A R T I C L E I N F O A B S T R A C T
Keywords: Background: Despite the well-dened eects of sensory stimulation, the knowledge of the pure eects of af-
Traumatic brain injury fective stimulation is scarce.
Aect Objective: To evaluate the eects of family-centered aective stimulation on the level of consciousness among
Sensory comatose patients with brain injuries.
Stimulation
Design: This study was designed as a three-group double-blinded Randomized Controlled Trial.
Consciousness
Methods: Ninety consecutive patients with traumatic brain injuries and a Glasgow Coma Scale score of 58 were
Coma
selected. Patients were randomly allocated to an experimental, a placebo, and a control group using permuted
block randomization. Aective stimulation intervention was provided to patients in the experimental group by
their family members twice a day during the rst seven days of their hospitalization. In the placebo group, a
sensory stimulation program was implemented by a xed trained person who was not familiar with the patients.
Patients in the control group solely received sensory stimulation which was routinely provided to all patients.
The level of consciousness among the patients using the Glasgow Coma Scale and Coma Recovery Scale-Revised
were measured both before and after a family visit. The SPSS software (version 17.0) was used to analyze the
data through running the one-way and the repeated measure analyses of variance.
Results: Despite an insignicant dierence among the groups regarding baseline level of consciousness, the
results of the one-way analysis of variance revealed at the seventh day of this study, however, the level of
consciousness in the experimental group was signicantly higher (9.1 2.1) than the placebo (7.2 1.1), the
control groups (6.6 1.7) (P < 0.001), subsequently. Moreover, at the seventh day of the study, the Coma
Recovery Scale score in the experimental group (11.9 3.7) was signicantly greater than the placebo
(9.0 2.0) and the control (6.6 1.6) groups (P < 0.001). Recovery rate and eect size values also con-
rmed the greater eectiveness of aective stimulation compared with pure sensory stimulation.
Conclusion: Our ndings in this study indicate that early family-centered aective stimulation is more eective
than sensory stimulation in improving the level of consciousness among comatose patients with brain injuries.
Family-centered aective stimulation is recommended to be integrated into the nursing curricula and routine
care plans for comatose trauma patients in intensive care units.
What is already known about this topic? What this paper adds
Sensory stimulation is eective for improving the level of con- Family-centered aective stimulation is more eective than sensory
sciousness among comatose patients. stimulation in improving the level of consciousness among comatose
Although family involvement in care delivery is an accepted prac- patients.
tice in general hospital wards, it cannot be freely practiced in in- Family-centered aective stimulation is distinct and beyond con-
tensive care units due to the critical conditions of patients and the ventional sensory stimulation and can be integrated into routine
specialized care services provided in these units. care plans for comatose trauma patients in intensive care units.
Little is known about the eects of aective stimulation.
Corresponding author.
E-mail address: mohamade@modares.ac.ir (E. Mohammadi).
http://dx.doi.org/10.1016/j.ijnurstu.2017.05.014
Received 25 January 2017; Received in revised form 29 May 2017; Accepted 30 May 2017
0020-7489/ 2017 Elsevier Ltd. All rights reserved.
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
1.1. Theoretical framework The inclusion criteria were aiction by a traumatic brain injury, a
Glasgow Coma Scale (GCS) score of 58, and an age of 1865. Patients
Theoretically or physiologically; a distinction between sensory and who discharged from the ICU, died, or needed an emergency operation
family-centered aective stimulation is based on three aspects namely during the study were excluded. The sample size was calculated using
psychological, cognitive, and motivational. These aspects can accel- the results reported by Abbasi et al. (2009). They reported that the
erate cognitive processes, stimulate sympathetic nervous system, and means of GCS score in their groups were 6.8 1.4 and 7.8 0.70
later promote arousal (Harmon-Jones et al., 2013). Moreover, any type (Abbasi et al., 2009). So far, with a condence level of 0.95 and a power
of aective stimulation can aect the brain, particularly the reticular of 0.90, we concluded that 30 patients were needed for each study
activating system. Activation of this system increases sympathetic ac- group. Initially, an allocation protocol was developed using the per-
tivity throughout the neither body which in turn increases the levels of mutated block randomization technique (Pocock, 2013). Then, eligible
norepinephrine at nerve terminals and causes arousal and conscious- patients were consecutively recruited and randomly allocated to an
ness (Olausson et al., 2016). A critical point is about the possibility of experimental, a placebo, or a control group based on the permuted
45
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
Transfer (N=2)
Transfer (N=1)
46
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
Table 1 2004). It contains three components namely eye opening, verbal re-
Comparing the groups regarding the patients demographic characteristics and con- sponse, and motor response and is used to assess patients reactivity.
founding variables.
GCS score ranges from 3 to 15. Score 3 represents the lowest LOC and
Demographics Case Placebo Control p-value sever neurologic disorder while score 15 shows full consciousness and
normal reactivity. Despite its usefulness for LOC assessment, GCS
N % N % N % should be used in association with other neurologic assessment tools
(Gonzalez et al., 2004) because it has several limitations. For instance,
sex Men 22 73.3 24 80 22 73.3 0.78
Woman 8 26.7 8 20 8 26.7 it does not necessarily show the extent of traumatic brain injury, cannot
education Under- 4 13.3 4 13.3 4 13.3 0.68 assess pupillary response and minor changes in the power and mobility
graduate of the extremities, and cannot be used for patients who are intubated,
Diploma 18 60 21 70 17 56.7 receive medications that alter LOC, or suer from spinal cord injuries or
post 8 26 5 16.7 9 30
orthopedic traumas (Carlson and Louis, 2009). Thus, we used CRS-R in
graduate
Diagnosis Cerebral 16 53.4 18 60 18 60 0.89 addition to GCS for LOC assessment.
Hemorrhage
MIX 4 13.3 4 13.3 2 6.7
Contusion 6 20 6 20 9 30
2.4.2. CRS-R
*DAI 4 13.3 2 6.7 1 3.3 CRS-R is a 23-item scale the items of which are hierarchically ar-
Cause Car accident 10 33.3 10 33.3 11 36.7 0.83 ranged. It contains six subscales namely audition, vision, motor, com-
Motor 16 35.3 15 50 17 56.7 munication, speech, and arousal which are scored 04, 05, 06, 03,
accident
02, and 03, respectively. Scoring is performed based the presence or
Falling 4 13.3 5 16.7 2 6.7
Mean(SD) Mean(SD) Mean(SD) 0.78 absence of specic behavioral responses to standard sensory stimula-
Age 35.4(14.2) 34.5(14) 32.9(13.6) tion. Lower CRS-R scores show reexive activity while higher scores
**Apache 2 12.8(5.4) 12.5)5.4( 12.1(4.8) 0.85 denote cognitive behaviors (Giacino et al., 2004). The concurrent va-
GCS DAY 1 5.3(0.6) 5.1(0.3) 5.3(0.8) 0.29 lidity of the CRS-R was assessed using the Disability Rating Scale which
Score
CRS DAY1 5.5(0.9) 5.1(0.8) 5.2(1) 0.41
yielded a correlation coecient of 0.9. Moreover, the stability and in-
Score ternal consistency assessments of CRS-R revealed a test-retest correla-
tion coecient of 0.94 and a Cronbachs alpha of 0.84, respectively
* DAI: Diuse Axonal Injury. (Giacino et al., 2004). We assessed the reliability of the scale through
**Apache 2: Acute physiology and chronic health evaluation. the inter-rater reliability assessment method. Therefore, seventeen pa-
tients were recruited to a pilot study and an intervention was im-
block randomization protocol (Fig. 2). Allocation numbers were gen- plemented for them. After that, two raters concurrently assessed pa-
erated using a table of random numbers. The process of sampling took tients using CRS-R. The inter-rater correlation coecient was 0.90.
place from March 2015 to June 2016.
2.4.3. Apache II
2.3. Ethical considerations APACHE II is among the instruments which are used to determine
disease severity during the rst 24 h of ICU admission. It consists of
This study obtained an ethical approval from the Ethics Committee three main parts. The rst part contains twelve items on physiological
of Tarbiat Modares University, Tehran, Iran, on April 28, 2013. parameters and LOC. The possible score on each physiological para-
Moreover, it was registered in the Iranian Registry of Clinical Trials. meter is 04. Final LOC score is calculated through subtracting GCS
The approval and the registration codes were IR.TMU.REC.1394.22 and score from 15. The total score of the physiological domain of APACHE II
IRCT201503048069N3, respectively. The family members of all parti- is 059. The second part is a 06 score which is calculated based on the
cipants were informed about the study aim. All patients were free intended patients age group. The third part is related to chronic con-
whether to participate in or to withdraw from the attendance. ditions or organ failure. The scores of these three parts are summed to
Additionally, all associated data concerning the patients were re- obtain the total APACHE II score which ranges from 0 to 71. Higher
maining condential. APACHE II score shows severer health condition (Donahoe et al., 2009).
Table 2
Comparing the groups regarding patients GCS scores at seven measurement time points.
GCS GCS1 GCS2 GCS3 GCS4 GCS5 GCS6 GCS7 P-value For Within groups
Groups Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD)
Intervention 5.3(0.6) 5.4(0.8) 5.8(1.2) 7(1.4) 7.7(1.9) 8.8(1.9) 9.1(2.1) p > 0.001
Placebo 5.1(0.3) 5.1(0.3) 5.2(0.5) 5.9(0.6) 6(0.9) 6.9(1.1) 7.2(1.1) p > 0.001
Control 5.3(0.8) 5.3(0.8) 5.4(0.8) 5.7(1.1) 5.7(1.2) 5.9(1.3) 6.6(1.7) p > 0.001
P-value for Between groups p = 0.29 p = 0.24 p = 0.05 p > 0.001 p > 0.001 p > 0.001 p > 0.001
47
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
< 0.001
Control
Control
close family member with training about family-centered aective sti-
0.4
mulation. Training was about the rules and regulations of ICU, patients
condition, the necessity and importance of scheduled attendance at
GCS7 Group
Intervention
< 0.001 patients bedside, hand washing before entering and after leaving ICU,
< 0.001
how to establish the relationship with the patient, and how to provide
Placebo
Control
(fourteen times in total). The length of each session was 3045 min.
0.03
During the sessions, the in-charge nurse of the intended patient ac-
companied the family member in order to ensure the accurate im-
plementation of the intervention and to help and support the family
GCS6 Group
Intervention
member. At the rst session, either of the two trained nurses provided
> 0.001
> 0.001
Placebo
Control
0.513
Intervention
< 0.001
the patient for fteen minutes about happy daily events in the family,
Control
Control
Intervention
member massaged the patients hands and legs and performed passive
0.001
0.001
Control
0.663
GCS3 Group
Intervention
and interests.
Placebo
0.04
0.04
Control
0.933
0.430
tactile, and kinetic) stimulation twice a day for seven consecutive days
as follows: introduction (two minutes); providing information about
time and place (three minutes); talking about care services, treatments,
GCS2 Group
Intervention
and patients health status compared with previous days (ten minutes);
Placebo
0.251
Control
0.976
0.313
Intervention
all patients in the study setting. These services and stimulation included
0.424
0.424
windows.
P- value
P value
Patients LOC was assessed by the two trained nurses using the GCS
Table 3
and the CRS-R both before and after family visitation. Both nurses were
blind to the study aims and interventions.
48
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
Table 4
Comparing the groups regarding patients CRS-R scores at seven measurement time points.
CRS CRS 1 CRS 2 CRS 3 CRS 4 CRS 5 CRS 6 CRS 7 P-value For Within groups
Groups Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD) Mean(SD)
Intervention 5.4(0.8) 5.5(1.4) 6.2(1.8) 8(2.1) 9.1(2.5) 10.5(2.8) 11.9(3.7) p > 0.001
Placebo 5.1(0.8) 5.1(0.8) 5.3(0.9) 6.5(1.4) 6.8(1.6) 8.1(2.2) 9(2) p > 0.001
Control 5.2(1.3) 5.3(1.4) 5.5(1.5) 6(2.1) 6.1(2.1) 6.4(2.2) 6.6(1.6) p = 0.004
P-value for Between groups p = 0.56 p = 0.47 p = 0.07 p > 0.001 p > 0.001 p > 0.001 p > 0.001
2.6. Data analysis size. Results revealed that recovery rates in the experimental, placebo,
and control groups were 71%, 39%, and 24%. These ndings show that
Data were analyzed using the Statistical Package for Social Sciences recovery rate in the experimental group was higher than the placebo
(SPSS v. 17.0). Nominal and ordinal demographic variables were ana- and the control groups.
lyzed through the Chi-square test while interval and ratio demographic
variables were analyzed by running the one-way analyses of variance
(ANOVA). Moreover, the groups were compared with each other with 4. Discussion
regard to LOC scores via the one-way ANOVA with Tukeys post hoc
test. Besides, repeated measures ANOVA with Bonferroni post hoc test This study was designed to evaluate the eects of family-centered
were used to analyze the variations of LOC scores across the seven days aective stimulation on the LOC of comatose patients who had trau-
of the study intervention and P < 0.05 was signicant. matic brain injuries and were hospitalized in ICU. Findings indicated
the eectiveness of family-centered aective stimulation in improving
LOC among comatose patients from the third to the seventh day of the
3. Results intervention so much so that in these ve days, the dierences between
the groups with regard to patients LOC were statistically signicant.
The results of the Chi-square test and the one-way ANOVA showed Given the insignicant dierences between the groups regarding pa-
no signicant dierences between the groups with respect to the pa- tients age, gender, educational status, underlying conditions, cause of
tients gender, educational status, underlying conditions, causes of hospitalization, APACHE II score, and baseline GCS and CRS-R scores as
hospitalization, age, and baseline APACHE II, GCS, and CRS-R scores well as the presence of a control and a placebo group in the study,
(P > 0.05, Table 1). signicant dierences among the groups can be attributed to the pure
The results of the one-way ANOVA illustrated that the dierences eects of family-centered aective stimulation. Although most previous
among the groups regarding the GCS scores in the rst and the second studies reported the eectiveness of sensory stimulation in improving
days of the intervention were not statistically signicant while in days LOC, they were conducted using two-group designs (Davis and
37, the dierences were statistically signicant (Table 2). The results Gimenez, 2003; Hasanzadeh et al., 2012; Mahmoodi et al., 2013;
of the Tukeys post hoc test revealed that the dierences between the Mandeep, 2012; Oh and Seo, 2003; Abbasi et al., 2009; Bassampoor
experimental and the control groups as well as between the experi- et al., 2007; Gorji et al., 2014; Kalani et al., 2016; Karma and Rawat,
mental and the placebo groups regarding GCS scores were statistically 2006; Megha et al., 2013; Tavangar et al., 2016; Hoseini Azizi et al.,
signicant while the dierences between the placebo and the control 2016; Lombardi et al., 2002; Urbenjaphol et al., 2009; Goudarzi et al.,
groups were not statistically signicant (Table 3). The results of the 2010; Park and Davis, 2016; Puggina et al., 2011) and to evaluate just
repeated measures ANOVA also showed that the variations of GCS the eects of sensory stimulation. Thus, they failed to show the pure
scores in each study group were statistically signicant across the seven eects of aective stimulation. Taking together, our three-group
measurement time points. Moreover, a statistically signicant dier- double-blind randomized design is leading to the dierentiated study
ence was observed among the groups with respect to the mean GCS showing the eects of sensory and aective stimulation and indicated
scores across the seven measurement time points. In other words, LOC the greater eectiveness of aective stimulation in improving LOC
improvements in the experimental group (i.e. pure aective stimula- among comatose patients with traumatic brain injuries.
tion) were signicantly greater than the control and the placebo groups In contrary, some previous studies showed the positive eects of
(Table 2). sensory stimulation provided by family members on LOC while they
The results of the one-way ANOVA indicated that there were no failed to show the eects of sensory stimulation provided by unfamiliar
signicant dierences among the groups regarding CRS-R scores on persons on LOC.
days 13. However, the dierences between the groups were statisti- The results of the present study showed that after seven days of
cally signicant on days 47 (Table 4). The Tukeys post hoc test sensory and aective stimulation, the LOC of patients in the experi-
showed that in these days, CRS-R scores in the experimental group were mental group was signicantly higher than the patients in the placebo
signicantly greater than the control and the placebo groups while the and the control groups (P < 0.001). Moreover, the results of the re-
dierences between the placebo and the control groups were not sta- peated measures ANOVA for the within-subject factor of time in all
tistically signicant (Table 5). The repeated measures ANOVA also in- study groups were statistically signicant. In other words, the LOC in all
dicated that the variations of CRS-R scores in all study groups across the groups signicantly improved from the rst to the seventh days.
seven measurement time points were statistically signicant. Moreover, Therefore, LOC improvements in the experimental group were sig-
the results of this analysis showed a signicant dierence among the nicantly greater than the placebo and the control groups (P < 0.001).
groups regarding the variations of CRS-R scores across the seven mea- Consequently, recovery rates in the experimental, placebo, and control
surement time points. As shown in Table 4, improvements in the levels groups were 71%, 39%, and 24%, respectively. Moreover, eect sizes in
of CRS-R in the experimental group were signicantly greater than the the experimental and the placebo groups were 1.5 and 0.4, denoting the
placebo and the control groups. moderate and very large eect sizes of pure sensory stimulation and
Study ndings also showed that eect sizes in the experimental and family-centered aective stimulation, respectively. As mentioned in the
the placebo groups were 1.5 and 0.4, denoting moderate and very large theoretical framework section, any aective stimulation aects the
eect sizes, respectively. On the other hand, the recovery rate was also brain and particularly activates the reticular activating system, re-
calculated in order to provide more precise information about eect sulting in the increased sympathetic activity in dierent parts of the
49
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
< 0.001
Control
Control
0.081
consciousness and arousal (Khurana, 2006). Moreover, not only aec-
tive stimulation activates the nervous system, but also stimulates its
development. The human brain responds to kindly stimulation and
CRS 7 Group
Intervention
< 0.002 behaviors through releasing some hormones from the frontal cortex.
These hormones increase the overall activity of the brain and improve
Placebo
0.002
LOC (Eysenck and Flanagan, 2001).
In line with our ndings, Moattari et al. (2016) reported that re-
covery rate among patients who received sensory stimulation from fa-
> 0.001
mily members and nurses and patients who received no sensory sti-
Control
Control
0.028
mulation were 60%, 17%, and 7%, respectively. Likewise, in the former
study, eect sizes in the family- and nurse-centered sensory stimulation
groups were 1.2 and 0.2, respectively. Nonetheless, they considered
CRS 6 Group
Intervention
< 0.001
Placebo
Control
0.460
Intervention
< 0.001
Control
0.543
Intervention
Recovery rates in their experimental and control groups were 35% and
Placebo
16%, respectively while eect size was 0.4 (Davis and Gimenez, 2003).
0.009
0.009
Control
0.199
0.902
through both GCS and CRS-R. Our ndings revealed that the rst sig-
CRS 3 Group
Intervention
nicant dierences among the groups regarding GCS and CRS-R scores
Placebo
were in days 3 and 4, respectively. Most previous studies into the eects
0.083
0.083
Control
0.813
0.813
2013; Oh and Seo, 2003; Bassampoor et al., 2007; Gorji et al., 2014;
Kalani et al., 2016; Tavangar et al., 2016; Puggina et al., 2011). The
only study which used both GCS and CRS-R for evaluating the eects of
CRS 2 Group
Intervention
0.440
0.440
Control
0.729
0.991
Intervention
nal GSC score in our experimental group (i.e. 9.1 2.1), probably
Placebo
due to the fact that in Mandeeps study, sensory stimulation was pro-
0.651
0.651
P-value
50
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
51
F. Salmani et al. International Journal of Nursing Studies 74 (2017) 4452
Stimulation for Brain Injured Individuals in Coma or Vegetative State. The Cochrane Paradiso, M.A., Bear, M.F., Connors, B.W., 2016. Neuroscience: Exploring the Brain, 4 ed.
Library. Lippincott Williams & Wilkins, Hagerstwon, MD (p. 618.).
Lotze, M., Schertel, K., Birbaumer, N., Kotchoubey, B., 2011. A long-term intensive be- Park, S., Davis, A.E., 2016. Eectiveness of direct and non direct auditory stimulation on
havioural treatment study in patients with persistent vegetative state or minimally coma arousal after traumatic brain injury. Int. J. Nurs. Pract. 22, 391396.
conscious state. J. Rehabil. Med. 43 (3), 230236. Pessoa, L., 2005. To what extent are emotional visual stimuli processed without attention
Mahmoodi, G.R., HosseinZadeh, E., Vakili, M.A., KazemNejad, K., Mohammadi, M.R., and awareness? Curr. Opin. Neurobiol. 15 (2), 188196.
Taziki, M.H., et al., 2013. The eect of voice auditory stimulation on the con- Phelps, E.A., 2006. Emotion and cognition: insights from studies of the human amygdala.
sciousness of the coma patients suering from head injury. J. Res. Dev. Nurs. Annu. Rev. Psychol. 57, 2753.
Midwifery 10 (1), 19. Pocock, S.J., 2013. Clinical Trials: a Practical Approach. John Wiley & Sons.
Mandeep, P.K., 2012. Eectiveness of early intervention of coma arousal therapy in Puggina, A.C.u.G, da Silva, M.J.l.P., Santos, J.L.F., 2011. Use of music and voice stimulus
traumatic head injury patients. Int. J. Head Neck Surg. 3, 137142. on patients with disorders of consciousness. J. Neurosci. Nurs. 43 (1), 816.
Megha, M., Harpreet, S., Nayeem, Z., 2013. Eect of frequency of multimodal coma sti- Schnakers, C., Magee, W.L., Harris, B., 2016. Sensory stimulation and music therapy
mulation on the consciousness levels of traumatic brain injury comatose patients. programs for treating disorders of consciousness. Front. Psychol. 7, 16.
Brain Inj. 27 (5), 570577. Tavangar, H., Shahriary-Kalantary, M., Salimi, T., Jarahzadeh, M., Sarebanhassanabadi,
Mitchell, M., Chaboyer, W., Burmeister, E., Foster, M., 2009. Positive eects of a nursing M., 2016. Eect of family members voice on level of consciousness of comatose pa-
intervention on family-centered care in adult critical care. Am. J. Crit. Care 18 (6), tients admitted to the intensive care unit: a single-blind randomized controlled trial.
543552. Adv. Biomed. Res. 4, 106.
Moattari, M., Shirazi, F.A., Shari, N., Zareh, N., 2016. Eects of a sensory stimulation by Udekwu, P., Kromhout-Schiro, S., Vaslef, S., Baker, C., Oller, D., 2004. Glasgow Coma
nurses and families on level of cognitive function, and basic cognitive sensory re- Scale score, mortality, and functional outcome in head-injured patients. J. Trauma
covery of comatose patients with severe traumatic brain injury: a randomized control Acute Care Surg. 56 (5), 10841089.
trial. Trauma Mon. 21 (4), 110. Urbenjaphol, P., Jitpanya, C., Khaoropthum, S., 2009. Eects of the sensory stimulation
Oh, H., Seo, W., 2003. Sensory stimulation programme to improve recovery in comatose program on recovery in unconscious patients with traumatic brain injury. J.
patients. J. Clin. Nurs. 12 (3), 394404. Neurosci. Nurs. 41 (3), E10E16.
Olausson, H., Wessberg, J., McGlone, F., 2016. Aective Touch and the Neurophysiology Vuilleumier, P., 2005. How brains beware: neural mechanisms of emotional attention.
of CT Aerents. Springerpp. 421. Trends Cogn. Sci. 9 (12), 585594.
52