• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of cmajCMAJ Information for AuthorsCMAJ Home Page
CMAJ. Mar 19, 2002; 166(6): 749–754.

Listening to injured workers: how recovery expectations predict outcomes — a prospective study

Donald C. Cole,*† Michael V. Mondloch,* Sheilah Hogg-Johnson,*† and The Early Claimant Cohort Prognostic Modelling Group



Rigorous evidence on factors affecting the prognosis of work-related soft-tissue injuries remains limited. Although shown to be important for a wide variety of clinical conditions, recovery expectations have rarely been assessed as prognostic factors for workers with soft-tissue injuries. We examined the predictive role of various measures of recovery expectations among workers with injuries resulting in time off work.


We identified a prospective cohort of 1566 injured workers shortly after they filed a claim for their injury with the Ontario Workers' Compensation Board (OWCB). They had soft-tissue injuries to the back or upper or lower extremities, had new, lost-time claims from May to November 1993 and were still off work at the time of the first interview. We interviewed participants by telephone within 3 weeks after the injury and measured their recovery expectations (perceptions regarding progress, expected change in condition, expected time until return to usual activities and expectations regarding return to usual job) along with other, potentially important prognostic factors. The primary outcome was total time receiving 100% wage-replacement benefits during the year following injury, obtained from OWCB administrative files. Self-reported measures of pain, health-related quality of life and functional status, obtained up to 4 times during the year following injury, were both independent predictors and secondary outcomes.


The 4 measures of recovery expectations together explained one-sixth of the variation in time receiving benefits. All but expectations regarding return to usual job were individually predictive of time receiving benefits. Judging one's recovery as much better than expected resulted in a 30% (95% confidence interval [CI] 9%–46%) faster rate of stopping receiving benefits (and likely returning to work) compared with judging one's recovery as much worse than expected. Similarly, participants who expected to return to usual activities within 3 weeks had a 37% (95% CI 26%–47%) faster rate of stopping receiving benefits than those who responded “Don't know” to this question, and participants who stated that they were fully recovered or would get better soon had a 25% (CI 5%–40%) faster rate than those who thought they would never get or stay better. Positive recovery expectations were also associated with reductions in pain grade and improvement in functional status outcomes.


Expectations regarding recovery may provide useful information on the complex process of recovering from work-related soft-tissue injuries. For clinicians, patients' negative or uncertain expectations may indicate the need for further probing and intervention on psychosocial factors to facilitate recovery.

Injuries that do not result in fractures or dislocations, often called soft-tissue injuries, may have a highly variable clinical course, from rapid resolution to chronic persistence. When the soft-tissue injuries are judged to be work-related, this variability can become stressful for the patient, the physician and workers' compensation authorities. Since soft-tissue injuries account for the vast majority of lost-time workers' compensation claims in Canada,1 better evidence for important prognostic factors is sorely needed.2,3 Whereas job demands and severity of injury, pain and functional impairment may be obvious prognostic factors for most clinicians, psychosocial factors may be more difficult to grapple with.4

Among psychosocial factors, patients' concerns about “ability to manage at work” were found to be independently predictive of work status at 1 year among patients with chronic low-back pain.5 Yet little other high-quality scientific evidence on the prognostic significance of recovery expectations exists for patients with soft-tissue injuries, as documented in a recent systematic review of the prognostic role of patients' recovery expectations across a wide variety of clinical conditions.6 Hence, we examined the predictive role of various measures of recovery expectations in a prospective cohort of workers with injuries resulting in time lost from work. We hypothesized that injured workers with more positive expectations would stop receiving wage-replacement benefits (i.e., return to work) sooner and report better health-related outcomes (e.g., reduced pain) than workers with less positive expectations.


The Ontario-wide cohort of injured workers with lost-time claims for soft-tissue disorders was part of a larger program evaluation,7 conducted in conjunction with the Ontario Workers' Compensation Board (OWCB) (now the Workplace Safety & Insurance Board). Since the OWCB has the statutory authority to ask claimants about relevant issues in order to inform program and policy development, the study protocol did not undergo external ethical review. However, all respondents were free to decline participation and were assured that participation status would not affect their access to OWCB services. Furthermore, confidentiality of their individual responses was guaranteed, with only aggregate data from analyses provided to the OWCB. The sampling strategy, eligibility criteria and participation rates are described in detail elsewhere.8

Potential participants were identified shortly after claim registration with the OWCB (n = 12 182). They were screened for eligibility through review of their claim file. To be eligible, participants had to have a soft-tissue injury to the back or upper or lower extremity, have a new, lost-time workers' compensation claim from May to November 1993 and still be off work at the time of the first interview. A total of 5866 claimants were ineligible. The remaining 6316 workers were telephoned within 21 days after the injury to explain the study and obtain informed consent; 3381 people could not be contacted, and 1102 people refused to participate. Refusers were more likely to be male, to earn more, to be employed by a large, self-insured employer and to have shorter duration of benefits than participants. Subsequently excluded were 259 people who had already returned to work or whose claim was a reopening of one registered within the previous 30 days, and 8 people with very incomplete baseline information. Thus, included in the present analysis were 1566 people with new claims who were still off work at the time of first interview. Most of the participants (62.4%) had demographic, employment and claim characteristics that were remarkably similar to those of all OWCB claimants with soft-tissue injuries.8 Intentional undersampling of men resulted in relatively equal numbers of women and men in the cohort. Repeated telephone contacts were made 4, 16 and 52 weeks after the injury.

We measured the participants' recovery expectations using several questions adapted from previously published questions9,10 or developed de novo based on concepts arising out of qualitative work.11 Questions were designed to be relatively neutral with respect to the determinants of recovery expectations in order to allow injured workers to take “all things into consideration.” Two questions addressed clinical course. The first focused on whether the patient was progressing as he or she expected,9 and the second focused on expected change in the injury or condition.10 Two questions addressed future functioning. The first asked the expected time until return to usual activities,10 and the second, whether the patient expected to return to his or her usual job.11

We examined 4 key outcomes. The cumulative number of calendar days during which the worker received 100% benefits within the year after the injury came from the OWCB benefits database and was therefore available for all participants. Pain was measured with the use of the Chronic Pain Grade,12 modified to a suitable time frame. Generic health-related quality of life was measured with the use of the 36-item Short Form Medical Outcomes Survey (SF-36).13,14 Injury-specific or body-area-specific functional status was measured with the use of the modified American Shoulder and Elbow Surgeons instrument for upper extremity injuries,15 the Western Ontario McMaster instrument for lower extremity injuries16 and the Roland instrument for back problems.17

A wide variety of clinical, demographic, work-related and other variables may hypothetically affect both injured workers' expectations of recovery and outcomes.18 Among clinical factors we included pain grade, health-related quality of life, functional status and the presence of serious coexisting conditions (measured on the baseline interview). Demographic factors included sex, income, marital status and age. Work-related factors included industrial sector, job heaviness/demands according to the National Occupation Classification19,20 and offers of workplace accommodation to facilitate return to work.

We conducted analyses in 3 stages. First, to refine the recovery expectation information, we used descriptive statistics and tests of association across measures, including χ2 tests and Cramer's V, a measure of association for nominal level data. There were significant statistical associations among the 4 questions (χ2 tests, all p values < 0.001). However, because the associations were weak (Cramer's V range 0.09–0.24) we included all the questions in the multivariable analyses.

Second, to estimate the relation between recovery expectations and total time receiving benefits, we included the recovery expectation questions as explanatory variables in a Cox regression analysis and adjusted for the potential confounding effects of clinical, demographic and work-related variables (n = 1332 with complete information). Adjustment resulted in loss of significance of the “Don't know” category in the question regarding expectations to return to one's usual job. Results are reported with hazard rate ratios or the ratio of the instantaneous conditional rate of stopping receiving benefits at any particular time. To gauge the relative importance of the 4 recovery expectation variables, as measured by percentage of explained variation in time receiving benefits, we carried out calculations using the method described by Schemper.21 Marginal effects describe the percentage of variation explained in separate regressions for each prognostic factor. Partial effects describe the increase in the percentage of variation explained by including the factor in a regression model that already includes the other prognostic factors. Variation explained by marginal and partial effects were as follows: perceptions regarding progress: 8.1% marginal effects, 3.0% partial effects; expected change in condition: 6.1% and 1.2% respectively; expected time until return to usual activities: 9.2% and 3.4% respectively; and expectations regarding return to usual job: 2.5% and 0.7% respectively. Expectations regarding return to usual job accounted for significantly less variation than perceptions regarding progress or expected time to return to usual activities (t > 2.56, p < 0.05). When we analyzed each of the 4 variables individually, expectations regarding return to usual job explained less variation than the other variables (t > 2.76, p < 0.05). None of the other differences was statistically significant.

Third, among the participants with complete data, we examined the relations between initial expectations regarding recovery and changes in pain grade (n = 704), SF-36 scores (n = 706) and functional status (n = 701). Loss to follow-up varied across baseline response categories of expected change in condition, with greater loss of patients who responded that they did not know the future course of their condition and less loss of patients who expected to not get better (p < 0.05). Loss to follow-up was not associated with responses to the other baseline measures of recovery expectations (all p values > 0.05). We evaluated relations between initial recovery expectations and changes in health outcomes using the interaction terms between time and baseline recovery expectations from multivariable repeated-measures analysis of variance, with separate analyses of variance for each outcome.


The personal, work-related and clinical characteristics of the cohort are presented in Table 1. The injury was a new one for the vast majority of participants. The median time receiving wage-replacement benefits, and therefore off work, was 55 days. The participants lost to telephone follow-up at 52 weeks (n = 611 [39%]) had a lower mean duration receiving benefits and a higher mean baseline score on the physical dimensions of the SF-36 than those not lost to follow-up, which indicates a selection bias toward inclusion of those worse off in the analyses of changes in pain grade and functional status.

Table thumbnail
Table 1

At the first telephone interview, more participants thought they were progressing worse than expected (n = 765) than better than expected (n = 605), and most participants (n = 825) thought they would get better slowly (Table 2). About two-thirds of the participants responded “Don't know” to the question regarding expected time until return to usual activities. Despite such uncertainty over the duration of their problems, over three-quarters of the participants thought they would recover enough to return to their usual job.

Table thumbnail
Table 2

Together, the 4 recovery expectation variables explained 15.5% of the variation in time receiving benefits. In other words, knowing a worker's expectations would reduce a clinician's prognostic uncertainty by about one-sixth. Expectations regarding return to usual job accounted for the least variation. Three of the 4 variables (perceptions regarding progress, expected change in condition and expected time until return to usual activities) were significant predictors of time receiving benefits, even after variables other than those regarding recovery expectations were adjusted for (Table 2). In general, workers' expectations of slower recovery or greater uncertainty were consistently associated with longer periods off work and receiving wage-replacement benefits.

Injured workers started with relatively similar levels of pain across different recovery expectation categories. Nevertheless, over time, those who expected to return to their usual activities in less than 3 weeks showed the greatest reduction in pain severity by 4 weeks, and their pain levels remained lower at the end of the year than those of the participants who responded that they did not know when they would be able to return to usual activities (p = 0.007) (Fig. 1). Positive recovery expectations, as indicated by the response “Much better” on the question regarding progress (p = 0.012) and “Get better soon” on the question regarding expected change in condition (p < 0.001), were also associated with reductions in reported pain levels over time. More rapid and greater improvement among participants with more positive expectations regarding progress, return to usual job and change in condition than among those with more negative expectations was also noted for injury-specific functional status measures and the physical component of the SF-36 score but not the mental component of the SF-36 score.

figure 19FF1
Fig. 1: Mean pain grade at baseline and during the year following injury for workers with soft-tissue injuries, as a function of expected time until return to usual activities (measured at baseline). Error bars represent 95% confidence intervals.


For 3 of 4 measures of injured workers' recovery expectations, workers with more positive expectations spent less time off work and receiving workers' compensation benefits than those with more negative expectations, even when we controlled for other suspected prognostic factors.18 These findings suggest that recovery expectations are not simply an indirect measure of other factors but that they have a direct influence on the recovery process. Our study provides strong evidence regarding this relation according to quality criteria for prognostic studies.22 Our results are consistent with the 1 other prospective study of adequate methodologic quality.5

Despite these strengths, some caveats remain. Stopping receiving 100% wage-replacement benefits may not always be associated with return to work, as other administrative reasons may arise (e.g., noncooperation with medical treatment or with return to work plans). With respect to self-reported outcomes (pain grade, functional status and the physical component of the SF-36 score), loss to follow-up was substantial. For one of the recovery expectation questions, loss to follow-up was associated with baseline values, which indicated a selection bias. Potential differences in the “administrative” (i.e., the OWCB view) and “clinical” course of compensable soft-tissue injuries may not be fully documented here, although the similar role of injured workers' expectations in predicting recovery for both sets of outcomes is reassuring. Furthermore, an important number of injured workers were uncertain about their recovery. Future research into how recovery expectations are established would be useful. The limited associations between the various measures of recovery expectations and their independent explanation of various health outcomes suggest that the questions were not equivalent and did not represent a single construct. Some of these differences may be due to how the responses were scaled. For example, the lack of prediction by the variable “expectations regarding return to usual job” may have been due to the limited range of responses possible (3 versus 5 for progressing as expected [Table 2]). Alternatively, such a question may be more useful later on in the recovery process, when recovering workers are more open to returning to work. Reliability testing of the questions has not been done and may be difficult given the time-varying nature of recovery expectations. Nor has validity testing been carried out to assess overlap with other constructs in the recovery expectations literature1 or with fear-avoidance beliefs.23 Hence, further refinements of the questions to include comparable scaling and reliability and validity testing are warranted. Trials of assessment and counselling interventions to modify expectations and improve outcomes could then be conducted to test experimentally the role of changed recovery expectations.

Although injured workers may largely be reflecting prognostic information that their treating clinician has shared with them, they may also be providing their judgement of a variety of relevant factors. In the clinical setting, patients' recovery expectations may indicate the need for further probing of the unique constellation of factors — personal, social and work-related — that contribute to a person's recovery. Clinicians should thus explore potential barriers to recovery among injured patients who expect that recovery will be slow, who feel they are not doing as well as they had expected or who are uncertain regarding their eventual return to work. Listening to injured workers' expectations may not only be therapeutic in the clinician–patient encounter but also uncover barriers to return to work that the patient and the physician could deal with together.24


This project was sponsored by the Institute for Work & Health, an independent, not-for-profit research organization that receives support from the Ontario Workplace Safety & Insurance Board.


‡A list of members of the group appears at the end of the article.

This article has been peer reviewed.

Contributors: Drs. Cole and Mondloch were responsible for conceiving the study, collecting, analyzing and interpreting the data, and writing and revising the manuscript. Dr. Hogg-Johnson was responsible for conceiving and designing the study, interpreting the data and revising the manuscript.

Competing interests: None declared.

Correspondence to: Dr. Donald C. Cole, Institute for Work & Health, 800-481 University Ave., Toronto ON M5G 2E9; ac.no.hwi@elocd


1. Sullivan TJ, Frank J. Restating disability or disabling the state: four challenges. In: Sullivan TJ, editor. Injury and the new world of work. Vancouver: University of British Columbia Press; 2000. p. 3-24.
2. Frank JW, Brooker AS, DeMaio SE, Kerr MS, Maetzel A, Shannon HS, et al. Disability resulting from occupational low back pain. Part II: What do we know about secondary prevention? A review of the scientific evidence on prevention after disability begins. Spine 1996;21:2918-29. [PubMed]
3. Cole DC, Hudak PL. Prognosis of nonspecific work-related musculoskeletal disorders of the neck and upper extremity. Am J Ind Med 1996;29:657-68. [PubMed]
4. Miller M. The main feature: the employee. In: Isernhagen SJ, editor. Work injury: management and prevention. Rockville (MD): Aspen Publishers; 1988. p. 347-54.
5. Sandstrom J, Esbjornsson E. Return to work after rehabilitation. The significance of the patient's own prediction. Scand J Rehabil Med 1986;18:29-33. [PubMed]
6. Mondloch MV, Cole DC, Frank JW. Does how you do depend on how you think you'll do? A systematic review of the evidence for a relation between patients' recovery expectations and health outcomes. CMAJ 2001;165(2):174-9. [PMC free article] [PubMed]
7. Sinclair SJ, Hogg-Johnson SA, Mondloch MV, Shields SA. The effectiveness of an early active intervention program for workers with soft tissue injuries: the Early Claimant Cohort study. Toronto: Institute for Work & Health; 1997. [PubMed]
8. Sinclair SJ, Hogg-Johnson SA, Mondloch MV, Shields SA. The effectiveness of an early active intervention program for workers with soft tissue injuries: the Early Claimant Cohort study. Spine 1997;22:2919-31. [PubMed]
9. Atlas SJ, Deyo RA, Keller RB, Chapin AM, Patrick DL, Long JM, et al. The Maine Lumbar Spine Study, Part II. 1-year outcomes of surgical and nonsurgical management of sciatica. Spine 1996;21:1777-86. [PubMed]
10. Philips HC, Grant L. Acute back pain: a psychological analysis. Behav Res Ther 1991;29:429-34. [PubMed]
11. Tarasuk V, Eakin JM. “Back problems are for life”: perceived vulnerability and its implications for chronic disability. J Occup Rehabil 1994;4(1):55-64. [PubMed]
12. Von Korff M, Ormel J, Keefe FJ, Dworkin SF. Grading the severity of chronic pain. Pain 1992;50(2):133-49. [PubMed]
13. Ware JE Jr, Snow KK, Kosinski M, Gandek B. SF-36 health survey manual and interpretation guide. Boston: The Health Institute; 1993.
14. Ware JE Jr, Kosinski M, Keller S. SF-36 physical and mental health summary scales: a user's manual. Boston: The Health Institute, New England Medical Center; 1994.
15. Richards RR, An KN, Bigliani LU, Friedman R, Gartsman JM, Gristina AG, et al. A standardized method for the assessment of shoulder function. J Shoulder Elbow Surg 1994;3(6):347-52. [PubMed]
16. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt J. Validation study of WOMAC: a health status instrument for measuring clinically-important patient-relevant outcomes following total hip or knee arthroplasty in osteoarthritis. J Orthop Rheumatol 1994;1:95-108. [PubMed]
17. Roland M, Morris R. A study of the natural history of back pain — part I: development of a reliable and sensitive measure of disability in low-back pain. Spine 1983;8:141-4. [PubMed]
18. Hogg-Johnson S, Cole DC, and the Early Claimant Cohort Design Group, Prognostic Modelling Workgroup. Early prognostic factors for duration on benefits among workers with compensated occupational soft tissue injuries. Toronto: Institute for Work & Health; 1998. Institute for Work & Health Working Paper no 64R1.
19. Employment and Immigration Canada. National Occupational Classification: occupational descriptions. Ottawa: Canada Communications Group; 1993.
20. Employment and Immigration Canada. National Occupational Classification: career handbook. Ottawa: Canada Communications Group; 1996.
21. Schemper M. The relative importance of prognostic factors in studies of survival. Stat Med 1993;12(24):2377-82. [PubMed]
22. Hudak PL, Cole DC, Haines AT. Understanding prognosis to improve rehabilitation: the case of lateral elbow pain. Arch Phys Med Rehabil 1996;77:586-93. [PubMed]
23. Waddell G, Newton M, Henderson I, Somerville D, Main CJ. A fear-avoidance beliefs questionnaire (FABQ) and the role of fear-avoidance beliefs in chronic low back pain and disability. Pain 1993;52:157-68. [PubMed]
24. Frank JW, Sinclair S, Hogg-Johnson SA, Shannon HS, Bombardier C, Beaton DE, et al. Preventing disability from work-related low-back pain: new evidence gives new hope — if we can just get all the players onside. CMAJ 1998;158(12): 1625-31. [PMC free article] [PubMed]

Articles from CMAJ : Canadian Medical Association Journal are provided here courtesy of Canadian Medical Association
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...