DOI 10.1007/s00420-008-0346-9
O R I G I N A L A R T I CL E
Received: 6 March 2007 / Accepted: 25 June 2008 / Published online: 9 August 2008
© Springer-Verlag 2008
123
428 Int Arch Occup Environ Health (2009) 82:427–434
123
Int Arch Occup Environ Health (2009) 82:427–434 429
documentation of the workplace in addition to the usual Patient history of clinical data coding was done in retro-
patient history and clinical Wndings at study inclusion. spect blinded to job or sick leave status. This was used to
Throughout the rehabilitation speciWc contacts were named evaluate the eVect strength on sick leave changes (pre vs.
at all cooperating institutions. post-treatment, patients vs. controls), duration of complaints
Outpatient rehabilitation lasted from 3 to 4 weeks with (onset to rehabilitation), age, body-mass-index as well as the
daily treatment of approximately 4–6 h per day. After the dichotomized factors smoking habits, regular sports,
Wrst 8 days of treatment the most reasonable way of work- recorded psychosocial burden as independent risk factors.
place reintegration was evaluated (Fig. 1), and occupational We evaluated the relative strength of each factor by mul-
physicians were informed accordingly. Regularly multidis- tiple linear regression retaining all variables with a signiW-
ciplinary team conferences were held with therapists and cant eVect for the F statistic. For statistical data analysis
Airbus health professionals to achieve optimal rehabilita- single missing values were substituted by variable mean
tion results and overcome communication diYculties. values to allow multiple linear regression analysis. Analysis
For all patients annual sick leave days were calculated was conducted with SPSS 15.0 (SPSS Inc, Chicago, USA).
by the appropriate health insurance fund. Sick leave data We consider data in boxplots as outliers if they do not Wt
were available for Airbus employees up to 5 years before within 1.5–3 box-lengths from the 75th percentile or 25th
treatment and up to 3 years following treatment. percentile, extreme values are cases with the values more
123
430 Int Arch Occup Environ Health (2009) 82:427–434
than 3 box-lengths from the 75th percentile or 25th percen- reports. The most common suggestion for ongoing treat-
tile. SigniWcance level was deWned as P = 0.05. ment was “training by the patient itself” (75/79) without
The study was approved by the Ethics Committee of the further speciWcation.
Luebeck University (Germany) and informed consent had
been obtained from all subjects. Sick leave duration
123
Int Arch Occup Environ Health (2009) 82:427–434 431
Table 3 Correlations between outcome and potential risk indicators for musculoskeletal disease
Sick leave Sick leave Changes
change before in sick
(days/year) rehabilitation leave (%)
123
432 Int Arch Occup Environ Health (2009) 82:427–434
Table 4 Excluded variables in the multiple linear regression model for risk factor inXuence on sick leave days
Excluded variable Beta In t Sig. Partial
correlation
Graded work rtw schedule (0 = no, 1 = yes) 0.024 0.226 0.822 0.026
Age (years) 0.033 0.336 0.738 0.039
BMI (kg/m2) ¡0.082 ¡0.832 0.408 ¡0.095
Smoking habits (0 = non smoker, 1 = smoker) ¡0.064 ¡0.609 0.544 ¡0.070
Psychosocial burden (0 = absent, 1 = present) ¡0.005 ¡0.049 0.961 ¡0.006
Regular physical activity (0 = no, 1 = yes) ¡0.179 ¡1.858 0.067 ¡0.208
Duration of complaints before rehabilitation (years) 0.174 1.805 0.075 0.203
Sick leave days change before and after intervention (mean days per year) were set as dependent variable, with risk factors as independent variable
tested for inXuence. Sick leave change pre/post-treatment (mean days/year), predictor in the model: sick leave pre-treatment
before intervention indicating that this pattern of work changes (optimized communication, workplace customiza-
return applied mostly for severely handicapped workers. tion of rehabilitation procedures) could easily be evaluated
All other factors tested including physical activity and for other workplace situations as well as the eVect strength
duration of complaints showed no signiWcant correlation to over longer time periods.
the sick leave days change (Table 4). Another advantage over conventional rehabilitation
schedules was the increasing interest of the therapists in
occupational aspects concerning functional limitations and
Discussion work demands.
Limitations of our study are the small sample size, the
There is strong evidence that duration of work disability is lack of a control group and the low number of persons
signiWcantly reduced by contact between healthcare pro- observed for 2 or more years after therapy. Most data were
vider and workplace, although the available evidence for collected by insurances or industrial medical services for
the sustainability of these eVects is limited (Bendix et al. routine purposes and had to be recoded for statistical analy-
1998; Franche et al. 2005; Steenstra et al. 2006). We have sis; this resulted in only a few parameters with complete
investigated the inXuence of optimized rehabilitation and reliable documentation. Therefore, some factors
schedules focused on speciWc work burden and optimal reported to be relevant had to be omitted from the statistical
communication between employer and healthcare provider; analysis, e.g. history of pain, individual (pain or work)
our data suggest a major reduction of sick leave days for at beliefs, behavioural changes, functional limitations, extent
least 2 years after rehabilitation and the disruption of a of physical activity, psychosocial stressors. Also, some
trend of increasing sick leave days over the years before clinical as well as radiological Wndings were not considered
intervention. The presence of this increase indicates the as relevant factors (Bigos et al. 1992; Boos et al. 2000;
necessity to improve the work ability in workers with back Savage et al. 1997; van Tulder 1997 and Waddell and
complaints. Burton 2001).
The sick leave reduction shown here exceeds the eVects Other known risk factors for the incidence of musculo-
of non-invasive therapy regimen in musculoskeletal dis- skeletal diseases are age, sex, body mass index (Bernard
eases (van Tulder et al. 2006; Weiler et al. 2006; Heymans 1997) or smoking (Cassou et al. 2002); however, the pre-
et al. 2007); in addition, we had a 97% return-to-work ratio dictive value of these factors was low in our cohort. In
1 year post-treatment, whereas Anema et al. (2004) and patient records from both occupational and family physi-
Heymans et al. (2007) reported a 70–75% rtw rate as nor- cians most patient history and physical examination data
mal. Our results suggest that sick leave duration before are incompletely documented, especially for healthy work-
rehabilitation is more predictive than the reintegration ers; thereby these data usually are of poor quality. In Ger-
model with or without graded work, but this result may be many occupational medical examination is voluntary as
biased by the fact that workers with high sick leave prefera- long as no speciWc health threat exists, even if work-related
bly got a graded work rtw-schedule. related complaints are common and well known.
The changes implemented in this rehabilitation proce- From our data average sick leave days should be
dure are rather simple and consist of detailed workplace regarded as predictors for the necessity of rehabilitation
descriptions handed out to the therapists. A brief report procedures; and are more likely to be reduced signiWcant
after 8 days of treatment enabled the possibility of early the higher the initial number of sick leave days was. Both
workside rtw-action; therefore the eVectivity of these political opinion and insurance companies emphasize fac-
123
Int Arch Occup Environ Health (2009) 82:427–434 433
tual data, suggesting that data from self report question- 3–4 months. Occup Environ Med 61:289–294. doi:10.1136/
naires are weak and irrelevant; on the other hand even oem.2002.006460
Bendix AE, Bendix T, Haestrup C, Busch E (1998) A prospective, ran-
return-to-work and sick leave data can only be compared domized 5-year follow-up study of functional restoration in
with caution due to diVerences in data collection and evalu- chronic low back pain patients. Eur Spine J 7:111–119.
ation (van der Giezen et al. 2000). doi:10.1007/s005860050040
Known barriers to work return for patients with muscu- Bernard BP (ed) (1997) Musculoskeletal disorders and workplace fac-
tors: a critical review of epidemiologic evidence for work-related
loskeletal diseases not addressed in this study are chronic musculoskeletal disorders of the neck, upper extremity, and low
pain (Straaton et al. 1998; Buchner et al. 2007), psychoso- back. National Instiute for Occupational Safety and Health. http:/
cial factors, the patients pain beliefs and pain attitudes /www.cdc.gov/niosh/docs/97-141/ (21.05.2008)
(Main and Williams 2002; Marhold et al. 2002; Swinkels- Bigos SJ, Spengler DM, Martin NA, Zeh J, Fisher L, Nachemson A
(1986) Back injuries in industry: a retrospective study. III. Em-
Meewisse et al. 2006) and speciWc work requirements like ployee related factors. Spine 11:52–56. doi:10.1097/00007632-
repetitive work strain. Those factors should be considered 198601000-00014
as relevant in further studies. We suggest the use of out- Bigos SJ, Battie MC, Spengler DM, Fisher LD, Fordyce WE, Hansson
come measurements reported by Mannion et al. (2005) and T et al (1992) A longitudinal, prospective study of industrial back
injury reporting. Clin Orthop Relat Res 279:21–34
Elfering (2006) based on the core factor set as deWned by BLS Bureau of Labour Statistics of the U.S. (2006) Department of La-
Pincus et al. (2008). They ought to focus on factors possi- bour Nonfatal Occupational Injuries and Illnesses Requiring Days
bly inXuencing a lasting rehabilitation eVect. Other health Away From Work. http://www.bls.gov/news.release/osh2.nr0.htm
related questionnaires can be limited because short self (21.05.2008)
Boos N, Semmer N, Elfering A, Schade V, Gal I, Zanetti M et al (2000)
administered instruments have the same sensitivity to pre- Natural history of individuals with asymptomatic disc abnormal-
dict further sick leave as longer ones (Lindberg et al. 2008); ities in magnetic resonance imaging. Spine 25:1484–1492.
however our results indicate that questions for preceding doi:10.1097/00007632-200006150-00006
sick leave should be included. Boshuizen HC, Bongers PM, Hulshof CTJ (1992) Self-reported back
pain in fork-lift truck and freight-container tractor drivers ex-
posed to whole-body vibration. Spine 17:59–65. doi:10.1097/
00007632-199201000-00010
Conclusions Buchner M, Neubauer E, Zahlten-Hinguranage A, Schiltenwolf M
(2007) The inXuence of the grade of chronicity on the outcome of
multidisciplinary therapy for chronic low back pain. Spine
Our data support the notion that occupational physicians 32:3060–3066
should improve contacts both to the workplace and the Burton AK, Tillotson KM, Troup JDG (1989) Prediction of low-back
medical provider; improving communication and individu- trouble frequency in a working population. Spine 14:939–946.
alizing the rehabilitation leads to sustainable reduction of doi:10.1097/00007632-198909000-00005
Cassou B, Derriennic F, Monfort C, Norton J, Touranchet A (2002)
sick leave time and a very high rate of successful workplace Chronic neck and shoulder pain, age, and working conditions:
reintegration. The concept has been useful for rehabilitation longitudinal results from a large random sample in France. Occup
purposes in this study, possibly because relevant pain Environ Med 59:537–544. doi:10.1136/oem.59.8.537
related fear and self-conWdence were addressed in realistic Elfering A (2006) Work-related outcome assessment instruments. Eur
Spine J 15:S32–S43. doi:10.1007/s00586-005-1047-7
workplace simulations. Franche RL, Cullen K, Clarke J, Irvin E, Sinclair S, Frank J (2005) The
Institute for Work and Health (IWH) Workplace-Based RTW
Acknowledgments We acknowledge the support by the “Techniker Intervention Literature Review Research Team: workplace-based
Krankenkasse” (TK) and the German pension insurance (“Deutsche return-to-work interventions: a systematic review of the quantita-
Rentenversicherung Nord”) in sick leave days calculations and Wnanc- tive literature. J Occup Rehabil 15:607–631. doi:10.1007/s10926-
ing of the rehabilitation treatment, respectively. 005-8038-8
Fredriksson K, Alfredsson L, Ahlberg G, Josephson M, Kilbom Å,
ConXict of interest statement None. Wigaeus Hjelm E et al (2002) Work environment and neck and
shoulder pain: the inXuence of exposure time. Results from a pop-
ulation based case-control study. Occup Environ Med 59:182–
188. doi:10.1136/oem.59.3.182
References Frost P, Bonde JPE, Mikkelsen S, Andersen JH, Fallentin N,
Kaergaard A et al (2002) Risk of shoulder tendinitis in relation to
Andersen JH, Haahr JP, Frost P (2007) Risk factors for more severe re- shoulder loads in monoto-nous repetitive work. Am J Ind Med
gional musculoskeletal symptoms. Arthritis Rheum 56:1355– 41:11–18. doi:10.1002/ajim.10019
1364. doi:10.1002/art.22513 van der Giezen AM, Bouter LM, Nijhuis FJN (2000) Prediction of re-
Andersson GBJ (1981) Epidemiologic aspects on low-back pain in turn-to-work of low back pain patients sicklisted for 3–4 months.
industry. Spine 6:53–60. doi:10.1097/00007632-198101000- Pain 87:285–294. doi:10.1016/S0304-3959(00)00292-X
00013 Häkkinen M, Viikari-Juntura E, Martikainen R (2001) Job experience,
Anema JR, Cuelenaere B, van der Beek AJ, Knol DL, de Vet HCV, van work load, and risk of musculoskeletal disorders. Occup Environ
Mechelen W (2004) The eVectiveness of ergonomic interventions Med 58:129–135. doi:10.1136/oem.58.2.129
on return-to-work after low back pain; a prospective two year co- Heymans MW, Ford JJ, McMecken JM, Chan A, de Vet HCW, van
hort study in six countries on low back pain patients sicklisted for Mechelen W (2007) Exploring the contribution of patient-reported
123
434 Int Arch Occup Environ Health (2009) 82:427–434
and clinician based variables for the prediction of low back work spine and low back pain, age and oc-cupation in males. Eur Spine
status. J Occup Rehabil 17:383–397. doi:10.1007/s10926-007- J 6(2):106–114. doi:10.1007/BF01358742
9084-1 Schierhout GH, Meyers JE, Bridger RS (1995) Work related musculo-
Hoogendoorn WE, Bongers PM, de Vet HCW, Ariens GAM, van skeletal disorders and ergonomic stressors in the South African
Mechelen W, Bouter LM (2002) High physical work load and low workforce. Occup Environ Med 52:46–50
job satisfaction increase the risk of sickness absence due to low Schmidt CO, Kohlmann T (2005) What do we know about back pain
back pain: results of a prospective cohort study. Occup Environ symptoms? Epidemiological data regarding prevalence, inci-
Med 59:323–328. doi:10.1136/oem.59.5.323 dence, course, risk factors/Was wissen wir über das Symptom
Koes BW, van Tulder MW, Thomas S (2006) Diagnosis and treatment Rückenschmerz? Epidemiologische Ergebnisse zu Prävalenz,
of low back pain. BMJ 332:1430–1434. doi:10.1136/ Inzidenz, Verlauf, Risikofaktoren. Z Orthop Unfall 143:292–298.
bmj.332.7555.1430 doi:10.1055/s-2005-836631
Lindberg P, Josephson M, Alfredsson L, Vingard E (2008) Compari- Steenstra IA, Anema JR, Bongers PM, de Vet HCW, Knol DL, van
sons between Wve self-administered instruments predicting sick Mechelen W (2006) The eVectiveness of graded activity for low
leaves in a 4-year follow-up. Int Arch Occup Environ Health. back pain in occupational health-care. Occup Environ Med
(doi:10.1007/s00420-008-0326-0) 63:718–725. doi:10.1136/oem.2005.021675
Main CJ, Williams AC (2002) Musculoskeletal pain. BMJ 325:534– Straaton KV, Maisiak R, Wrigley JM, White MB, Johnson P, Fine PR
537. doi:10.1136/bmj.325.7363.534 (1998) Barriers to return to work among persons unemployed due
Mannion AF, Elferin A, Staerkle R, Junge A, Grob D, Semmer NK et al to arthritis and musculoskeletal disorders. Arthritis Rheum
(2005) Outcome assessment in low back pain: how low can you 39:101–109. doi:10.1002/art.1780390114
go? Eur Spine J 14:1014–1026. doi:10.1007/s00586-005-0911-9 Svensson HO, Andersson GBJ (1989) The relationship of low-back
Marhold C, Linton SJ, Melin L (2002) IdentiWcation of obstacles for pain, work history, work environment, and stress. Spine 14:517–
chronic pain patients to return to work: evaluation of a question- 522. doi:10.1097/00007632-198905000-00008
naire. J Occup Rehabil 12:65–75. doi:10.1023/A:1015056429505 Swinkels-Meewisse IE, Roelofs J, Verbeek AL, Oostendorp RA, Vla-
Nahit ES, Taylor S, Hunt IM, Silman AJ, Macfarlane GJ (2003) Pre- eyen JW (2006) Fear-avoidance beliefs, disability, and participa-
dicting the onset of forearm pain: a prospective study across 12 tion in workers and non-workers with acute low back pain. Clin J
occupational groups. Arthritis Rheum 49:519–525. doi:10.1002/ Pain 22:45–54. doi:10.1097/01.ajp. 0000148626.84874.93
art.11202 van Tulder MW, Assendelft WJJ, Koes B, Bouter L (1997) Spinal
Osborne RH, Nikpour M, Busija L, Sundararajan V, Wicks IP (2007) radiographic Wndings and nonspeciWc low back pain: a systematic
Prevalence and cost of musculoskeletal disorders: a population review of observational studies. Spine 22:427–434. doi:10.1097/
based, public hospital system healthcare consumption approach. J 00007632-199702150-00015
Rheumatol 34:2466–2475 van Tulder MW, Koes B, Malmivaara A (2006) Outcome of non-inva-
Pincus T, Burton AK, Vogel S, Field AP (2002) A systematic review sive treatment modalities on back pain: an evidence-based re-
of psychological factors as predictors of chronicity/disability in view. Eur Spine J 15:S64–S81. doi:10.1007/s00586-005-1048-6
prospective cohorts of low back pain. Spine 27:E109–E120. Waddell G, Burton AK (2001) Occupational health guidelines for the
doi:10.1097/00007632-200203010-00017 management of low back pain at work: evidence review. Occup
Pincus T, Santon R, Breen A, Burton A, Unterwood M (2008) A re- Med 51(2):124–135. doi:10.1093/occmed/51.2.124
view and proposal for a core set of factors for prospective cohorts Walker-Bone K, Cooper C (2005) Hard work never hurt anyone: or did
in low back pain: a consensus statement. Arthritis Rheum 59:14– it? A review of occupational associations with soft tissue muscu-
24. doi:10.1002/art.23251 loskeletal disorders of the neck and upper limb. Ann Rheum Dis
Punnett L, Gold J, Katz JN, Gore R, Wegman DH (2004) Ergonomic 64:1391–1396. doi:10.1136/ard.2003.020016
stressors and upper extremity muskuloskeletal disorders in auto- Weiler SW, Hartmann B, Josenhans J, Hanse J, Hauck A, von Bod-
mobile manufacturing: a one year follow up study. Occup Envi- mann J et al (2006) Rehabilitation of construction industry work-
ron Med 61:668–674. doi:10.1136/oem.2003.008979 ers considering special work-place demands–results of the
RKI (2006) Krankheitsfolgen In: Health in Germany/Gesundheit in “RehaBau” pilot study. Rehabilitation 45:309–313. doi:10.1055/
Deutschland. Robert Koch Institut, ISBN 3-89606-173-9, p 58 s-2006-940002
Savage RA, Whitehouse GH, Roberts N (1997) The relationship be-
tween the magnetic resonance imaging appearance of the lumbar
123