Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Return to Work Coordination Programmes for Work Disability: A Meta-Analysis of Randomised Controlled Trials

  • Stefan Schandelmaier ,

    Affiliation Academy of Swiss Insurance Medicine, University Hospital Basel, Basel, Switzerland

  • Shanil Ebrahim,

    Affiliation Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada

  • Susan C. A. Burkhardt,

    Affiliation Academy of Swiss Insurance Medicine, University Hospital Basel, Basel, Switzerland

  • Wout E. L. de Boer,

    Affiliation Academy of Swiss Insurance Medicine, University Hospital Basel, Basel, Switzerland

  • Thomas Zumbrunn,

    Affiliation Clinical Trial Unit, University Hospital Basel, Basel, Switzerland

  • Gordon H. Guyatt,

    Affiliation Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada

  • Jason W. Busse,

    Affiliations Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada, Department of Anesthesia, McMaster University, Hamilton, Ontario, Canada

  • Regina Kunz

    Affiliation Academy of Swiss Insurance Medicine, University Hospital Basel, Basel, Switzerland

Return to Work Coordination Programmes for Work Disability: A Meta-Analysis of Randomised Controlled Trials

  • Stefan Schandelmaier, 
  • Shanil Ebrahim, 
  • Susan C. A. Burkhardt, 
  • Wout E. L. de Boer, 
  • Thomas Zumbrunn, 
  • Gordon H. Guyatt, 
  • Jason W. Busse, 
  • Regina Kunz



The dramatic rise in chronically ill patients on permanent disability benefits threatens the sustainability of social security in high-income countries. Social insurance organizations have started to invest in promising, but costly return to work (RTW) coordination programmes. The benefit, however, remains uncertain. We conducted a systematic review to determine the long-term effectiveness of RTW coordination compared to usual practice in patients at risk for long-term disability.

Methods and Findings

Eligible trials enrolled employees on work absence for at least 4 weeks and randomly assigned them to RTW coordination or to usual practice. We searched 5 databases (to April 2, 2012). Two investigators performed standardised eligibility assessment, study appraisal and data extraction independently and in duplicate. The GRADE framework guided our assessment of confidence in the meta-analytic estimates. We identified 9 trials from 7 countries, 8 focusing on musculoskeletal, and 1 on mental complaints. Most trials followed participants for 12 months or less. No trial assessed permanent disability. Moderate quality evidence suggests a benefit of RTW coordination on proportion at work at end of follow-up (risk ratio = 1.08, 95% CI = 1.03 to 1.13; absolute effect = 5 in 100 additional individuals returning to work, 95% CI = 2 to 8), overall function (mean difference [MD] on a 0 to 100 scale = 5.2, 95% CI = 2.4 to 8.0; minimal important difference [MID] = 10), physical function (MD = 5.3, 95% CI = 1.4 to 9.1; MID = 8.4), mental function (MD = 3.1, 95% CI = 0.7 to 5.6; MID = 7.3) and pain (MD = 6.1, 95% CI = 3.1 to 9.2; MID = 10).


Moderate quality evidence suggests that RTW coordination results in small relative, but likely important absolute benefits in the likelihood of disabled or sick-listed patients returning to work, and associated small improvements in function and pain. Future research should explore whether the limited effects persist, and whether the programmes are cost effective in the long term.


Long-term sickness absence secondary to illness or injury is associated with reduced quality of life [1], [2], and considerable socioeconomic costs [3][9]. Both patients who are unable to work and the society benefit from return to work (RTW) [2]. However, RTW often requires overcoming challenges, including coping with on-going health problems, re-establishing work functioning, and finding suitable alternative work if a previous job is no longer available [10]. Lack of cooperation between patients, employers, healthcare providers, and insurers may also complicate RTW [1], [10]. The Organisation for Economic Co-operation and Development (OECD) postulated in 2010 that “more people with disability could work if they were helped with the right supports at the right time” through better “cross-agency co-operation” and “systematic and tailored engagement with clients” [1].

Following this intuitively appealing approach, social and private insurers have increasingly implemented RTW coordination services for people receiving wage replacement benefits [11]. RTW coordination, however, demands considerable effort from the affected individual, health professionals, and employers, often without compensation, and is associated with substantial direct costs for insurers. Involved parties thus require reliable information about the effectiveness of RTW coordination to gauge whether RTW coordination is warranted [1].

Existing systematic reviews of RTW interventions have not focused on RTW coordination [12][22]. Therefore, we conducted a systematic review and meta-analysis of randomised controlled trials (RCTs) addressing the effectiveness of RTW coordination compared to usual practice on disability, RTW, function, quality of life and satisfaction in employees receiving wage replacements benefits.


Document S1 shows the protocol of the review.

Eligibility Criteria

Eligible studies met the following criteria: (1) random allocation of adult participants to RTW coordination or usual care, (2) inclusion of participants of whom at least 80% were continuously off work (full or part time sick leave or on disability benefit) for at least four weeks and employed at the time of sick listing, and (3) report of disability status or RTW as an outcome. We defined RTW coordination as involving a direct assessment leading to an individually tailored RTW plan implemented by a RTW-coordinator or team who coordinates services and communication among involved stakeholders.

We excluded employer initiated RTW coordination programmes because they typically focus on prevention of sick leave, and encounter fewer barriers in implementing workplace-directed interventions than insurance or third party RTW coordinators.

Identification of Studies and Data Collection

We carried out a systematic search of MEDLINE, EMBASE, CINAHL, PsycINFO, and the Cochrane Central Register of Controlled Trials from inception to April 2, 2012. Our search strategy combined possible synonyms of RTW coordination (e.g. case management or multidisciplinary rehabilitation), sick leave and disability with a filter for RCTs (see Document S2). We screened reference lists of relevant articles to identify additional eligible trials. Two reviewers independently and in duplicate screened titles and abstracts in any language, reviewed articles in full text, and extracted data from eligible trials. They resolved discrepancies by discussion to achieve consensus. We contacted study authors if information about eligibility criteria, methodological components, or outcome data was incomplete or conflicting.

Assessment of Risk of Bias

Two reviewers independently assessed randomisation sequence generation, concealment of allocation, blinding of participants, RTW coordinators, and outcome assessors, completeness of data, whether participants were analysed in the group to which they were initially randomised, and whether selective outcome reporting occurred. Cluster RCTs were assessed for recruitment bias [23], and appropriate statistical analysis [23]. We assessed blinding of outcome assessment and completeness of data separately for RTW outcomes and patient reported outcomes (PROs). We used a modified Cochrane risk of bias instrument [23], with response options of “definitely yes”, “probably yes”, “probably no”, and “definitely no” with definitely and probably yes ultimately assigned high risk of bias and probably and definitely no assigned low risk of bias [24]. Because of the small number of studies for each outcome, we were unable to address publication bias or explore explanations for variability in results [23].

Data Analysis

We conducted random effects meta-analyses (MAs) using RevMan 5.1 [25] and R 2.15.0 [26]. If available, we used baseline-adjusted effect estimates. In case of missing values, we analysed the available data without imputations to prevent biased weighting of studies [23]. We used I2 to estimate heterogeneity [23].

We expressed pooled effects of dichotomous outcomes as risk ratios and calculated illustrative absolute risk differences by using the median baseline risk. We pooled effects of continuous outcomes as differences between group means (mean differences).

We felt the most important outcome was RTW that persisted over the long term; if we found varying measures of RTW, we therefore focused on the one that best reflected long-term outcome. If studies with time to event outcomes failed to report hazard ratios (HR), we extracted individual patient data from survival curves, verified the extraction by re-plotting, and then calculated the HR and associated 95% confidence interval (CI). If data extraction was not possible, we calculated HRs and 95% CIs based on log-rank-tests [27].

Five reviewers independently grouped all PROs by consensus into 9 categories: Overall function, physical function, social function, mental function, general health, pain, depression, anxiety, and patient satisfaction. We preferred change scores to end scores in order to correct for possible baseline differences, but we pooled both types of scores as change scores were not available for all trials. We transformed PROs expressed in different units to units on the scale of the most familiar instrument before we pooled mean differences [28]. This allowed us to enhance the interpretation of the summary effect by considering an anchor based minimal important difference (MID) on that instrument. Specifically, we rescaled overall function into the 0 to 100 scale of the Oswestry Disability Index (MID = 10 [29][34]), physical, mental and social function into the 0 to 100 scale of the SF-36 (MIDs = 8.4, 7.3, and 11.7 [35], respectively) and pain into a 0 to 100 visual analogue scale (MID = 10 [36]). In a second step, we used the rescaled outcomes to calculate the proportion of participants who improved by at least one MID in each group of each trial which allowed us to calculate and pool risk differences (RD) [28].

We conducted sensitivity analysis if a study reported several definitions of a RTW-outcome, e.g. full-time and part-time RTW versus full-time only (specified in footnotes of table 3). If more than one study reported several definitions, we conducted meta-analyses of all possible combinations, that is six for proportion at work at end of study and six for proportion ever returned to work.

Reporting and Rating Quality of Evidence

The PRISMA statement [37] guided our reporting and the GRADE framework [38] guided our assessment of confidence in the meta-analytic estimates.


Identification of Eligible Trials and Data Collection

Of 2459 citations, 15 articles [39][55] describing 9 RCTs proved eligible (figure 1). We approached 12 authors of whom 10 replied and 7 provided additional information about 7 studies [39][44], [46] (footnotes in tables 1, 2, 3, 4).

Figure 1. Study selection.

Last update of electronic search to April 2, 2012.

Characteristics of Included Trials

Table 1 shows characteristics of studies and populations. Participants were consenting volunteers in all but one study in which participants received no official information about the intervention [46]. Table 2 shows characteristics of interventions and comparisons. No study specified the financial resources available to the RTW coordinators for patient support. In five studies [39], [40], [43], [45], [46], some participants assigned to practice as usual may have received RTW coordination.

Table 1. Characteristics of studies and populations (at time of randomisation).

Table 3 shows details of the reported outcome measures. The outcome proportion at work at end of study best reflected long-term in contrast to time until stable RTW and proportion ever returned to work that provided information regarding the first episode of RTW or the first episode of RTW of a specific duration, and sickness absence days that expressed the duration of all episodes of sickness absence.

Risk of Bias

Table 4 presents our assessment of risk of bias. See footnotes of table 4 for unclear or incomplete reporting of outcomes that we could not clarify with authors. Most studies concealed allocation and conducted an analysis-as-randomised. Blinding of personnel, participants and assessors of patient reported outcomes (self-administered questionnaires) was impossible. Loss to follow-up was substantial in most studies.

Effects and Confidence in Estimates

Table 5 shows the evidence profile of the meta-analytic estimates of important outcomes and Table S1 the summary of findings table for all outcomes. The heterogeneity was low across all outcomes but risk of bias (high attrition or selective reporting), imprecision and indirectness limited our confidence in the estimates.

All pooled effects of RTW outcomes significantly favoured RTW coordination (figure 2). The proportion at work at end of study increased by a factor of 1.08 (95% confidence interval (CI) 1.03 to 1.13, moderate confidence). This corresponds to an absolute effect of 5 in 100 more individuals returning to work (95% CI 2 more to 8 more). The pooled hazard ratio of time until stable RTW was 1.34 (95% CI 1.12 to 1.36, moderate confidence). The proportion of ever returning to work increased by a factor of 1.07 (95% CI 1.00 to 1.13, low confidence), corresponding to 4 more per 100 (95% CI, 0 more to 8 more). Total sickness absence days decreased by 36 workdays per year (95% CI, 17 to 56, moderate confidence). Sensitivity analysis did not reveal any substantial differences in our pooled estimates or heterogeneity.

Figure 3 shows meta-analyses of PROs. Expressed on a 0 to 100 scale, RTW coordination improved mean overall function by 5.2 (95% CI 2.4 to 8.0; MID = 10, moderate confidence), physical function by 5.3 (95% CI 1.4 to 9.1; MID = 8.4, moderate confidence), pain by 6.1 (95% CI 3.1 to 9.2; MID = 10, moderate confidence), mental function by 3.1 (95% CI 0.7 to 5.6; MID = 7.3, moderate confidence) and social function by 3.1 (95% CI –0.6 to 6.8; MID = 11.7, low confidence). When we used the MIDs to calculate risk differences, RTW coordination increased the proportion of participants who improved considerably in overall function by 9% (95% CI 4 to 15%), physical function by 8% (95% CI 2 to 14%), pain by 8% (95% CI 2 to 13%), mental function by 6% (95% CI 0 to 11%), and social function by 4% (95% CI –2 to 10%).

Figure 3. Patient reported outcomes.

Individual trials’ outcomes expressed on a 0 to 100 scale. RTW coord. = return to work coordination. MID = minimal important difference.

Figure S1 shows the output of the RevMan software including the raw data.


We found moderate quality evidence that RTW coordination interventions result in small relative increases in RTW. Assuming a typical risk of 43 in 100 individuals not returning to work, this small relative effect implies an absolute effect of 5 in 100 more returning to work. If maintained over the long term, many would consider this an important benefit. We also found moderate quality evidence that the intervention results in small improvements in function and pain. We found no evidence that one type of RTW coordination programme was superior to another.

Our findings gain credence from the rigor of the review. We performed a comprehensive search, adjudicated eligibility and extracted data independently and in duplicate, obtained additional information from 7 authors, performed appropriate primary and sensitivity analyses and evaluated confidence in estimates of effect using the GRADE approach [38].

Our review has limitations. First, given the small number of studies for each outcome, we were unable to address publication bias. Second, we pooled change and end scores for the PROs. In theory, standard deviations of the two scores might differ substantially, leading to different weighting of individual studies in the meta-analysis [23]. However, there is evidence that SDs of change scores often do not appreciably differ from end scores [56]. Third, results from two cluster RCTs uncorrected for intra-cluster dependency may have spuriously increased precision, thus overweighting these studies in the meta-analysis.

Comparison with Other Systematic Reviews

Our study selection partly overlaps with related systematic reviews that defined RTW interventions from different points of view. They compared usual practice to RTW interventions that either included a specific workplace component [12][15], applied RTW-interventions to a population with a specific health condition [16][19], or explored them within a specific country only [20][22]. Two of these systematic reviews (with 3/42 [17] or 0/10 [13] studies overlapping) addressed RTW coordination in a subgroup analysis (RTW coordination as a subgroup of RTW-interventions). Both suggested that RTW coordination improved RTW [13], [17] whereas effects on PROs remained unclear [13]. However, much like other related reviews, they did not perform a meta-analysis. Reasons included poor study quality [15] or high heterogeneity in the RTW interventions [15], [17], [18]. Only one systematic review (1/6 studies overlapping) conducted a meta-analysis, concluding with low confidence that RTW interventions with an active workplace involvement improve RTW outcomes [12].

Other reviews also noted limitations in the evidence that we identified. Evidence regarding the effectiveness of RTW interventions suffers from poor descriptions of interventions and controls [12], [13], insufficient information beyond one year follow-up [13], [18], and paucity of studies on participants with mental health problems [12], [13]. Further, a systematic review of 34 RCTs (3 overlapping) and 8 cohort studies found evidence of possible publication bias [17].

Applicability of Findings

Applicability of the results is enhanced by recruitment through insurance registers that ensured a representative selection of claimants. The prompt initiation of interventions after work absence and the high intensity of support are consistent with the OECD recommendations that social insurances or corresponding benefit authorities should apply RTW coordination at an early stage and resources should shift from passive benefits towards RTW programmes [1].

Diversity and limitations in the description of both RTW coordination interventions, and the nature of usual practice, advise on cautious interpretation and application of our results. Most studies focused on organisational features, such as composition of the team, distribution of roles, and standardisation of initial assessment. Interventions differed in degree of standardisation, and in the roles and backgrounds of intervention providers. Information regarding training and experience of RTW coordinators, resources available, and adherence of coordinators and participants were typically lacking. Descriptions of the usual practice controls were even more limited.

The striking consistency of results from study to study in virtually all outcomes ameliorated the unease about variability in interventions and controls. If variability were very important, one would not expect to see such consistency.

All but 2 studies [42], [45] (85% of participants in the review) focused on claimants with musculoskeletal complaints. Recent statistics from high-income countries show that new disability claimants with psychiatric disorders (30 to 40%) have outnumbered those with musculoskeletal complaints [1]. Although the results from the two studies that did enrol a substantial proportion [42] or an exclusive sample [45] of claimants with psychiatric complaints showed similar results to other studies, generalizing results to these populations is questionable.

Judging the importance of our measured relative effect size is challenging. An absolute difference in the proportion at work at end of study - of the order of 5% suggested by the results of this review - could be important if maintained over the long term. Indeed, many are likely to agree that an absolute reduction in the proportion on long-term disability would be important. However, follow-up was generally too short to inform results over the long-term. Only one study assessed work stability after initial work resumption but reported the results incompletely [47].

Two studies conducted an economic analyses based on the outcome cumulative sickness absence [39], [50] one year after randomisation. They both concluded that RTW coordination compared to usual practice was cost effective from a societal perspective, that is by considering the cost of the intervention, health care utilisation, and loss of productivity. The societal perspective leaves out the cost of wage replacement, which is considered a redistribution of wealth, and, therefore, does not inform about the impact of RTW coordination on social security savings. In contrast, an economic analysis from an insurance perspective would integrate this information. Cost effectiveness from an insurance perspective may occur only in the long-term and depend mainly on savings related to fewer disability pensions [57].

Implications for Research

Results to date suggest small but possibly important benefits of RTW coordination. Determining the long-term benefits and the cost effectiveness of the programmes will require trials with low risk of bias (concealment, blinding of outcome assessors and statisticians, minimal missing data), measuring long-term outcomes of work force retention and long-term disability (including pensions). This would also enable extending the research on comparing different definitions of RTW outcomes [58]. We require studies in specific populations that represent the majority of disabled individuals, including both musculoskeletal and psychiatric problems. We strongly encourage researchers of RTW interventions to describe interventions, comparisons, and settings more systematically to enable comparability of studies and facilitate transfer into practice.

Supporting Information

Figure S1.

RevMan output for all outcomes including raw data.


Table S1.

Summary of findings for all outcomes.



We thank the study authors Johannes Anema, Judith Bosmans, Ute Bültmann, Peter Donceel, Susan Purdon, Michel Rossignol, and William Shaw for their unhesitant cooperation and for providing data; Urs Brügger, Julian Higgins and Stephen Walter for valuable advice.

Author Contributions

Conceived and designed the experiments: RK SS SB JB GG. Performed the experiments: SS RK SB WdB SE JB. Analyzed the data: TZ SS RK GG. Wrote the paper: SS SE SB WdB TZ JB GG RK. Developed the search strategy: SS RK. Performed the study selection: SS SB RK WdB. Appraised study quality: RK SS. Extracted study characteristics and outcome data: SE RK SS TZ. Performed consensus exercises: SE JWB SS RK WdB. Contributed to the interpretation and discussion of the results: SS SE SB WdB TZ JB GG RK.


  1. 1. OECD (2010) Sickness, Disability and Work: Breaking the Barriers. A Synthesis of Findings across OECD Countries. Paris: OECD Publishing. p.
  2. 2. Gordon Waddell, Burton KA (2006) Is Work Good for Your Health and Well-Being? London: The Stationery Office. p.
  3. 3. Frymoyer JW, Cats-Baril WL (1991) An overview of the incidences and costs of low back pain. Orthop Clin North Am 22: 263–271.
  4. 4. Nachemson A (1994) Chronic pain - the end of the welfare state? Qual Life Res 3 Suppl 1: S11–17.
  5. 5. Abenhaim L, Suissa S (1987) Importance and economic burden of occupational back pain: a study of 2,500 cases representative of Quebec. J Occup Med 29: 670–674.
  6. 6. Labriola M, Lund T (2007) Self-reported sickness absence as a risk marker of future disability pension. Prospective findings from the DWECS/DREAM study 1990–2004. Int J Med Sci 4: 153–158.
  7. 7. Lund T, Kivimäki M, Labriola M, Villadsen E, Christensen KB (2008) Using administrative sickness absence data as a marker of future disability pension: the prospective DREAM study of Danish private sector employees. Occup Environ Med 65: 28–31
  8. 8. Kivimäki M, Forma P, Wikström J, Halmeenmäki T, Pentti J, et al. (2004) Sickness absence as a risk marker of future disability pension: the 10-town study. J Epidemiol Community Health 58: 710–711
  9. 9. Gjesdal S, Bratberg E (2003) Diagnosis and duration of sickness absence as predictors for disability pension: Results from a three-year, multi-register based and prospective study. Scand J Public Health 31: 246–254
  10. 10. Young AE, Roessler RT, Wasiak R, McPherson KM, van Poppel MNM, et al. (2005) A developmental conceptualization of return to work. J Occup Rehabil 15: 557–568
  11. 11. Young AE, Wasiak R, Roessler RT, McPherson KM, Anema JR, et al. (2005) Return-to-work outcomes following work disability: stakeholder motivations, interests and concerns. J Occup Rehabil 15: 543–556
  12. 12. van Oostrom SH, Driessen MT, de Vet HC, Franche R-L, Schonstein E, et al. (2009) Workplace interventions for preventing work disability. Cochrane Database Syst Rev. doi:10.1002/14651858.CD006955.pub2.
  13. 13. Franche R-L, Cullen K, Clarke J, Irvin E, Sinclair S, et al. (2005) Workplace-based return-to-work interventions: a systematic review of the quantitative literature. J Occup Rehabil 15: 607–631
  14. 14. Kuoppala J, Lamminpää A (2008) Rehabilitation and work ability: a systematic literature review. J Rehabil Med 40: 796–804
  15. 15. Carroll C, Rick J, Pilgrim H, Cameron J, Hillage J (2010) Workplace involvement improves return to work rates among employees with back pain on long-term sick leave: a systematic review of the effectiveness and cost-effectiveness of interventions. Disabil Rehabil 32: 607–621
  16. 16. Meijer EM, Sluiter JK, Frings-Dresen MHW (2005) Evaluation of effective return-to-work treatment programs for sick-listed patients with non-specific musculoskeletal complaints: a systematic review. Int Arch Occup Environ Health 78: 523–532
  17. 17. Palmer KT, Harris EC, Linaker C, Barker M, Lawrence W, et al. (2011) Effectiveness of community- and workplace-based interventions to manage musculoskeletal-related sickness absence and job loss: a systematic review. Rheumatology (Oxford, England): 1–13. doi:10.1093/rheumatology/ker086.
  18. 18. Hlobil H, Staal JB, Spoelstra M, Ariëns GAM, Smid T, et al. (2005) Effectiveness of a return-to-work intervention for subacute low-back pain. Scand J Work Environ Health 31: 249–257.
  19. 19. Elders LA, van der Beek AJ, Burdorf A (2000) Return to work after sickness absence due to back disorders–a systematic review on intervention strategies. Int Arch Occup Environ Health 73: 339–348.
  20. 20. Clayton S, Bambra C, Gosling R, Povall S, Misso K, et al. (2011) Assembling the evidence jigsaw: insights from a systematic review of UK studies of individual-focused return to work initiatives for disabled and long-term ill people. BMC Public Health 11: 170
  21. 21. Bambra C, Whitehead M, Hamilton V (2005) Does “welfare-to-work” work? A systematic review of the effectiveness of the UK’s welfare-to-work programmes for people with a disability or chronic illness. Soc Sci Med 60: 1905–1918
  22. 22. Hayday S, Rick J, Carroll C, Jagger N, Hillage J (2008) Review of the Effectiveness and Cost Effectiveness of Interventions, Strategies, Programmes and Policies to Help Recipients of Incapacity Benefits Return to Employment (Paid and Unpaid). Brighton: Institute for Employment Studies. p.
  23. 23. Higgins JPT, Green S (2011) Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. Higgins JPT, Green S, editors The Cochrane Collaboration. p.
  24. 24. Akl EA, Sun X, Busse JW, Johnston BC, Briel M, et al. (2012) Specific instructions for estimating unclearly reported blinding status in randomized trials were reliable and valid. J Clin Epidemiol 65: 262–267
  25. 25. The Nordic Cochrane Centre, The Cochrane Collaboration (2011) Review Manager (RevMan). Copenhagen. p.
  26. 26. R Development Core Team (2011) (n.d.) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. p. Available:
  27. 27. Parmar MK, Torri V, Stewart L (1998) Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med 17: 2815–2834.
  28. 28. Thorlund K, Walter SD, Johnston BC, Furukawa TA, Guyatt GH (2011) Pooling health-related quality of life outcomes in meta-analysis - a tutorial and review of methods for enhancing interpretability. Research Synthesis Methods 2: 188–203
  29. 29. Lauridsen H, Hartvigsen J, Manniche C, Korsholm L, Grunnet-Nilsson N (2006) Responsiveness and minimal clinically important difference for pain and disability instruments in low back pain patients. BMC Musculoskeletal Disorders 7: 82.
  30. 30. Copay AG, Glassman SD, Subach BR, Berven S, Schuler TC, et al. (2008) Minimum clinically important difference in lumbar spine surgery patients: a choice of methods using the Oswestry Disability Index, Medical Outcomes Study questionnaire Short Form 36, and pain scales. Spine J 8: 968–974.
  31. 31. Ostelo RWJG, De Vet HCW (2005) Clinically important outcomes in low back pain. Best Pract Res Clin Rheumatol 19: 593–607.
  32. 32. Fisher K (2008) Assessing clinically meaningful change following a programme for managing chronic pain. Clin Rehabil 22: 252–259
  33. 33. Hägg O, Fritzell P, Nordwall A (2003) The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J 12: 12–20.
  34. 34. Ostelo RWJG, Deyo RA, Stratford P, Waddell G, Croft P, et al. (2008) Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine 33: 90–94
  35. 35. Kosinski M, Zhao SZ, Dedhiya S, Osterhaus JT, Ware JE Jr (2000) AID-ANR10>3.0.CO;2-M.
  36. 36. Dworkin RH, Turk DC, McDermott MP, Peirce-Sandner S, Burke LB, et al. (2009) Interpreting the clinical importance of group differences in chronic pain clinical trials: IMMPACT recommendations. Pain 146: 238–244
  37. 37. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. J Clin Epidemiol 62: 1006–1012
  38. 38. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, et al. (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 336: 924–926
  39. 39. Bültmann U, Sherson D, Olsen J, Hansen CL, Lund T, et al. (2009) Coordinated and tailored work rehabilitation: a randomized controlled trial with economic evaluation undertaken with workers on sick leave due to musculoskeletal disorders. J Occup Rehabil 19: 81–93
  40. 40. Lambeek LC, van Mechelen W, Knol DL, Loisel P, Anema JR (2010) Randomised controlled trial of integrated care to reduce disability from chronic low back pain in working and private life. BMJ 340: c1035
  41. 41. Rossignol M, Abenhaim L, Séguin P, Neveu a, Collet JP, et al. (2000) Coordination of primary health care for back pain. A randomized controlled trial. Spine 25: 251–258.
  42. 42. Purdon S, Stratford N, Taylor R, Natarajan L, Bell S (2006) Impacts of the Job Retention and Rehabilitation Pilot. Changes.
  43. 43. Feuerstein M, Huang GD, Ortiz JM, Shaw WS, Miller VI, et al. (2003) Integrated case management for work-related upper-extremity disorders: impact of patient satisfaction on health and work status. J Occup Environ Med 45: 803–812
  44. 44. Davey C (1994) The implementation and evaluation of a rehabilitation co-ordinator service for personal injury claimants.
  45. 45. van der Feltz-Cornelis CM, Hoedeman R, de Jong FJ, Meeuwissen JA, Drewes HW, et al. (2010) Faster return to work after psychiatric consultation for sicklisted employees with common mental disorders compared to care as usual. A randomized clinical trial. Neuropsychiatr Dis Treat 6: 375–385.
  46. 46. Donceel P, Du Bois M, Lahaye D (1999) Return to work after surgery for lumbar disc herniation. A rehabilitation-oriented approach in insurance medicine. Spine 24: 872–876.
  47. 47. Lindh M (1997) A randomized prospective study of vocational outcome in rehabilitation of patients with non-specific musculoskeletal pain: A multidisciplinary approach to patients identified after 90 days of sick-leave. Scan J Rehab Med 29: 103–112.
  48. 48. Davey C (1993) Evaluating a rehabilitation co-ordinator service for personal injury claimants. International journal of rehabilitation research 16: 49–53.
  49. 49. Farrell C, Nice K, Lewis J, Sainsbury R (2006) Experiences of the Job Retention and Rehabilitation Pilot. Pensions.
  50. 50. Lambeek LC, Bosmans JE, Van Royen BJ, Van Tulder MW, Van Mechelen W, et al. (2010) Effect of integrated care for sick listed patients with chronic low back pain: economic evaluation alongside a randomised controlled trial. BMJ 341: c6414
  51. 51. Lambeek LC, Anema JR, van Royen BJ, Buijs PC, Wuisman PI, et al. (2007) Multidisciplinary outpatient care program for patients with chronic low back pain: design of a randomized controlled trial and cost-effectiveness study. BMC public health 7: 254
  52. 52. Lambeek LC, van Mechelen W, Buijs PC, Loisel P, Anema JR (2009) An integrated care program to prevent work disability due to chronic low back pain: a process evaluation within a randomized controlled trial. BMC musculoskeletal disorders 10: 147
  53. 53. Lincoln AE, Feuerstein M, Shaw WS, Miller VI (2002) Impact of Case Manager Training on Worksite Accommodations in Workers’ Compensation Claimants With Upper Extremity Disorders. Journal of Occupational and Environmental Medicine 44: 237–245
  54. 54. Shaw WS, Feuerstein M, Lincoln AE, Miller VI, Wood PM (2001) Case management services for work related upper extremity disorders. Integrating workplace accommodation and problem solving. AAOHN J 49: 378–389.
  55. 55. van der Feltz-Cornelis CM, Meeuwissen J a C, de Jong FJ, Hoedeman R, Elfeddali I (2007) Randomised controlled trial of a psychiatric consultation model for treatment of common mental disorder in the occupational health setting. BMC health services research 7: 29
  56. 56. Busse JW, Montori VM, Krasnik C, Patelis-Siotis I, Guyatt GH (2009) Psychological Intervention for Premenstrual Syndrome: A Meta-Analysis of Randomized Controlled Trials. Psychother Psychosom 78: 6–15
  57. 57. Busch H, Bodin L, Bergström G, Jensen IB (2011) Patterns of sickness absence a decade after pain-related multidisciplinary rehabilitation. Pain 152: 1727–1733
  58. 58. Steenstra IA, Lee H, de Vroome EMM, Busse JW, Hogg-Johnson SJ (2012) Comparing Current Definitions of Return to Work: A Measurement Approach. J Occup Rehabil [Epub ahead of print]. doi:10.1007/s10926-011-9349-6.