RETURN TO WORK (RTW) is an important stage in rehabilitation after traumatic brain injury (TBI) for a number of reasons. First, being employed has been associated with better quality of life in TBI survivors.1,2 Second, the financial costs associated with unemployment after TBI are substantial, given that TBI disproportionately affects young people of working age.3 In New Zealand, injury compensation is covered by Accident Compensation Corporation (ACC). In addition to medical and rehabilitation costs, ACC pays 80% of preinjury income as wage compensation while people rehabilitate from an injury. Delayed RTW (common in this population) is therefore very costly for both the individual and the society that funds the compensation scheme. According to Accident ACC injury statistics 2006, the cost of ongoing claims (ie, claims carried over from previous years) due to brain injury from July 2005 to June 2006 (excluding treatment-only claims and medical fees) was NZ$46 930 000-that is approximately NZ$32 800 per person with TBI.4 The United States records similarly enormous figures due to lost productivity and wages, which, together with costs associated with care and management of TBI, are estimated to be $22 billion annually.3
Estimates of long-term sustained RTW after TBI are low (generally around 30%), with some variation depending on the characteristics of the sample.5,6 Recent literature reviews have concluded that RTW rate improves with vocational intervention, and that even those with severe injuries can achieve vocational success with specialized interventions.7,8 In their recent quantitative synthesis, Kendall et al8 suggest that people who had vocational intervention showed a quicker RTW than those who did not. Furthermore, they concluded that the RTW rate for persons who have received intervention remains above the "natural course" RTW rate over many years.8 Important outstanding questions, therefore, concern which vocational interventions are the most effective for TBI survivors and how interventions differ in strengths, limitations, and application in TBI populations.
Over the last 25 years, various types of vocational rehabilitation approaches have emerged for people with TBI. In the last 10 years, several literature reviews have examined these approaches and the evidence that supports them. However, these have tended to be selective reviews that focus on a specific theory or approach, for example, supported employment, cognitive rehabilitation.3,9-13 Thus, it is still unclear how well recent evidence about vocational rehabilitation after TBI addresses the question of effectiveness and "best practice" (or which approach is best for whom). The present review examines the literature from 1990 to 2007 addressing approaches to vocational rehabilitation for people after TBI to assess these and to evaluate the evidence around the question of what is considered "best practice."
METHODS
Our literature search was based on the principles of systematic review.14 This involved defining topic variables and criteria for literature to be included, employing a systematic strategy for literature search and critique, and quality evaluation of the retrieved articles to draw conclusions based on the evidence. Given the heterogeneous nature of study designs, samples and outcomes and the general lack of empirical evidence regarding effectiveness of vocational rehabilitation programs, the resulting review should be considered descriptive. It was also considered that the potential value of a descriptive synthesis in this case (eg, for identifying strengths and limitations)15 warranted such an approach. This review attempts to
1. identify the main approaches to vocational rehabilitation after TBI described in the literature and the similarities and differences among these approaches;
2. assess the evidence for the effectiveness of these approaches; and
3. discuss the strengths, limitations, and applications of the various approaches.
DEFINITIONS
For this topic review, TBI was defined as injury to the brain that is caused by trauma, resulting in ongoing functional impairment as described in the first 3 domains of the International Classification of Functioning, Disability and Health (ICF) model: impairments of body functions, impairments of body structures, and impairments relating to activities and participation.16TBI vocational rehabilitation was defined as a rehabilitation approach designed to serve survivors of TBI, with the main goal of achieving a vocational outcome (ie, return to work, employment, or job retention).
DATA SOURCES
Data sources and key search words for this review (Table 1) were selected according to relevance to the topic of interest. A wide range of electronic databases was searched. These were the Ovid Journals database, which includes all journals on Ovid databases (MEDLINE, PsycINFO, CINAHL, AMED, Health and Psychosocial Instruments), Evidence-Based Medicine databases (Ovid), and Web of Science. A manual search through the bibliographies of relevant articles was also done to identify articles missed in database searches. Finally, manual searching for articles published from the year 2000 to July 2007 was also carried out for 4 particularly relevant journals: Brain Injury, Journal of Head Trauma Rehabilitation, Journal of Vocational Rehabilitation, and Journal of Occupational Rehabilitation.
DATABASE SEARCH STRATEGY
The strategy for the electronic journal database search was organized by population, intervention, and outcome variables15 to allow for a focused review that was broad enough to include studies without control groups. The search was applied using both keyword searches and subject heading searches where possible. The search was further limited to (1) articles available in English from January 1990 to July 2007 and (2) articles relating to adults aged 18 to 65 years. No limit was put on study design. The keywords used for all the databases are shown in Table 1.
EVALUATION OF RELEVANCE
Articles identified in the search were then assessed for relevance to the review topic. First, abstracts were scanned for relevant title and keywords. Second, abstracts of identified articles were read, and full articles that were potentially relevant to the review question were obtained. Qualitative articles that explored the experience of returning to work after TBI were included to augment findings, particularly regarding survivor views relating to the third aim of evaluating the strengths, limitations, and applications of the various approaches.
CRITICAL APPRAISAL
All articles obtained were critically appraised in 2 stages. First, the PICOT framework17 was applied on the initial reading of each article to identify the population researched, intervention, comparison, the outcome, and the timeframe. This was done to identify the broad components of the research for critique and comparison. Second, to evaluate the quality of articles that reported research evidence, a standardized scoring tool was used, a research typology developed to evaluate evidence for the National Service Framework for Long Term Conditions in the UK.18 This tool was chosen because it was designed to evaluate a broad range of research designs and to assess evidence relating to long-term neurological conditions. Evaluation using this tool is achieved by scoring questions (0 = no, 1 = in part, or 2 = yes) regarding (1) appropriateness of the research question, aims, and design for achieving the research objectives, (2) clarity of design and methods, (3) adequacy of data for supporting conclusions, and (4) generalizability of findings. The tool specifies the range of scores (out of 10) that indicate whether a study is of low (0-3), medium (4-6), or high (7-10) quality. In addition to the quality score, the tool also allows one to determine whether the evidence came from primary research or secondary data analysis and whether it reported evidence directly from the population concerned. Where articles contained evidence pertaining to several areas of inquiry, appraisal was carried out only for the evidence with relevance to the question of the effectiveness and appropriateness of the particular approach to vocational rehabilitation after TBI. Strength of evidence for conclusions overall was based on the National Service Framework research typology18 grading: (1) moderate evidence for a conclusion when there was more than 1 high-quality study relating to it, at least 1 with direct applicability; (2) weak evidence when there was 1 high-quality study or more than 1 medium-quality study, at least 1 with direct applicability; and (3) insufficient evidence when fewer than 2 medium-quality studies were identified.
For qualitative studies included to augment findings, rigor was assessed using the framework outlined by Mays and Pope.19 This consists of a checklist of attributes proposed as key aspects of good-quality, rigorous, and qualitative studies, including clear description of theoretical framework and methods, detailed description of analysis, and sufficient data stated to support conclusions. Studies meeting all of the relevant criteria for rigor according to this framework were included in the review.
Once the search strategy and evaluation procedure were developed, a researcher (J.K.F.) undertook the primary searching, screening, and appraisal. A second researcher (K.M.M) oversaw the process and assessed articles on those occasions where relevance or rigor was questionable. Any differences between the 2 researchers regarding quality scoring were resolved through discussion.
RESULTS
During the database keyword searches, Ovid Journals returned 2334 results, Web of Science returned 395 results, and Evidence Based Medicine (Ovid) returned 38 results. Manual searching returned no new articles, although some useful background information was obtained regarding descriptions of vocational rehabilitation processes and factors potentially contributing to successful outcomes. An additional database search was undertaken for articles by authors identified in these background articles (returning no new articles that met inclusion criteria). From the abstract reading, 35 research articles, 6 reviews, and 10 practice guide-type articles (sometimes including a case example) were identified as relevant, and the articles were obtained. Of these, 12 articles were deemed not to meet inclusion criteria during the first reading (because the population or outcomes evaluated were not covered by our definitions stated above). Of the 23 research articles that were evaluated for quality, 14 (61%) reported research evaluating existing services and 5 (22%) reported secondary data analysis of those data that were routinely collected from rehabilitation programs or state databases. Of 20 quantitative articles, only 2 were classified as clinical trials (1 with matched controls, 1 with no control group). Three qualitative articles relating to the experience of RTW after brain injury were identified and retrieved. Conclusions are based on the results of scoring by using the National Service Framework for Long Term Conditions tool. Studies judged to be of low quality were not used to inform conclusions.
Question 1: What approaches to vocational rehabilitation after traumatic brain injury are described in the literature?
Three broad models of vocational rehabilitation after brain injury were clearly described in the literature reviewed. The categorization of these approaches into 3 broad categories was based on the models used to underpin the approach (as described below):
1. Program-based vocational rehabilitation based on the New York University (NYU) Medical Center Head Trauma Program model,20
2. The individual placement model of supported employment, and
3. A case coordination approach.
Figure 1 illustrates the basic components of each model and relationships between models.
Program-based vocational rehabilitation model
This approach is characterized by a module-based program aimed at maximizing vocational outcome. There is some variation in the specific components of the modules, but programs are generally based on the NYU Medical Center Head Trauma Program model, which was first described by Ben-Yishay et al20 in 1987. This model contains 3 sequential modules: (1) intensive individualized work skills rehabilitation and interventions within a structured program environment, (2) guided work trials, and (3) assisted placement with transitional job support. Some programs also included follow-up at specific time points following discharge. An overview of the components described in different programs is given in Appendix A. The key differences between this model and the other 2 main approaches are illustrated in Figure 1.
Supported employment model
This approach involves job placement, on-the-job training, and long-term support and job skills reinforcement through on-the-job coaching.21 The "individual placement model" was adapted for people with brain injury by Wehman et al22 in the late 1980s. The key aspects of this approach are quick job placement (based on assessment of abilities, limitations, interests, and work environment) with minimal preemployment training, individualized training and advocacy on the work site by a job coach, job coaching on a one-to-one basis until job competence is reached, and long-term monitoring of performance by the job coach.21 All job training is done on the work site by a job coach rather than before placement.23 In addition, job coach support is not time-limited and is provided on a long-term basis according to individual needs.24 The justification for this intensive intervention is that the severity of disability experienced by the people who receive this type of support is such that they are unlikely to manage in the work environment without it.24 This is reflected in the eligibility criteria for service (see Appendix A). As illustrated in Figure 1, the key differences between supported employment and the other 2 main approaches are (1) intervention is delivered entirely on-the-job and (2) intervention time and extent are not specified or limited.
Case coordination model
Case coordination refers to a holistic approach in which vocational rehabilitation is part of an overall rehabilitation program that is individualized to suit specific needs. Individuals are overseen by a case coordinator who assesses them for service needs and refers accordingly.25 This approach often includes referral for various aspects of vocational rehabilitation described in the other models such as vocational counseling, preemployment training, assisted job placement, and on-the-job support. It can also include referral to specific programs or supported employment where available and appropriate. Key distinguishing features, however, are the monitoring of progress by a case coordinator and integration of vocational rehabilitation into a holistic rehabilitation plan. An overview of each of the case coordination models identified from the literature for this review is included in Appendix A. The main differences between the case coordinated model the other two approaches are (a) the focus on early intervention and continuity of care and (b) coordination of vocational rehabilitation with other post-acute rehabilitation services (see Fig 1).
Question 2: What can we conclude about effectiveness of the various approaches to vocational rehabilitation after traumatic brain injury?
Program-based vocational rehabilitation model
Owing to its influence on later articles, Ben-Yishay et al's original description of the NYU approach20 was included for quality evaluation despite being outside the time frame for the search. Overall, 7 research articles20 were identified that employed program-based approaches (see Table 2). Of these, 5 reported some medium-quality evidence as assessed by the quality scoring tool.18 No high-quality studies were identified. Quality evaluation for these studies is presented in Table 2. On the basis of these studies, the following conclusions are suggested:
* There is weak evidence that people with TBI have better vocational outcomes (ie, employment, level of support, wages, etc) after completing a specialist brain injury vocational program as compared with what they would be expected to achieve before undergoing the program.6,20,29,31
* There is weak evidence that people with TBI who complete this type of vocational rehabilitation program are more likely to gain competitive employment (paid work at a competitive wage) following the program, earn higher wages, and work more hours as compared with those who do not receive vocational rehabilitation.31
* There is weak evidence that of those who gain employment after completion of the program, approximately half remain employed at 1 year following placement.20,29
Supported employment model
Eight research articles and 4 review articles were identified that met the inclusion criteria (see Table 3). In addition, 7 descriptive practice guide-type articles were identified, some with case examples36-42-these were not included in quality evaluation. There were 5 medium-quality studies and 1 high-quality study identified that examined this model. The results of quality evaluation are presented in Table 3. The following overall conclusions are suggested from the studies in this area:
* There is weak evidence that providing supported employment allows some individuals who have not been employed postinjury to be employed specifically within the supported employment model, and for some individuals this can be maintained for several years.21,24,32,33
* There is weak evidence that for people with TBI, provision of supported employment correlates with gaining employment that lasts at least 90 days.35
Unfortunately, these conclusions are quite limited and tell us little about the effectiveness of the supported employment approach to vocational rehabilitation after TBI. It should also be noted that the target outcome in supported employment is different from target outcomes associated with other approaches. For supported employment, the outcome is competitive-wage employment with ongoing job coach support.
Case coordination model
Six research articles examining the case coordinated approach to vocational rehabilitation after TBI were identified that met the inclusion criteria (see Table 4). In addition, 3 articles were found that offered descriptions of case coordinated approaches,49-51 including one that reported audit results but not outcome figures in general51: these were not included in quality evaluation but are described in Appendix A. Two studies examining service provision in relation to employment outcome using data from the Rehabilitation Services Administration database in the United States were also identified.52,53 These studies were not included because the correlations identified could potentially be attributed to a number of other factors. In particular, differences in outcomes for different services could result from provision of services based on individual needs. There were 3 high-quality studies and 1 medium quality study identified for this approach. Table 4 outlines quality evaluation for these studies. From these studies we suggest the following overall conclusions:
* There is moderate evidence that the vocational case coordination model for vocational rehabilitation after brain injury produces higher employment and productivity outcomes than previously reported in the literature. This refers to approximately 80% in some kind of community-based employment (including supported, sheltered, voluntary, and education), with approximately 50% in paid competitive employment without any supports when followed up 1 year after placement.5,44
* There is weak evidence that people who receive intervention within the first year following injury are placed in employment more quickly than people who do not receive intervention within 12 months from the time of injury. Although the article supporting this finding43 was of high quality overall, it should be noted that correlation between injury severity and injury timing of vocational services has not yet been investigated.
Results of Malec and Moessner5 also suggest that referral by case coordinators to supported employment and transitional employment (to get people into the workplace earlier) can lead to competitive employment without support later on (competitive employment at 1-year follow-up rose 32%, while transitional employment dropped 28% and supported employment dropped 10%. There was only a 6% rise in the number of unemployed persons at the end of the first year following placement).
COMPARISONS AMONG APPROACHES
No research literature was found that compared the 3 main approaches in terms of effectiveness for achieving vocational outcomes. It would indeed be difficult to compare models and approaches with one another owing to differences in populations served and definitions of successful outcome. To illustrate this, an overview of populations served and outcome definitions is included in Appendix A.
INFORMATION FROM QUALITATIVE STUDIES
Three qualitative studies were identified that investigated the experience of returning to work after TBI and how this affected people's lives. Two of these studies54,55 were assessed as rigorous,19 with the third being rejected owing to a lack of justification for the unusual methodology.56 Both Levack et al55 and Johansson and Tham54 reported that RTW was viewed by people with TBI as "having achieved normality." Johansson and Tham54 found that work was reevaluated following injury and was given a new meaning. This reevaluation was influenced by the reduced capacity to do previously enjoyed work tasks and the cost-benefit tension regarding managing fatigue. This cost-benefit tension is reinforced in the study by Levack et al,55 which reported that some people made large sacrifices in their personal lives to maintain what they were doing at work. This often led to employment ending with catastrophic personal events. Interestingly, in some cases, this collapse of the work situation did not happen until several years after the individual returned to work. This is an important consideration for follow-up monitoring following RTW services. Levack et al55 also reported that people interpreted success in vocational rehabilitation in a much broader sense than only whether they gained employment. Success sometimes included gaining ability or regaining lost ability to do various other activities, whether or not they gained employment.
DISCUSSION
We now address question 3, consider some issues encountered with this literature, and offer suggestions for further investigation.
Question 3: What are the strengths, limitations, and applications of the various approaches to vocational rehabilitation after TBI?
Because of the lack of high-quality evidence (from which to draw firm conclusions) and the notable absence of studies comparing the different approaches, the question of best practice is still unanswered. One conclusion we suggest can be drawn from the evidence is that vocational rehabilitation is warranted. It is likely that this could be most effectively executed within a case coordination context, because this is the approach currently with the strongest outcome evidence. Although some evidence suggests that assisted job placement and on-the-job supports may be associated with improved outcomes,47 it is still unclear which components of vocational rehabilitation are most effective, and in which contexts.
STRENGTHS AND LIMITATIONS OF THE DIFFERENT APPROACHES
Appendix A outlines the populations, timeframes, and general descriptions of the programs described in the literature reviewed. The key strengths and limitations of each type of approach are outlined in Table 5.
THE STRENGTHS AND LIMITATIONS OF INDIVIDUALIZED INTERVENTION
There is a consensus within the literature that individualized vocational intervention is important for people with TBI.11,39,59,60 Each of the 3 models uses an individualized intervention plan to some extent. However, we are given very little detail about what was actually provided, making assessment of any of these approaches difficult. It is quite possible that the specifics of intervention received, even within the same model, are quite different between individuals and between providers. This highlights a concern that individual staff and providers may be key to the provision of a good-quality service, even when the general model has demonstrated effectiveness. The importance of the individual practitioner role is reinforced by evidence that the therapeutic relationships between rehabilitation staff and their clients are important to outcome.29 This further increases the complexity of the problem. It may be that considering methods of context-dependent rehabilitation used in other populations with brain injury (eg, Ylvisaker et al61) could contribute to resolving some of these issues.
THE MEANING OF VOCATIONAL SUCCESS--IS THERE SOMETHING MISSING?
There appear to be differences in the literature with regard to the definition of "successful" outcome in vocational rehabilitation. Some vocational programs that were reviewed focused solely on RTW as a discreet 'one-off' outcome. However, those in supported employment would argue that RTW is just the beginning, and that intervention must persist to ensure job stability. Job retention is an issue that has often been poorly addressed, particularly in the literature on program-based approaches, despite evidence that it is both important and of concern for this population.62 Furthermore, qualitative studies have shown that there are challenges and experiences associated with RTW after TBI that may not show up in the statistics currently reported by vocational rehabilitation programs, such as the experience in the work environment and the costs associated with trying to achieve the "normality" of being employed.54,55 There is a need, therefore, to look at the trajectory for people who return to work after TBI to assess not only their job retention but also their experience of work postinjury and the effect it has on their lives.
FURTHER INVESTIGATION
On the basis of our review, the following are suggested as topics warranting future investigation:
* Research toward developing standardized measures of how people with TBI in various circumstances can be assisted and supported by vocational rehabilitation options. This would involve research into what interventions work in what circumstances, and how we can assess their potential effectiveness. Especially within a case coordination model, this would go some way toward reducing the problems of effective intervention being somewhat dependent on individual staff skill and experience.
* Research exploring the most effective ways of providing interventions that are suited to each individual, perhaps drawing on literature from other areas of brain injury rehabilitation.
* Comparing trajectories of people who return to work through the different approaches regarding their job retention and the impact of work on their lives to assess the longer-term impact of vocational rehabilitation after TBI.
* Greater research emphasis on exploring the impact of employment after TBI for the person receiving the intervention to inform further research toward identifying services with greater effectiveness and acceptability for survivors of TBI.
Vocational rehabilitation is clearly an area of importance for TBI rehabilitation, and well-designed and focused research could bring huge benefits in practice.
CONCLUSIONS
Approaches to vocational rehabilitation after TBI
There are 3 main models described in the literature for vocational rehabilitation after TBI:
* Program-based job skills training and placement assistance
* Individual placement model of supported employment
* Case coordinated rehabilitation
Each model has a slightly different focus, with program-based rehabilitation focusing on job skills training, supported employment on training on-the-job and ongoing work site support, and case coordinated employment on coordination of rehabilitation services and early intervention.
CURRENT EVIDENCE AND IMPLICATIONS FOR RESEARCH AND PRACTICE
It appears that evidence for the effectiveness of various RTW interventions remains limited. In the absence of definitive evidence regarding efficacy, we suggest that the framework proposed here could be used to evaluate which of the approaches available would be most suitable for the particular individual or population of interest. Consideration of the strengths, weaknesses, and applications for each approach presented in this article may assist with this task.
In addition, we propose that further investigation is warranted, and indeed overdue, to identify the efficacy of the different models for different TBI populations. In particular, research with regard to (1) developing a standardized measure of vocational rehabilitation needs; (2) exploring provision of intervention, which is both standardized and allows for individualization, and; (3) exploration of the long-term impact of vocational rehabilitation and employment after TBI would be particularly pertinent.
REFERENCES