nurses talk to you, can you always hear them clearly? I can always clearly hear what people say Occasionally high levels of noise make it hard to hear Often high levels of noise make it hard to hear
7. Do you consider that it is possible to hold a private conversation here? YES NO 8. If you answered yes to question 7 and wanted to hold a private conversation, would you: Use your normal voice Lower your voice Take some other precautionary measure If so please specify 9. Do you ever feel that there is too little sound in here? YES NO 10. Do you suffer from any hearing impairment that you know of? YES NO If you have any further comments regarding noise, please write them in the space below.
THANK YOU FOR TAKING THE TIME TO COMPLETE THE QUESTIONNAIRE

267
5.
Copy of email response from the National Research Ethics Service for the NHS (NRES)
RE: Research Enquiry NRES Queries Line [nres.queries@npsa.nhs.uk] To: Cc: Shiers, Nicola
Your query was reviewed by our Queries Line Advisers. We would classify this as a type of service evaluation and it does not require REC review.
Regards
Queries Line National Research Ethics Service National Patient Safety Agency 4-8 Maple Street London W1T 5HD
Website: www.nres.npsa.nhs.uk Email: queries@nres.npsa.nhs.uk
Ref: 04 /31
268
6.
Ethical approval received from London South Bank University (scan of relevant signatures)
269
14. Appendix B
1. 2. 3. 4. 5. 6. Great Ormond Street Hospital parent / patient questionnaire comments Bedford Hospital staff questionnaire comments Bedford Hospital patient questionnaire comments Addenbrookes Hospital staff questionnaire comments Addenbrookes Hospital patient questionnaire comments Comforting sounds
270
1.
Great Ormond Street Hospital parent / patient questionnaire comments

Comments made are listed below: General level is ok but sometimes conflicting TV/radio can be annoying. Would free earphones be appropriate? Noise of a working area is never a problem. Sometimes silence is nice. Children communicating and other carers is very enjoyable and helpful. Headphones for children after lights out? Some TVs can be very loud. Would be helpful if the entrance door to the ward could be closed at night time. This would eliminate a lot of the noise from the corridor and the other wards. Some noise depends on where your childs bed is situated, so because we were away from the nurses station the phone and conversation did not affect us, but the doorbell was quite noisy. Patients have TVs very loud and have them on during the night. Each patient should be issued with ear phones so that TV/radio noise does not disturb others in the wards. The drilling has been a mild nuisance and the fire alarm going off for 35 minutes at 5.20am made sleep impossible. The utility room opposite our room is used frequently during the day and night and the door makes a loud bang every time which kept us both awake. Quiet closers on the doors would help. It is quite difficult to determine what is too noisy as it is so dependent on what situation your child is in at the time. We have spent the last three days with our daughter recovering from spinal surgery. We are in a 4-bed bay with two other patients and next to a child who has special needs and benefits from loud music. She is also very vocal and both we and our daughter found this intrusive when she was feeling very unwell. This was not the fault of the other family, but was not conducive to our well being. Suggestions: Provide wireless headphone (perhaps two sets per bed to allow a visitor to listen as well as the patient). Limit the volume on each TV. Put speakers closer to the patient when 3 ceiling mounted TVs are set to different channels it is a real cacophony of sound. To illustrate how difficult it was at times, our daughter was lying with her fingers in her ears. 271
2.
Bedford Hospital staff questionnaire comments

I wish the curtains between patients beds could be sound proofed so you could have a confidential / private conversation without the rest of the bay hearing. I am partially deaf so find it difficult to hear when there is a lot of background noise. I often have to ask other staff members to repeat themselves or to clarify what they have said. I dont think there is too much of a problem on the ward regarding noise. Did you know? There are crickets living in the elevators at Bedford hospital, chirping at night.
3.
Bedford Hospital patient questionnaire comments

Medical ward patients: It appears that if a patient is noisy and aggressive their attentions are responded to quicker, that is bed pan, bathing etc. Quiet patients are often ignored or completely forgotten as the staff are over pushed and failing in essential human contact in most of their functions. Perhaps quiet areas could be assigned and less aggressive patients attended to quicker and more efficiently rather than excluded by aggressive nutters who dominate the ward by vocal, noisy and constant requests for un-needed help . We only had one night of noise it should be noted that this was over the course of an 8 day hospital stay. Kitchen door left open after meal times so loud banging / washing up can be heard. Cleaning staff shouting at each other from one end of the room to kitchen. Staff making personal calls on mobiles. I dont really want to know whom they met in the pub last night and whether they will go out for a curry or Chinese. Im very grateful for the kindness of all the staff from kitchen staff to doctors. Only sometimes noisy when a large family get around the bed, for example 6 or 7 people. Angry patients mouthing at everyone and not taking full responsibility for their own misgivings. I accept that there is bound to be a certain level of noise day or night, and understand the difficulties that staff have to do their job and at the same time try not to disturb the patients. 272
Surgical ward patients When the background noise is low as at night, individual noises are more disturbing The bay is very close to the entrance door and the reception desk. There is continual traffic past bay 1 to the other bays. Perhaps the desk should be located before the entrance door. I think staff can be very inconsiderate during the night laughing out loud slamming doors to us who are not necessarily worn out they sound like clanging cymbals. It is not easy to sleep when it is very light. As a poor normal light sleeper I find the lights left on during the night really hard. Have to use alarm to get light switched off. The ward is the quietest ward I have been on. Wherever you are these days you expect to hear a certain amount of noise such as traffic, and you take it for granted. One wouldnt want to live in a totally silent world, missing bird song and human voices. Thank you. The buzzer [nurse call] in the nurse station is a little disturbing. (This patient was in the 4bed bay directly opposite). Visiting very unclear times: are they the same at weekends? & external noise very bad I found the noise at night more of a problem than my previous visit my bed is the closest to the nurse station so I get all the buzzer noises [nurse call] / phone calls etc and lots of walking up and down. Some shoes squeak more than others! May I suggest that trolleys are fitted with pneumatic tyres would cut out a lot of clatter. I know cleaning has to be done and appreciate the cleanliness, but could it be done a little quieter? Please thank the staff for all their help, assistance and very friendly attitude. Doors are ill-fitting and a constant annoyance. Other sources of noise annoyance mentioned cited by one patient who was in a single room outside building noises; water pipes / pumping sounds; truck outside moving bins/rubbish siren sounds when reversing; continual humming from electrical machine outside; police/ambulance sirens throughout the day; obviously staff not shown how to
273
move rubbish bins / bags quietly. The doors are the main problem banging, and staff agree too. I accepted noise as part and parcel of ward life. Patient staying in a single room and listed squeaky hinges on doors as annoying. Patient is disturbed by mobile phones ringing during the night and suggests that they should be put on silent.
4.
Addenbrookes Hospital staff questionnaire comments

Often construction noise can be heard on the ward. Obviously inevitable with planned expansion, but this disrupts patients rest and peace and quiet (D8) What may make noise felt more is the space (lack of) and narrowness of the ward (D8) When visiting hours were restricted and all visitors came at a certain time it became quite noisy within bays (D8) The most annoying noise I can think of is relatives / others pushing the doorbell; then if you dont answer straight away they ring again and again. It is a noise that is too loud and unnecessary (M4) Noise and excessive noise is everywhere in the ward environment none more irritating than the specific sounds use to catch attention, especially those of the telephone which not only catches attention but causes some discomfort and stress when other tasks take over. If these sounds were different soothing but alertive and not repetitive, this may help reduce stress (M4) There is no emergency alarm sound in the staff room, so if youre in there you cant hear an emergency call (N3)
5.
Addenbrookes Hospital patient questionnaire comments

Im glad research into this is being done as I greatly believe in the visual and acoustic environment has great impact on ones recovery. (D8) Noises at night often make it hard to sleep. (D8) Women in high heels walking in the corridors.(D8) I probably make most of it, so I apologise to other patients and the staff. Also I am not that bothered about other patients crying out now, you get used to it. (D8) 274
A set of earplugs should be issued to each person. (D8) One prevalent cause of noise is dictated by fashion and could not easily be changed. Todays version of clear, open, optimistic and especially young person in hospital employment is uninhibited and class-free. This is fine in many ways, but does leave a range of opportunities for serious and comforting discourse socially unrepresented and indeed of quite unfamiliar character. (D8) Pleased with treatment from staff. Concerned as a younger person should there be a separate bay for the dementia patients as fairly distressing at times when crying out. (D8) The thumping of the external building work could not only be heard, but also felt coming up through the floor and made conversations with visitors very difficult. (D8) When a patient is continually shouting and crying out in pain then I would suggest trying to isolate them from other patients. (D8) Noise levels at night are extremely high and have kept me awake for 2 nights. (N3) I feel that all staff respect patients privacy and deal with any problems that arise. Sometimes visiting hours can be a bit noisy but to be expected, unless doing it to annoy others. (N3) Patients talking on mobile phones and land lines late at night. Patients watching TV via their laptops. Staff unwilling and unable to ask patients to stop using their mobiles and be quiet. Foul language while using phones and in general conversation between visitors and patients. Sleep is very important for recovery and there should be a set time for lights out etc. Visiting times should be enforced! (N3) I have a hearing problem that hearing aids can sometimes make surrounding noise very disturbing, therefore I am not good at judging noise levels generally. (N3) Some of the staff wearing heels and the noise of the doorbell. No one answering the phone so it is always going off. (M4) The turning on of TVs / radios when on speaker is very annoying and inconsiderate, especially first thing in the morning. (M4) You have to give and take. (M4) Maybe a hospital volunteer or WI member could be on call to answer the phone. When people are sick you are going to get noise. 275
Foreign accents are especially difficult for the elderly.
6.
Comforting sounds Bedford Hospital

Traffic Life going on outside Nurse activity Nurses talking quietly Knowing someone is near Knowing staff are there Tea trolley (2 responses) Music on the radio (3 responses) Nurses' voices People's voices Silence Specific person's ring tone
Addenbrookes Hospital
Knowing staff are there (4 responses) Floor cleaner (sent patient to sleep) Birds outside Listening to music /radio (3 responses) Noise of trains on nearby railway (2 responses) Singing, laughter, good humour Nurses comforting patients (2 responses) Tea trolley and nurses pleasant good mornings (2 responses) Voices of wife and daughters Droning noises nebulizers, Hoovers etc Aircraft
276
Appendix C
Corner corrections
Throughout the study, care was taken in the positioning of the microphone and associated equipment was so as to minimise its impact on staff duties and patient care. There were also constraints regarding of the length of cable between the microphone and the environmental case housing the sound level meter (5m). In most instances these constraints meant that the microphone was situated close to the edge of room, often in the corner, with the microphone suspended from the ceiling on a 300mm bracket. Due consideration was given to the possible increase in sound pressure that may occur as a result of wall or corner reflections. A number of tests were carried out to investigate this, with simultaneous LAeq measurements made both in the centre of the room and at the main equipment location, that is close to the corner or wall of the room. The results are shown in the table below. It can be seen that the average difference between the measured levels was low (0.61 dB), and thus it was felt that no correction to measured levels was required.
Position
Nurse station, D8 - in front of nurse station 02.06.10 Nurse station, D8 - microphone suspended in corner 12-bed bay, D8 - centre of the ward 9.6.10 12- bed bay, D8 - microphone suspended close to wall 4-bed bay, D8 - centre of the ward 23.06.10 4-bed bay, D8 - microphone suspended in corner of ward 6-bed bay, surgical ward, Bedford - centre of ward 24.06.10 6-bed bay, surgical ward, Bedford - microphone suspended in corner Nurse station, surgical ward, Bedford - in front of nurse station 24.06.10 Nurse station, surgical ward, Bedford - microphone suspended above nurse station 4-bed bay 1, surgical ward, Bedford - centre of the ward 01.07.10 4-bed bay 1, surgical ward, Bedford - microphone suspended in corner 7-bed bay, D8 - centre of the ward 01.07.10 7-bed bay, D8 - microphone suspended in front of wall Single room 3, surgical ward, Bedford - microphone in centre of room by bed end 07.07.10 Single room, surgical ward, Bedford - microphone suspended over the window Single room 1, surgical ward, Bedford - microphone in centre of room by bed end 14.07.10 Single room 1, surgical ward, Bedford - microphone on mini tripod on light over mirror 4-bed bay 2, surgical ward, Bedford - centre of the ward 21.07.10 4-bed bay 2, surgical ward, Bedford - microphone suspended in corner 3-bed bay, D8 - centre of the ward 21.07.10 3-bed bay, D8 - microphone suspended in centre of ward close to wall 15.36.04 1.18 15.37.22 59.7 11.51.55 3.02 11.54.57 44.2 60.7 1.0 11.34.10 3.02 11.37.12 53.2 44.0 -0.2 12.28.12 3.01 12.31.13 50.5 53.5 0.3 15.23.39 3.01 15.26.40 61.4 50.3 -0.2 11.06.22 3.01 11.09.23 46.0 62.0 0.6 13.38.10 5.24 13.43.34 56.7 47.8 1.8 13.10.03 3.02 13.13.05 46.6 56.8 0.1 17.00.46 2.59 17.03.45 63.1 48.0 1.4 10.12.56 3.07 10.15.03 59.1 63.5 0.4 12.34.56 3.04 12.38.00 56.9 59.0 -0.1
Date
Start time Elapsed Time (m:ss) End Time
LAeq (dB) Difference

58.5 1.6
Average difference = 0.61 dB
277

PHD Thesis Nshiers

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

PHD Thesis Nshiers

Diunggah oleh

Hak Cipta:

Format Tersedia

ACOUSTIC DESIGN FOR INPATIENT FACILITIES IN HOSPITALS

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

2.6 2.7 Chapter 3 3.1 3.2

Conclusions ..... 29 The effects of noise on staff and patients . 30 i

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

4.4 Chapter 5 5.1 5.2

Chapter 6 6.1 6.2 6.3 6.4

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

6.9 6.10 6.11 6.12 6.13 6.14 6.15

6.18 6.19 6.20

Follow up discussions 84 Conclusions . 87 iii

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

Chapter 7 7.1 7.2 7.3 7.4

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

7.12 7.13 Chapter 8 8.1 8.2 8.3 8.4

8.5 8.6 Chapter 9 9.1 9.2 9.3

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

Further analysis of high level noise sources 186

9.11 9.12 9.13 Chapter 10 10.1 10.2 10.3

10.5 10.6 10.7 Chapter 11 11.1 11.2

Acoustic Design for Inpatient Facilities in Hospitals Table of Contents __________________________________________________________________________________________________________

11.5 11.6 Chapter 12 12.1 12.2

12.4 12.5 12.6 12.7 Chapter 13 13.1 13.2

References ... 253 vii

Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________

210 210 211

Acoustic Design for Inpatient Facilities in Hospitals List of Figures ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals List of Tables ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals List of Tables ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals Acknowledgements __________________________________________________________________________________________________________

Acoustic glossary and definitions

Sound pressure level (dB)

Sound absorption table

Acoustic Design for Inpatient Facilities in Hospitals Abstract ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals Introduction ___________________________________________________________________________________________________________________

1.2. Aims and objectives

Acoustic Design for Inpatient Facilities in Hospitals Introduction ___________________________________________________________________________________________________________________

Acoustic Design for Inpatient Facilities in Hospitals Introduction ___________________________________________________________________________________________________________________

Acoustic standards and guidance

2.2. UK design guidance

2.2.1. HTM 08-01

2.3. European healthcare design guidance

2.4. World Health Organisation guidelines

Night time (8 hours)

Day time and evenings (16 hours)

Interference with rest and recovery

2.5. Control of Infection

2.5.2. National Standards of Cleanliness for the NHS

2.5.3. HFN 30 Infection Control in the built environment

3. Previous research on hospital noise

3.2. Noise measurement studies

Limitations of noise measurement studies

Holmberg and Coon

State psychiatric hospital, US

36.5 hrs in total at different times of day

A&E departments at 4 hospitals in Phoenix, US

16.00 20.00 (4 hours)

Bailey and Timmons

Paediatric ICU in a UK teaching hospital