Clinical Rehabilitation

http://cre.sagepub.com Return to work after major trauma

Herman R Holtslag, Marcel W Post, Chris van der Werken and Eline Lindeman Clin Rehabil 2007; 21; 373 DOI: 10.1177/0269215507072084 The online version of this article can be found at: http://cre.sagepub.com/cgi/content/abstract/21/4/373

Published by:
http://www.sagepublications.com

Additional services and information for Clinical Rehabilitation can be found at: Email Alerts: http://cre.sagepub.com/cgi/alerts Subscriptions: http://cre.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.co.uk/journalsPermissions.nav Citations http://cre.sagepub.com/cgi/content/refs/21/4/373

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Clinical Rehabilitation 2007; 21: 373383

Return to work after major trauma

Herman R Holtslag University Medical Centre Utrecht, Department of Rehabilitation Medicine and Rudolf Magnus Institute of Neuroscience, Marcel W Post Rehabilitation centre De Hoogstraat, Chris van der Werken University Medical centre Utrecht, Department of Surgery and Eline Lindeman University Medical Centre Utrecht, Department of Rehabilitation Medicine, Utrecht, The Netherlands Received 19th June 2006; returned for revisions 2nd August 2006; revised manuscript accepted 3rd September 2006.

Objective: To quantify the prevalence of return to work after major trauma, and to investigate the determinants of postinjury work status. Design: Prospective cohort study. Setting: University Medical Centre Utrecht, a level 1 trauma centre in the Netherlands. Method: All severely injured (ISS 16) adult (age 16 ) trauma survivors admitted from January 1999 to December 2000 who were full-time employed at the time of the injury were selected for follow-up (n 214). Response rate was 93%. Outcome was assessed at a mean of 15 months (SD 1.5) after injury. Multivariate logistic regression analyses identied determinants at hospital discharge and at follow-up. Results: Following injury 58.4% of the patients (n 125) were able to return to full-time employment, 21.5% had a part-time job, and 20.1% did not return to work. Univariate analysis yielded the following signicant determinants of postinjury work status: age, comorbidity, injury severity score, brain injury, spinal cord injury, length of stay in an intensive care unit, hospital stay, discharge destination, percentage of permanent impairment (according to the fourth American Medical Association guide (AMA)), limitations in activities of daily living and cognitive complaints. Logistic regression analyses (23% explained variance) identied spinal cord injury, duration of hospital stay, discharge destination and age as determinants of return to work at hospital discharge. At follow-up, determinants of return to work included AMA, activities of daily living, cognitive complaints and being discharged home (51% explained variance). Conclusions: Around 60% of the patients returned to their pre-injury work status after major trauma. The return to work rate was only partly explained by disability at follow-up. Independent determinants of return to work differ with the time of assessment.

Introduction
Return to pre-injury work status (return to work) is the origin of rehabilitation medicine and the ultimate
Address for correspondence: Herman Holtslag, University Medical Centre Utrecht, Department of Rehabilitation Medicine, PO Box 85500, HP F00.810, 3508 GA Utrecht, The Netherlands. e-mail: h.r.holtslag@umcutrecht.nl 2007 SAGE Publications

expression of successful social reintegration after major trauma, because work increases a persons sense of self-worth and personal fullment.1 Active involvement in society (by having a job) and nancial independence (by having a paid job) seem closely related to health and well-being.2 Further, from an economic viewpoint, the societal costs of productivity loss due to disability might considerably exceed the direct costs of medical care.3,4
10.1177/0269215507072084

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

374

HR Holtslag et al. Patients and methods This prospective cohort study was conducted at the University Medical Centre Utrecht, one of the 10 level 1 trauma centres in the Netherlands, serving a catchment area with a population of 1.1 million. The Utrecht trauma care region is an urbanized area in the centre of the Netherlands with a population density of 813 inhabitants per square kilometre. The study included severely injured patients treated by emergency care from January 1999 to December 2000, with an Injury Severity Score (ISS)20 of at least 16. One investigator (HRH) performed all follow-up evaluations. The time of follow-up was beyond one year after trauma, because disabled workers in the Netherlands may become eligible for a permanent disability pension after one year of sick leave. Patients were asked to participate by written invitation. If they did not respond, the investigator contacted them, their relatives or general practitioners by telephone. After giving written informed consent, patients were sent several self-administered questionnaires and were invited to visit the outpatient clinic. If patients were unable to do so, the same investigator visited them at their homes or institutions. Patients were encouraged, if applicable, to ask their spouses to assist them in completing the questionnaires. The medical ethics review committee of our hospital approved the study protocol, and informed consent was obtained from all patients. Outcome assessment and determinants of return to work The primary outcome parameter was return to work, dened as complete or nearly complete ( 80%) return to former full-time paid employment. Return to work was dichotomized as complete return to work versus part-time or no return to work. We also dichotomized the type of work, into physically demanding jobs, that is, production work (blue collar), and physically nondemanding jobs, that is, administrative or managerial work (white collar). The sociodemographic factors we recorded included gender, age and educational level. The latter was divided into primary school and lower secondary vocational education, senior secondary vocational education, and higher education at research universities and universities of professional education. Age was dichotomized into under 35 years and 35 years or over, in accordance with the median age of our sample.

Most studies of return to work after major trauma have focused on specic injury types, such as spinal cord injury,4 traumatic brain injury5 and lower extremity injury.6 Large generic major trauma studies investigating return to work have been rare. One of the rst studies on return to work or productivity after major trauma was published in the late 1980s.7 Reported return to work rates after major trauma range from 50 to 90%.8,9 In part, this variation is explained by differences between studies in the instruments used, the timing of measurements, inclusion criteria (e.g. working age only, or including students and housewives), and denitions of return to work (e.g. whether or not including part-time work, working with modications, returning to a different position or employer). Studies with a mean follow-up of 12 years after major trauma found lower return to work rates (5060%)7,10,11 than studies with a mean follow-up of ve years or more (6075%).3,8,9,1214 Some studies included return to part-time work or return to school in the return to work outcome.7,9,12 Others selected more severely or less severely injured patients for their studies.9,13,15 Several studies have investigated determinants of return to work outcome after major trauma. Age was found to be an important determinant in several studies, as were educational level, pre-injury income and social support.3,7,12,13,15,16 Physical or psychological limitations and comorbidity negatively affected functional outcome after specic injury types (e.g. orthopaedic trauma).3,13,1719 Other relevant determinants were the injury severity score, duration of stay at the intensive care unit, length of hospitalization and discharge destination.10,1215 Finally, localization of the injury (e.g. brain, spinal cord or extremity) was found to be a signicant determinant in most studies. Surprisingly, only few studies10,13 included a wide range of possible determinants and used multivariate analyses to select the most relevant determinants. The aim of the present study was twofold. First, we attempted to quantify the prevalence of return to work in a large group of severely injured patients in the Netherlands who were employed at the time of their injury, as assessed at least one year after the injury. The second aim was to investigate the individual and combined inuences of sociodemographic factors, physical factors, injury-related factors, hospital factors and disability at follow-up as determinants of post-injury work status.

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Return to work after major trauma The physical factors were body mass index, physically demanding job and comorbidity. For the latter we used a list of 26 chronic diseases from the Health and Labour Questionnaire,21 a validated instrument, from which we deleted depression, as we only wanted to evaluate physical comorbidity. The injury-related factors were injury severity score (ISS), dichotomized into ISS below 25 versus ISS 25 and higher,17 and injury localization such as brain, chest or abdomen, spinal cord, lower or upper extremity. Brain injury was dened as an injury to the head and/or facial fracture(s), combined with loss of consciousness lasting longer than 15 min or a Glasgow Coma Scale rating lower than 15 or post-traumatic amnesia lasting longer than 24 hours. A patient could be included in more than one type of injury group. Hospital factors were stay at the intensive care unit, length of hospital stay and discharge destinations (home and rehabilitation centre). Intensive care unit stay was dichotomized into yes or no (median two days). The length of hospital stay was dichotomized at three weeks, because this cut-off point was between the mean (23 days) and median (16 days). We also tried using a cut-off point for hospital stay of two weeks, but this hardly inuenced the outcome. The disability factors measured at follow-up were divided into physical and cognitive factors. To measure physical disability, we used the percentage of permanent impairment (AMA range 0100%) based on the Guide to Evaluation of Permanent Impairment, fourth edition, of the American Medical Association.22 This method evaluates an injured patients medical impairment based on objective factors of disability, as opposed to subjective complaints. The guidelines provide a xed percentage of impairment for each structural or functional impairment an individual can have like, for example, lower leg amputation (28%) or spinal cord injury (75%). The Groningen Activity Restriction Scale (GARS)23,24 covers activities of daily living (ADL) and mobility: 11 questions measure grades of difculties a person may experience, with response options ranging from Yes, I can do this completely and independently, without any difculty (1) to No, I cannot do this without someones help (4). Patients with a maximum score of 11 points may be regarded as completely independent, whereas a score of 44 points implies complete dependence. The GARS has been used in several studies in the Netherlands and in the

375

multicentre, longitudinal European study on incapacitating diseases called EURIDISS.23,24 Cognitive disability was assessed by the Head Injury Symptom Checklist (HISC).25 This instrument lists 12 symptoms as part of the sequelae of brain injury: headaches, fatigue, dizziness, blurred vision, concentration, hypersensitivity to noise and light, irritability, temper, memory problems, anxiety and insomnia. Scores range from 0 (no problems) to 12. At follow-up, patients were asked whether they were currently experiencing any of these symptoms. Statistical analysis Univariate Statistical independence between all dichotomized determinants and the dichotomous outcome variable (return to work) was tested by Pearson chi-square test. If a signicant chi-square test showed a dependency, Nagelkerkes R-square was calculated. This provides an indication of the amount of explained variance in return to work. Multivariate Backward logistic regression analysis was used to evaluate the combined effect of statistically signicant determinants of return to work. The signicant univariate determinants were used in the multivariate models. Cramrs V was used to evaluate correlations between signicant determinants. If a correlation 0.70 was found between a pair of determinants, only one was entered in the multivariate analysis. Because of possible selection bias, we decided not to enter discharge to rehabilitation centre in the multivariate analysis. We evaluated two models, based on the two assessment time-points: one at hospital discharge (prospective) and one at follow-up (crosssectional). All data were analysed using Statistical Package for the Social Sciences (SPSS), version 12.

Results
At the time of injury, 214 patients were in full-time employment. They included 184 men, with a mean age of 35.1 years, and 30 women, with a mean age of 32.8 years (Table 1). The mean injury severity score

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

376

HR Holtslag et al.
Table 1 Patient and injury characteristics of 214 major trauma patients who were full-time employed before their injury Gender Male Female Age, mean in years (SD) 1625 years 2635 years 3645 years 4655 years 5665 years Educational level Primary school LBO MBO HBO Injury severity score, mean (SD) Cause of injury Trafc Work Sports Private/home Others Duration of ICU stay, mean (SD) Duration of hospital stay, mean (SD) Discharge destination Home Rehabilitation centre (inpatient) Another hospital Others Job at injury Physically demanding (blue collar) Physically non-demanding (white collar)

184 (86%) 30 (14%) 34.8 (11.6) 49 (23%) 79 (37%) 42 (20%) 32 (15%) 12 (5%) 29 92 49 42 (14%) (43%) (23%) (20%)

25.0 (11.1) 150 30 18 10 6 (70%) (14%) (8%) (5%) (3%)

6.4 days (12.3), 117 patients (55%) 23.3 days (21.7) 148 46 14 6 (69%) (22%) (6%) (3%)

137 (64%) 75 (36%)

LBO, lower secondary vocational education and training; MBO, secondary vocational education and training; HBO, higher education.

was 25, while 40.2% had a score of 25 or higher. Seventy per cent (n 150) of the accidents had happened in trafc. After hospital admission, more than half of the patients needed treatment in an intensive care unit (ICU). Their mean length of stay at the intensive care unit was 6.4 days. Mean length of hospital stay was 23 days. The discharge destination was home in about 70% (n 148) of cases, whereas over a quarter of the patients were enrolled in a clinical (n 48) or outpatient (n 7) rehabilitation programme. Almost two-thirds had been in a managerial or administrative job before their injury. Mean (SD) time of follow-up was 15 (1.6) months post injury, ranging from 12 to 18 months.

Return to work and univariate analyses A total of 58.4% of the patients were able to return to their former job full time, while 21.5% had a parttime job and 20.1% were unemployed at the time of follow-up. Many determinants were signicantly associated with return to work in the univariate analyses (Table 2). The only signicant sociodemographic factor was age, the only signicant physical factor was comorbidity and the most signicant injury-related factors were injury severity score and having a brain or spinal cord injury. All hospital factors were signicantly associated with return to work (intensive care unit stay, hospital stay and discharge destination), as were

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Return to work after major trauma

Table 2 Determinants of return to work after major trauma (n Full-time return to work Sociodemographic factors: Gender Male (n 184) Female (n 30) Age 35 (n 125) 35 (n 89) Education level Lower (n 121) Higher (n 91) Physical factors: Body mass index 25 (n 60) 25 (n 59) Physically demanding job PJ (n 75) PJ (n 137) Comorbidity C (n 126) C (n 88) Injury-related factors: Injury severity score 25 (n 128) 25 (n 86) Injury localization Brain BI (n 100) BI (n 114) Chest and abdomen CAI (n 112) CAI (n 102) Spinal cord SCI (n 200) SCI (n 14) Lower extremity LEI (n 92) LEI (n 122) Upper extremity UEI (n 170) UEI (n 44) Hospital factors: ICU stay ICU (n 97) ICU (n 117) Hospital stay HS (n 129) 21 days HS (n 85) Discharge home H (n 66) H (n 148) Discharge to rehab centre RC (n 168) RC (n 46) 214), pre-injury full-time ( 80%) workers Part-time or no return to work

377

P-value/Nagelkerke

106 (57.6%) 15 (50.0%) 79 (63.2%) 42 (47.2%) 70 (57.9%) 50 (54.9%)

78 (42.4%) 15 (50.0%) 46 (36.8%) 47 (52.8%) 51 (42.1%) 41 (45.1%)

0.391 0.003/0.054 0.409

37 (61.7%) 37 (62.7%) 42 (55.5%) 79 (57.3%) 80 (63.5%) 41 (46.6%)

23 (38.3%) 22 (37.3%) 33 (44.5%) 58 (42.7%) 46 (36.5%) 47 (53.4%)

0.252 0.507 0.001/0.079

80 (62.5%) 41 (47.7%) 62 (62.0%) 59 (51.8%) 62 (55.4%) 59 (57.8%) 118 (59.0%) 3 (21.4%) 55 (59.8%) 66 (54.1%) 102 (60.0%) 19 (43.2%)

48 (37.5%) 45 (52.3%) 38 (38.0%) 55 (48.2%) 50 (44.6%) 43 (42.2%) 82 (41.0%) 11 (78.6%) 37 (40.2%) 56 (45.9%) 68 (40.0%) 25 (56.8%)

0.038/0.027

0.047/0.025 0.989 0.016/0.071 0.635 0.063

62 (63.9%) 59 (50.4%) 89 (69.0%) 32 (37.6%) 21 (31.8%) 100 (67.6%) 111 (66.1%) 10 (21.7%)

35 (36.1%) 58 (49.6%) 40 (31.0%) 53 (62.4%) 45 (68.2%) 48 (32.4%) 57 (33.9%) 36 (78.3%)

0.019/0.035 0.181

0.001/0.166 0.001/0.225 (continued )

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

378

HR Holtslag et al.

Table 2 (Continued ) Full-time return to work Disability at follow-up: AMA AMA 0 (n 93) AMA 0 (n 121) GARS-ADL GA 11 (n 133) GA 11 (n 81) HISCwA HISC 0 (n 89) HISC 0 (n 125) Part-time or no return to work

P-value/Nagelkerke

80 (86.0%) 41 (33.9%) 102 (76.7%) 19 (23.5%) 64 (71.9%) 57 (45.6%)

13 (14.0%) 80 (66.1%) 31 (23.3%) 62 (76.5%) 25 (28.1%) 68 (54.4%)

0.001/0.294 0.001/0.359 0.001/0.120

ICU stay, stay at intensive care unit; AMA, percentage of permanent impairment according to the fourth American Medical Association guide; GARS-ADL, Groningen Activity Restriction Scale-Activities of Daily Living; HISCwA, Head Injury Symptom Checklist, without anxiety. Bold indicates that association is statistically signicant. All determinants were analysed by chi-square test. Nagelkerke explained variation.

all disability measures at follow-up (e.g. AMA, GARS-ADL and HISC). The 125 respondents working full time at follow-up had a mean (SD) of 4.1% (10.4) permanent impairment, while 80 of them had no (0%) permanent impairment at all. The 89 respondents who were unemployed or working part-time at follow-up had a mean (SD) of 25.5% (24.2) permanent impairment while 13 had no (0%) permanent impairment. The mean (SD) recovery time to denitive return to work was 27.9 (13.2) weeks; 98 of the 125 started working on a therapeutic basis after a mean (SD) time of 15.3 (8) weeks. Most of those in employment at follow-up (102/125) had a minimum score of 11 on the GARS-ADL scale. The 89 respondents who were unemployed or part-time workers at follow-up scored 16.3 on the GARS-ADL scale (SD 6.9). Cognitive complaints were reported by 121 patients, viz., 84% of those unemployed at follow-up, 67% of those working part-time at follow-up, and 48% of those working full-time at follow-up. Most complaints concerned fatigue, memory, concentration, temper and irritability. Multivariate analyses Table 3 shows the direction, strength (odds ratio (OR)) and signicance of the multivariate associations between the bivariately signicant determinants and return to work. Because of collinearity (Cramrs V being 0.784), it was not possible to include both dis-

charge home and discharge to a rehabilitation centre in the regression analysis. In model A, based on determinants available at hospital discharge, the strongest determinant was spinal cord injury (OR 4.3), followed by being in hospital for more than three weeks, discharge destination and age over 35 years. The overall percentage of return to work correctly predicted was 74.3, and the total amount of explained variance was 23.0% (Nagelkerke). In model B, based on the assessment at follow-up, the most signicant and strongest determinant was the percentage of permanent impairment (AMA, OR 7.5), followed by activities of daily living, discharge destination and having cognitive complaints. Injury severity was not signicant (P-value 0.051) but a determinant to be taken into account. The percentage return to work correctly predicted was 77.1, and the total amount of explained variance was 50.7% (Nagelkerke). The area under the receiver operating curve for AMA was 0.835 (95% condence interval (CI) 0.107 0.223), with an optimal cut-off point of 4.

Discussion
In this study, return to work rates and determinants of return to work of people working full time prior to sustaining major trauma were investigated at a mean time of 15 months. Just over half (58.4%) of the patients were able to return to their former work,

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Return to work after major trauma

379

Table 3 Determinants of post-injury employment (n 214), Backward-LR model. Model A at hospital discharge, model B at follow-up with functional measurement of disability factors Factors Odds ratio Sociodemographic: Age Physical: Comorbidity Injury-related: Injury severity score Injury localization Brain Spinal cord Hospital: ICU stay Hospital stay 21 days Discharge home Disability at follow-up: AMA GARS-ADL HISCwA 2.11 4.28* 1.0717.2 1.004.46 1.89* Model A 95% CI 1.033.46 Odds ratio Model B 95% CI

2.65** 0.41*

1.355.18 0.200.84

0.42* 7.51*** 3.69** 2.63*

0.190.95 3.2917.1 1.717.95 1.245.58

ICU stay, stay in intensive care unit; AMA, percentage permanent of impairment according to the fourth American Medical Association guide; GARS-ADL, Groningen Activity Restriction Scale-Activities of Daily Living; HISCwA, Head Injury Symptom Checklist, without Anxiety; 95% CI, 95% condence interval; Bold indicates that association is statistically signicant: *P 0.05, **P 0.01, ***P 0.001.

while 21.5% had a part-time job and 20.1% were unemployed at follow-up. Univariate analysis showed that determinants of post-injury work status were age, comorbidity, injury severity score, brain injury, spinal cord injury, intensive care unit stay, duration of hospital stay, discharge destination, AMA, activities of daily living and cognitive complaints. Determinants found in multivariate analyses at the time of hospital discharge were spinal cord injury, followed by duration of hospital stay, discharge destination and age. At follow-up, the most signicant and strongest independent determinant for return to work was permanent impairment (AMA), followed by limitations in activities of daily living (GARS-ADL), having cognitive complaints (HISC), being discharged home and injury severity. Strengths and limitations This study has added to the available knowledge by examining a large unselected group of consecutive severely injured patients referred to a level 1 trauma hospital in the Netherlands. Other strong features

were the use of multivariate regression analysis to examine the determinants of return to work in this specic group of patients and the high response rate (93%) at follow-up. A limitation of the study was that it included patients drawn from only one large trauma centre, which potentially limits the opportunities to generalize the results. However, the composition of our study population was similar in terms of age, gender and injury severity to those of other trauma studies performed in the Netherlands,8,10 in other European countries1214 and in the USA.3,7,11 Another limitation of our study was the lack of a functional measurement at the time of discharge from hospital. Return to work rates Comparison between our results and those of others is hampered by differences in the instruments used, the timing of measurements and the inclusion criteria, as well as different denitions of return to work. In addition, return to work rates could be inuenced by personal factors such as age, motivation and

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

380

H Holtslag et al. 58.4% complete return to work, or 79.9% when parttime return to work was included, is well within the range of ndings from these other studies. Therefore the 15-month follow-up period appears to have yielded fairly stable results. Determinants of return to work after major trauma The literature mentions many different prognostic determinants of return to work based on univariate analyses, but in our opinion multivariate analyses are more valuable, because of the collinearity between determinants. It has been suggested that return to work is impeded by having a signicant physical disability, psychosocial impairment, cognitive impairment or personality change.29 The determinants we found at different time-points are discussed below. The rst assessment moment in our study was discharge from hospital. At that time, the strongest independent determinant of return to work was spinal cord injury, followed by duration of hospital stay, discharge destination and age. Spinal cord and brain injury are plausible determinants of return to work.3,12,14 Duration of hospital stay has also been reported as a determinant, as has age.3,12,13,16 At follow-up, the strongest independent determinant of return to work in the logistic regression analysis in our study was the AMA score. There has been little research on the AMA score, and it is at least inconsistent for lower extremity injuries.30,31 Some studies did not nd a correlation between AMA score and patients ability to return to work,14,32 but these studies used highly selected groups of patients with multiple injuries and multiple organ failure,14 or people with severe head injury.32 We found a moderate correlation (Cramrs V 0.529) between the dichotomized AMA and GARS-ADL scores. We also found a high percentage (83.5%) of explained variation for AMA score as the only determinant for return to work with a cut-off point close to zero (4), which endorses our dichotomization. The second strongest independent determinant of return to work in our multivariate analysis was activities of daily living (GARS-ADL). The GARS has so far only been used to study rheumatoid arthritis patients33 and was a signicant determinant for work status in our univariate analysis. A third independent determinant was

job satisfaction and environmental factors such as cultural differences between countries, social security systems or insurances, average unemployment rates and disability pensions. Comparisons of disability pensions at European level are complicated by the fact that each country has its own system for dening the disabled population. All Dutch employees are insured against illness, disability and unemployment. One year after trauma, a permanent disability pension can be granted, irrespective of the circumstances and the question who is responsible, to a maximum of 70% of the last earned wages. The average unemployment rate for Europe was 9.1% in 1999, with the lowest rate in Luxembourg (2.4%) and one of the highest in Spain (12.5%). The 1999 unemployment rate in the Netherlands was 3.2%, more or less comparable to those in the UK (5.9%) and the USA (4.2%).26 In our study, follow-up assessment took place at 15 months (range 1218) post injury. The time at which permanent impairment and disability are assumed to be present is beyond 12 months, when patients have reached a steady state.10,27 Recovery might, however, take longer than a year in the case of certain types of injury or complications. There are also patients who cannot return to their pre-injury job and therefore need further retraining rst.28 Studies using the same follow-up periods as ours (1218 months)10,11 have found return to work rates of 64% and 74%. Two studies with follow-up periods of 2.5 years3 and 3 years12 differed considerably from ours, return to work rates of 55% and 81%, respectively. However, the former3 included only patients with hospital charges over US$100 000, implying a more severely injured group (mean ISS 33) with a longer hospital stay. The latter12 also included patients attending school, and excluded patients with solitary brain injury or spinal cord injury. When we performed similar analyses of our data after exclusion of spinal cord injury and solitary brain injury, we still found return to work rates of 60%, although literature reports lower rates for spinal cord and brain injury. Only a few long-term ( 5 years) studies on major trauma are available8,13,14 and only one extends beyond one decade.9 The return to work rates they found were 73%, 60%, 74%, 63% and 66%, respectively. The return to work rates reported in these studies show that longer follow-up terms did not yield higher return to work rates, and that our nding of

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Return to work after major trauma cognitive complaints. Brain injuries or cognitive complaints have indeed often been described as negative prognostic factors.7,12 The fourth independent determinant of return to work in our multivariate analyses was discharge home. This parameter has not been investigated before, perhaps because it can be regarded as an outcome parameter: if patients are discharged home it means that they have recovered sooner and more completely from their injuries than patients discharged to a rehabilitation facility, and will thus be able to return to work sooner and more fully.4 A nal independent determinant of return to work was the injury severity score, a nding that has also been reported by several other studies.7,10,12 Determinants of return to work after specic injury types Most studies on return to work after trauma have been performed among patients with specic injury types such as brain injury,5,29,34 spinal cord injury,28,35 extremity injury,6,36 multiple organ failure and longterm intensive care.14 The reported return to work rates vary tremendously, viz. between 25% and 80%: 3046% after severe brain injury,25,32 6088% after mild to moderate brain injury,5,25,34 40% after spinal cord injury4,37 and 70% after unilateral lower extremity fracture.38 In spinal cord injury studies, several patient factors have been associated with more favourable employment outcomes. The most prominent factors include younger age, higher level of education, more years with SCI, less severe injury and being Caucasian.4 Patients who had a physically more demanding job before the injury are less likely to return to work,37 while patients who move to a new employer need more time to resume work.35 One quarter of people with severe lower extremity injuries have not returned to work after one year.6,31 Factors inuencing return to work were examined in the study by MacKenzie et al.6 After accounting for the extent of impairment, they found that patient characteristics that correlated with higher return to work rates included younger age, higher educational level, higher income, the presence of strong social support and employment in a white-collar job that was not physically demanding. Receipt of disability compensation was highly negatively associated with return to work among people with lower extremity injuries.6,39

381

Clinical messages

After major trauma, 40% of people working full-time before the injury did not return to their former work status. Prediction of return to work by rehabilitation specialists at hospital discharge should be based on spinal cord injury, duration of hospital stay, discharge destination and age.

Recommendations The available literature and our results seem to indicate clearly that many major trauma patients are not able to return to their pre-injury level of productivity. Brain-damaged and spinal cord-injured patients in particular have difculties nding a suitable job. Rehabilitation specialists should give more attention to major trauma patients soon after admission, which is the best time to become involved and might be the moment to foster return to work.40 The time interval to partial and full return to work is not a xed period, a fact which should be considered in future research. Assessment of physical and cognitive disability at hospital discharge might improve the prediction of successful return to work at that time-point and thereby guide decisions about the most suitable follow-up care and reintegration guidance. Furthermore, we recommend including a questionnaire on mental and cognitive functions in future research. Questions about return to work should be implemented in a comprehensive instrument. Several generic subjective instruments measuring health-related work productivity have been very comprehensively described by Prasad et al.41 Consensus on instruments, time-points and a comprehensive set of determinants should be achieved in the near future. Competing interests None declared. Contributors HH is guarantor, initiator, data collector, designer, analyser, author. MP is designer, monitor, analyser, author. CW is designer, initiator, author. EL is monitor, author.

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

382

HR Holtslag et al.
considerable productivity loss at work among workers with high physical load jobs. J Clin Epidemiol 2005; 58: 51723. Tate DG. Workers disability and return to work. Am J Phys Med Rehabil 1992; 71: 9296. Bergeron E, Rossignol M, Osler T, Clas D, Lavoie A. Improving the TRISS methodology by restructuring age categories and adding comorbidities. J Trauma 2004; 56: 76067. Pollak AN, McCarthy ML, Bess RS, Agel J, Swiontkowski MF. Outcomes after treatment of highenergy tibial plafond fractures. J Bone Joint Surg Am 2003; 85-A: 1893900. McCarthy ML, MacKenzie EJ, Edwin D, Bosse MJ, Castillo RC, Starr A. Psychological distress associated with severe lower-limb injury. J Bone Joint Surg Am 2003; 85-A: 168997. Baker SP, ONeill B, Haddon W, Jr., Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 18796. van Roijen L, Essink-Bot ML, Koopmanschap MA, Bonsel G, Rutten FF. Labor and health status in economic evaluation of health care. The Health and Labor Questionnaire. Int J Technol Assess Health Care 1996; 12: 40515. American Medical Association. Guides to the evaluation of permanent impairment. AMA, 1996. Kempen GI, Miedema I, Ormel J, Molenaar W. The assessment of disability with the Groningen Activity Restriction Scale. Conceptual framework and psychometric properties. Soc Sci Med 1996; 43: 160110. Suurmeijer TP, Doeglas DM, Moum T et al. The Groningen Activity Restriction Scale for measuring disability: its utility in international comparisons. Am J Public Health 1994; 84: 127073. McLean A, Jr., Dikmen SS, Temkin NR. Psychosocial recovery after head injury. Arch Phys Med Rehabil 1993; 74: 104146. Eurostat. Employment rates in Europe. Eurostat, 2000. Accessed from http://epp.eurostat.ec.europa.eu Holbrook TL, Anderson JP, Sieber WJ, Browner D, Hoyt DB. Outcome after major trauma: 12-month and 18-month follow-up results from the Trauma Recovery Project. J Trauma 1999; 46: 76571. Krause JS. Years to employment after spinal cord injury. Arch Phys Med Rehabil 2003; 84: 128289. Wehman P, Targett P, West M, Kregel J. Productive work and employment for persons with traumatic brain injury: what have we learned after 20 years? J Head Trauma Rehabil 2005; 20: 11527. McCarthy ML, McAndrew MP, MacKenzie EJ et al. Correlation between the measures of impairment, according to the modied system of the American Medical Association, and function. J Bone Joint Surg Am 1998; 80: 103442.

References
1 Oppermann JD. Interpreting the meaning individuals ascribe to returning to work after traumatic brain injury: a qualitative approach. Brain Inj 2004; 18: 94155. 2 Andries F, Wevers CW, Wintzen AR et al. Vocational perspectives and neuromuscular disorders. Int J Rehabil Res 1997; 20: 25573. 3 Morris JA, Jr., Sanchez AA, Bass SM, MacKenzie EJ. Trauma patients return to productivity. J Trauma 1991; 31: 82733. 4 Krause JS, Kewman D, DeVivo MJ et al. Employment after spinal cord injury: an analysis of cases from the Model Spinal Cord Injury Systems. Arch Phys Med Rehabil 1999; 80: 1492500. 5 Yasuda S, Wehman P, Targett P, Cifu D, West M. Return to work for persons with traumatic brain injury. Am J Phys Med Rehabil 2001; 80: 85264. 6 MacKenzie EJ, Morris JA, Jr., Jurkovich GJ et al. Return to work following injury: the role of economic, social, and job-related factors. Am J Public Health 1998; 88: 163037. 7 MacKenzie EJ, Shapiro S, Smith RT, Siegel JH, Moody M, Pitt A. Factors inuencing return to work following hospitalization for traumatic injury. Am J Public Health 1987; 77: 32934. 8 van der Sluis CK, Eisma WH, Groothoff JW, ten Duis HJ. Long-term physical, psychological and social consequences of severe injuries. Injury 1998; 29: 28185. 9 Redmill DA, McIlwee A, McNicholl B, Templeton C. Long term outcomes 12 years after major trauma. Injury 2006; 37: 24346. 10 Vles WJ, Steyerberg EW, Essink-Bot ML, van Beeck EF, Meeuwis JD, Leenen LP. Prevalence and determinants of disabilities and return to work after major trauma. J Trauma 2005; 58: 12635. 11 Michaels AJ, Michaels CE, Smith JS, Moon CH, Peterson C, Long WB. Outcome from injury: general health, work status, and satisfaction 12 months after trauma. J Trauma 2000; 48: 84148. 12 Anke AG, Stanghelle JK, Finset A, Roaldsen KS, Pillgram-Larsen J, Fugl-Meyer AR. Long-term prevalence of impairments and disabilities after multiple trauma. J Trauma 1997; 42: 5461. 13 Kivioja AH, Myllynen PJ, Rokkanen PU. Is the treatment of the most severe multiply injured patients worth the effort? A follow-up examination 5 to 20 years after severe multiple injury. J Trauma 1990; 30: 48083. 14 Grotz M, Hohensee A, Remmers D, Wagner TO, Regel G. Rehabilitation results of patients with multiple injuries and multiple organ failure and long-term intensive care. J Trauma 1997; 42: 91926. 15 Meerding WJ, IJzelenberg W, Koopmanschap MA, Severens JL, Burdorf A. Health problems lead to 16 17

18

19

20

21

22 23

24

25 26 27

28 29

30

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010

Return to work after major trauma

31 Holtslag HR, Buskens E, Rommers GM, Prevo AJH, van der Werken C. Long-term outcome after lower extremity injuries in severely injured patients. Eur J Trauma 2006; 32: 36573. 32 Brooks N, McKinlay W, Symington C, Beattie A, Campsie L. Return to work within the rst seven years of severe head injury. Brain Inj 1987; 1: 519. 33 Doeglas D, Suurmeijer T, Krol B, Sanderman R, van Leeuwen M, van Rijswijk M. Work disability in early rheumatoid arthritis. Ann Rheum Dis 1995; 54: 45560. 34 van der Naalt J, van Zomeren AH, Sluiter WJ, Minderhoud JM. One year outcome in mild to moderate head injury: the predictive value of acute injury characteristics related to complaints and return to work. J Neurol Neurosurg Psychiatry 1999; 66: 20713. 35 Schonherr MC, Groothoff JW, Mulder GA, Eisma WH. Vocational perspectives after spinal cord injury. Clin Rehabil 2005; 19: 200208.

383

36 Seekamp A, Regel G, Tscherne H. Rehabilitation and reintegration of multiply injured patients: an outcome study with special reference to multiple lower limb fractures. Injury 1996; 27: 13338. 37 Tomassen PC, Post MW, van Asbeck FW. Return to work after spinal cord injury. Spinal Cord 2000; 38: 5155. 38 Parziale JR. Disability evaluation of extremity fractures. Phys Med Rehabil Clin N Am 2001; 12: 64757. 39 Lysgaard AP, Fonager K, Nielsen CV. Effect of nancial compensation on vocational rehabilitation. J Rehabil Med 2005; 37: 38891. 40 British Society of Rehabilitation Medicine. Vocational rehabilitation. British Society of Rehabilitation Medicine, 2006. Accessed from www.bsrm.co.uk 41 Prasad M, Wahlqvist P, Shikiar R, Shih YC. A review of self-report instruments measuring health-related work productivity: a patient-reported outcomes perspective. Pharmacoeconomics 2004; 22: 22544.

Downloaded from http://cre.sagepub.com at Vilnius University Library on April 14, 2010