Randomized comparison of a multidisciplinary job-retention vocational rehabilitation program with usual outpatient care in patients with chronic arthritis at risk for job loss




Work disability is a major consequence of inflammatory rheumatic conditions. Evidence regarding the effectiveness of interventions aimed at the prevention or reduction of work disability in rheumatic diseases is limited. We conducted a randomized controlled trial to investigate the effectiveness of a multidisciplinary job-retention vocational rehabilitation (VR) program in patients with a rheumatic condition who were at risk for job loss.


A total of 140 patients with a chronic rheumatic condition were randomly assigned to either a multidisciplinary job-retention VR program (n = 74) or usual outpatient care (UC) (n = 66). Patients in the VR group were assessed and guided by a multidisciplinary team, whereas patients in the UC group received care as initiated by their rheumatologist, supplemented with written information. The main outcome measure was the occurrence of job loss (complete work disability or unemployment); additional outcome measures included job satisfaction, pain, functional status, emotional status, and quality of life.


There was no difference between the 2 groups regarding the proportion of patients having lost their job at any time point, with 24% and 23% of the patients in the VR and UC groups, respectively, having lost their job after 24 months. Over the total period of 24 months, patients in the VR group had a significantly greater improvement of the fatigue visual analog scale and of emotional status (all P values < 0.05).


A job-retention VR program did not reduce the risk of job loss but improved fatigue and mental health in patients with chronic rheumatic diseases at risk for job loss.


The prevalence of work disability among persons with chronic rheumatic diseases is high. In patients with rheu- matoid arthritis (RA), work disability rates vary between 25% and 50% after 10 years of disease and increase to 90% in patients with longer disease duration (1–4). Work disability is also substantial in patients with other rheumatic conditions, such as ankylosing spondylitis (AS) (5, 6) and systemic lupus erythematosus (SLE) (7).

Costs ensuing from work disability account for a large part of the total costs associated with rheumatic conditions (8, 9). In addition to the economic consequences of work disability, the noneconomic impact on a person and his or her family may be substantial. Work disability was found to be associated with lower levels of self esteem, life satisfaction, and perceived health status and higher levels of depression and pain (10–14).

Given the large impact of work disability, work retention issues have been identified as one of the aims of the treatment of patients with rheumatic conditions (15). In the United States and European countries, vocational rehabilitation programs are being offered to patients with the goal of preventing the loss of paid employment or returning patients to work. In contrast with the many studies on factors associated with work disability (4, 16), the number of publications reporting the results of vocational rehabilitation programs is limited (17–19). The results of the few available studies, the majority of which had an uncontrolled design, indicate an overall positive effect on vocational status (18). A recent randomized controlled trial on the effectiveness of a job-retention vocational rehabilitation program (2 1.5-hour sessions) in patients with rheumatic diseases demonstrated that such an intervention delayed and reduced job loss (17). That study did not include outcome measures reflecting the impact of the vocational rehabilitation program on quality of life.

The aim of the present study was to evaluate the effectiveness of a multidisciplinary job-retention vocational rehabilitation program on the prevention of job loss and on quality of life. For that purpose, we conducted a multicenter, randomized controlled trial among patients with chronic rheumatic diseases who were in paid employment and at risk for job loss.


Study participants.

Between March 1999 and June 2001, patients were recruited at the outpatient rheumatology departments of Leiden University Medical Center and 10 nonacademic hospitals within the region of Leiden, the Netherlands. Participants were between 18 and 63 years of age and had a chronic rheumatic disease (diagnosis of RA, AS, psoriatic arthritis, reactive arthritis, SLE, or scleroderma) (20–22). All patients had a paid job (working full time or part time or on sick leave, either with or without a partial disability pension) and were having a self-perceived, disease-related problem at work that threatened their ability to work. This situation was verified by asking every potential participant, “Do you have concerns that your rheumatic condition-related problems at work may result in job loss?” Exclusion criteria were reaching the pension-eligible age within 2 years or having another disease or situation influencing work ability. The medical ethics committees of all participating hospitals approved this trial and all patients gave written informed consent.

Study protocol.

This study was a randomized controlled trial comparing a job-retention vocational rehabilitation program (VR group) with usual outpatient care (UC group), with 24 months of followup. After enrollment and baseline assessments had been completed, participants were randomly allocated to either the multidisciplinary job-retention vocational rehabilitation program or to usual outpatient care. Randomization was done with stratification for center (academic hospital versus nonacademic hospital) and 3 diagnosis groups (RA; AS, psoriatic arthritis, or reactive arthritis; and SLE or scleroderma), according to a randomization list that was created by a random digit generator. All clinical assessments were performed by a trained research nurse who was blinded to the patients' treatment status. Assessments were performed at baseline and after 6, 12, 18, and 24 months of followup. To maintain allocation concealment, patients were instructed not to inform the principal investigator or the research nurse about the type of care they received.


The job-retention vocational rehabilitation program has been previously described in detail (23). In brief, the program was delivered at the department of rheumatology of the Leiden University Medical Center by a multidisciplinary team comprising a rheumatologist, a social worker, a physical therapist, an occupational therapist, and a psychologist. Moreover, an occupational physician who was linked to the occupational health service of the Leiden University Medical Center was connected to the team. This occupational physician was not involved in the guidance of individual patients, but had a general advisory role. The organization of the program was in the hands of a coordinator. All patients made at least 2 visits to the hospital in connection with the job-retention vocational rehabilitation program.

The intervention consisted of a systematic assessment followed by education, vocational counseling, guidance, and medical or nonmedical treatment. The basic assessment was performed by a rheumatologist (current level of disease activity and joint destruction, presence of extraarticular manifestations or comorbidity, and extent and severity of activity limitations; prognosis regarding future impairments and activity limitations) and the coordinator (education level and previous jobs; systematic registration of the problems encountered in the current working situation, using a list of potential challenges and psychosocial situations). If necessary, additional team members were asked to see the patient to gather more information about specific aspects of the work situation. Depending on the specific problems of the individual patient, the intervention further consisted of education (such as providing written and oral information about the Dutch social security system regarding sick leave and work disability), counseling and guidance (such as the identification of resources for adapting the work environment or work hours, promotion of work self efficacy), or treatment (such as adaptation of the medical treatment in consultation with the referring rheumatologist, exercise therapy, occupational therapy, functional training of relevant activities, or mental restoration).

All information concerning the patient's health status; work situation and work challenges; and the course of the process of education, counseling, guidance, or treatment was listed in a final report, which was then sent to the referring rheumatologist and the occupational physician connected with the patient's company if applicable. The total duration of the intervention varied, and lasted on average between 4 and 12 weeks.

Patients assigned to the UC group were treated and referred to other health professionals in relation to their working problem if regarded necessary by their rheumatologist. In addition, all patients received the same written information about the Dutch social security system regarding sick leave and work disability as patients in the VR group.

The referring rheumatologists were informed about the treatment allocation. In both groups, physicians had free choice with respect to their medical prescriptions and other treatment strategies. All medical treatment and the use of health services during the intervention period and 2-year followup were recorded in both groups.

Sociodemographic and disease characteristics and the use of health care services.

The following variables were recorded at baseline: age, sex, status of living (living with a partner yes/no), diagnosis, and disease duration. Comorbidity was measured with the index created by Charlson et al (24) and categorized as not present (Charlson index = 0) or present (Charlson index > 0).

Education level was divided into 3 categories based on the Dutch school system: primary education (0–8 years), secondary education (9–16 years), and higher vocational education/university (≥17 years). Information about the job characteristics included the level and type of objective physical and mental demands at work (25). Category 1 is characterized by predominantly mental demands and no physical demands; category 2 by occupations with a combination of physical effort (light or heavy; standing, walking, lifting, high physical strain on the low back) and mental effort; category 3 by light physical demands (standing, walking, lifting of light objects); and category 4 by heavy physical demands (lifting of heavy objects, handling of heavy tools, and stooping frequently in combination with standing or walking). In addition, the presence of material or immaterial adaptations at the workplace was recorded (yes/no).

Current sick leave was recorded, with sick leave defined as being absent from work as officially reported to the employer. At the time the study was conducted and according to the Dutch social security system, employees who were on full or partial sick leave for more than 1 year are entitled a full or partial work disability pension if permanent work disability had officially been determined. If a person was judged to be ≥80% impaired, this person was entitled full work disability, whereas those who were 15–80% impaired were entitled partial work disability. In the Dutch social security system, it is possible to receive a partial work disability pension and to remain in paid employment on a part-time basis, in which situation partial or complete sick leave may occur.

The use of health care services and visits to different health professionals, such as an occupational therapist or the clinical nurse specialist, were measured using a 3-month diary.

Endpoint measures.

The main outcome was the occurrence of job loss, defined as receiving an official full work disability pension or unemployment. The classification of job losses was based on the participants' records of their work status at every followup visit. Patients who were on full sick leave <1 year were classified as being in paid employment. In addition to job loss, the number of patients whose extent of disability pension had increased (by receiving an official full disability pension or by receiving a new or a larger official partial disability pension) was recorded at every time point.

Secondary outcome measures were satisfaction with the job, pain, fatigue, physical functioning, and quality of life. Satisfaction with the job was measured on a horizontal 10-cm visual analog scale (VAS). The anchor on the left indicated that patients were not at all satisfied and the anchor on the right indicated that they were fully satisfied with the job. The VAS was only to be filled in by patients who had worked at least 5 days in the last month.

The patient's global assessments of pain and fatigue were measured on a 10-cm VAS. The anchors on the left indicated no pain and no fatigue and the anchors on the right indicated severe pain and severe fatigue. To assess physical functioning, the Health Assessment Questionnaire (HAQ), a 20-item questionnaire comprising 8 domains of activities of daily living (26), was included.

Anxiety and depression were measured by means of a Dutch version of the Hospital Anxiety and Depression Questionnaire (HADS) (27), which contains two 7-item scales, one for anxiety and one for depression, both with a score range of 0–21.

Quality of life was measured using the RAND 36-item Health Survey (28). The RAND was converted into 2 summary scales: the physical and mental component summary scales. The RAND includes the same items as the Medical Outcomes Study Short Form 36 (SF-36), and although the scoring procedures are somewhat different, the effects on final scores are minimal (28).

Statistical analysis.

The sample size was calculated to allow detection of a 20% difference between the 2 groups. Assuming 10% job loss in the VR group and 30% job loss in the UC group, with 80% power based on a 2-sided test with a significance level of 0.05, 63 patients per group would be needed to detect a significant difference. Considering a dropout rate of 10%, 140 patients in total would be needed for the present study. Data management was performed using the Project Manager Software package version 6.1 (29). Data were automatically and integrally converted to SPSS 11.5 for Windows (SPSS, Chicago, IL) for statistical analysis.

Baseline characteristics and baseline values of outcome measures were compared using the Mann-Whitney U test, unpaired Student's t-test, or chi-square test where appropriate. The primary analyses of effectiveness were based on intention to treat as initially assigned. All available data were used. As a secondary analysis, a per-protocol analysis was performed, comparing the patients who did actually receive the treatment in the VR group with the patients in the UC group.

Regarding the primary outcome measure job loss, proportions of patients in both groups were compared at each time point using a chi-square test. A logistic regression model with time, randomization group, and a random person effect was used to compare the overall change in percentages over time between the groups (test for interaction between time and randomization group). The same procedure was followed to compare proportions of patients whose extent of disability pension had increased. To investigate the presence of subgroups of patients who would or who would not benefit from the intervention, tests for interaction between randomization group and age, diagnosis at baseline, and the presence of sick leave at baseline were performed in the logistic model.

With respect to the secondary outcomes, changes from baseline within each group with the 95% confidence interval were computed at each time point. Change scores were compared between the 2 groups with an unpaired Student's t-test. A linear mixed model with time as covariate was used to compare the difference in trend over the total 2-year followup period.


A total of 196 patients were assessed for eligibility. Fifty-six patients were excluded because they did not meet the inclusion criteria (n = 35), refused to participate (n = 12), or could not enter the study for other reasons (n = 9).

Of the 74 patients randomized to the VR program, 10 (14%) did not take part in the intervention (protocol violations) for various reasons: hospital visits too troublesome (n = 4), hospital admission (n = 2), myocardial infarction (n = 1), new job (n = 1), and unknown (n = 2). Four of these 10 patients did not show up at any of the followup visits (lost to followup at 6 months). Over the period of 2 years, 12 participants allocated to the UC group and 13 participants allocated to the VR group withdrew from the trial (Figure 1). Reasons for withdrawal were moving out of the area (n = 5), personal and family matters (n = 7), not responding to our repeated telephone calls (n = 3), death from a heart attack (n = 1), time constraints (n = 1), loss of interest (n = 5), or other reasons (n = 3). The baseline sociodemographic and disease characteristics of the 115 patients who completed the study did not differ from those of the 25 patients who withdrew (data not shown).

Figure 1.

Allocation of study participants and study withdrawal.

The baseline sociodemographic and disease characteristics of the 140 study participants are shown in Table 1. There were no statistically significant differences in any of the characteristics between the 2 groups.

Table 1. Baseline sociodemographic and clinical characteristics of 140 patients with chronic arthritis in a randomized controlled trial comparing a multidisciplinary job-retention vocational rehabilitation program with usual care*
 Vocational rehabilitation (n = 74)Usual care (n = 66)P
  • *

    Values are presented as number (percentage) unless otherwise indicated. SLE = systemic lupus erythematosus.

  • Chi-square test or Mann-Whitney U test where appropriate.

Age, median (range) years43 (21–57)44 (24–58)0.86
Female41 (55)38 (58)0.80
Living with partner61 (82)48 (73)0.18
 Rheumatoid arthritis34 (46)36 (55)0.54
 Ankylosing spondylitis, psoriatric arthritis, or reactive arthritis17 (23)12 (18) 
 SLE, scleroderma23 (31)18 (27) 
Duration of disease, median (range) months11.0 (0–158)19.5 (0–174)0.60
Comorbidity present (Charlson index >0)32 (43)28 (42)0.88
Education level   
 High15 (20)10 (15)0.54
 Medium37 (50)39 (59) 
 Low22 (30)17 (26) 
Current occupational category   
 Mental demands20 (28)24 (36)0.41
 Mixed mental/physical demands15 (20)13 (20) 
 Light physical demands20 (27)19 (29) 
 Heavy physical demands19 (25)10 (15) 
Adaptations at work due to rheumatic disease22 (29)15 (23)0.35
Partial work disability benefit12 (16)11 (17)0.94
Sick leave42 (57)35 (53)0.66
Complete sick leave21 (50)20 (57)0.80
Duration of sick leave in weeks, median (range)16 (1–52)18 (3–48)0.80
Duration of sick leave >6 weeks29 (39)28 (42)0.70
Duration of sick leave >40 weeks6 (8)2 (3)0.17

The use of health services during the intervention period and 2-year followup.

Over the first 6 months of the study as well as during the total 2 years of followup, no significant difference between the 2 groups was found with respect to the mean number of visits to the rheumatology nurse specialist, occupational therapist, physical therapist, social worker, psychologist, or the occupational physician (data not shown). However, patients in the UC group paid more visits to the rheumatologist in the first 6 months of the study (mean ± SD 2.8 ± 2.0) compared with the patients in the intervention group (mean ± SD 1.5 ± 1.9; P < 0.001).

Permanent job loss and increase in disability pension.

Over the total followup period, job loss occurred in both groups, predominantly in the first 12 months of followup (Table 2). All job losses were related to the rheumatic disease, and patients who lost their jobs could be classified as receiving a full work disability pension. None of the patients became unemployed for other reasons. There was no statistically significant difference in the proportion of patients with permanent job loss between the groups at any time point. Moreover, the mixed effects logistic regression model did not indicate a different trend over time between the 2 groups (test for interaction between time and intervention group: P = 0.13, test for main group effect: P = 0.86). In a secondary per-protocol analysis comparing the 64 patients who did receive the treatment in the VR group with the 66 patients in the UC group, there was no statistically significant difference in the proportion of patients with job loss between the groups at any time point or over the total followup period (data not shown).

Table 2. Job loss and increase in official disability pension (cumulative) in 140 patients with a rheumatic condition randomized to a multidisciplinary job-retention vocational rehabilitation program or usual care*
 Vocational rehabilitation (n = 74)Usual care (n = 66)P
  • *

    Values are the number/total (percentage).

  • Chi-square test.

Job loss   
 6 months6/66 (9)3/59 (5)0.39
 12 months12/64 (19)11/58 (19)0.97
 18 months11/59 (19)13/55 (24)0.51
 24 months14/59 (24)12/53 (23)0.89
Job loss or increase in official disability pension   
 6 months14/66 (21)4/59 (7)0.02
 12 months26/64 (41)19/58 (33)0.37
 18 months26/59 (44)23/55 (42)0.81
 24 months31/59 (53)23/53 (43)0.33

With respect to the primary outcome measure job loss, there was no significant interaction between randomization group and age, diagnosis at baseline, and the presence of sick leave at baseline as performed in a logistic model. Regarding deterioration of the working situation defined as either full work disablement or institution or increase of a partial disability pension, initially more patients in the VR group than in the UC group became either fully work disabled (job loss) or, to a greater extent, partially work disabled at 6 months of followup (P = 0.02). However, after 12 months this difference disappeared. Regarding this endpoint, over the entire period there was no statistically significant difference between the 2 groups (test for interaction between time and randomization group: P = 0.09, test for main effect: P = 0.27).

In contrast with the 54 patients whose disability pension decreased, 7 patients improved (3 in the VR group and 4 in the UC group). Two patients, both in the UC group, who were partially work disabled at baseline did not receive a disability pension anymore after 12 months of followup. In addition, 4 patients who were fully work disabled became partially work disabled (3 from the VR group after 12, 18, and 24 months and 1 from the UC group after 24 months of followup), whereas 1 patient in the UC group who was fully work disabled did not receive a disability pension after 12 months of followup.

Job satisfaction, physical and mental functioning, and quality of life.

Over the total 24-month followup period, patients in the VR group showed a significantly greater improvement in fatigue, the HADS depression and anxiety subscales, and mental health as measured by the RAND 36 summary scale mental health than patients in the UC group (Table 3). Moreover, there was a trend towards a greater improvement in job satisfaction in the VR group. Pain, functional ability, and physical health did not differ between the 2 groups over time.

Table 3. Clinical outcome data at baseline and change scores from baseline in 140 patients with chronic arthritis at risk for job loss randomized to a multidisciplinary job-retention vocational rehabilitation program or usual care*
 Baseline, mean ± SD6 months, mean (95% CI)12 months, mean (95% CI)18 months, mean (95% CI)24 months, mean (95% CI)P
  • *

    95% CI = 95% confidence interval; VAS = visual analog scale; VR = vocational rehabilitation program; UC = usual care; HAQ = Health Assessment Questionnaire; HADS = Hospital Anxiety and Depression Score; RAND SSC = RAND summary scale.

  • P > 0.05 for all differences between baseline values, unpaired Student's t-test.

  • Linear mixed model.

  • §

    VAS job satisfaction was only filled in by those patients who worked ≥5 days in the past month. Numbers of patients were 58, 46, 46, 37, and 37 (at baseline, 6, 12, 18, and 24 months of followup) for the intervention group and 46, 39, 41, 32, and 37 for the usual care group, respectively.

  • P < 0.05, unpaired Student's t-test.

Job satisfaction§ (VAS 0–10)      
 VR5.57 (2.55)0.18 (−0.92, 1.27)1.78 (0.85, 2.70)1.65 (0.55, 2.74)2.00 (1.25, 2.75)0.12
 UC5.53 (2.55)0.15 (−0.80, 1.10)0.53 (−0.48, 1.55)0.24 (−0.96, 1.45)0.88 (−0.33, 2.11) 
Pain (VAS 0–10)      
 VR4.37 (2.31)−0.70 (−1.40, 0.01)−0.31 (−1.08, 0.47)−0.43 (−1.19, 0.32)−0.59 (−1.28, 0.09)0.85
 UC4.71 (2.27)−0.20 (−0.81, 0.41)−0.58 (−1.28, 0.13)−0.33 (−1.00, 0.34)−0.42 (−1.16, 0.32) 
Fatigue (VAS 0–10)      
 VR6.11 (2.42)−0.23 (−0.92, 0.47)−0.58 (−1.29, 0.14)−0.48 (−1.20, 0.25)−1.23 (−1.91, 0.54)0.04
 UC5.43 (2.74)0.11 (−0.53, 0.75)−0.55 (−1.38, 0.28)−0.05 (−0.88, 0.77)−0.15 (−1.03, 0.73) 
HAQ (0–3)      
 VR0.76 (0.50)0.03 (−0.08, 0.13)−0.04 (−0.15, 0.06)0.00 (−0.11, 0.11)−0.01 (−0.14, 0.12)0.43
 UC0.83 (0.55)−0.04 (−0.16, 0.08)−0.07 (−0.19, 0.05)0.08 (−0.04, 0.21)−0.10 (−0.23, 0.03) 
HADS anxiety      
 VR7.20 (4.00)−0.30 (−1.08, 0.48)−0.83 (−1.78, 0.11)−0.94 (−1.87, −0.020)−1.83 (−2.86, −0.80)0.01
 UC6.80 (4.10)−0.43 (−1.39, 0.54)−0.25 (−1.37,0.89)−0.34 (−1.53, 0.89)−0.03 (−1.26, 1.34) 
HADS depression      
 VR6.10 (3.30)−0.02 (−1.05, 1.01)−0.46 (−1.50, 0.57)−0.64 (−1.71, 0.44)−1.66 (−2.72, −0.60)0.04
 UC5.70 (3.50)0.28 (−0.54, 1.10)0.02 (−0.89, 0.92)−0.21 (−1.36, 0.93)0.15 (−1.12, 1.42) 
RAND SSC physical health      
 VR40.64 (17.66)5.75 (−0.45, 11.95)13.6 (7.04, 20.18)13.78 (6.32, 21.25)13.72 (6.73, 20.71)0.63
 UC43.32 (19.03)5.96 (0.38, 11.53)11.7 (5.04, 18.39)9.32 (2.75, 15.90)11.69 (5.36, 18.02) 
RAND SSC mental health      
 VR59.59 (24.08)−1.40 (−8.40, 5.54)5.31 (−1.99, 12.61)11.20 (2.40, 20.06)13.61 (6.61, 20.60)0.01
 UC64.10 (23.31)1.72 (−5.05, 8.50)3.33 (−4.42, 11.08)3.60 (−4.78, 12.00)2.16 (−5.30, 9.62) 


The results of this randomized controlled trial of patients with rheumatic diseases who were at risk for job loss showed that participating in a vocational rehabilitation program had no effect on remaining in paid employment. However, there was a significant effect on fatigue and mental health as compared with the usual care.

To our knowledge, this is the second randomized controlled trial investigating the effectiveness of a vocational rehabilitation program for patients with rheumatic diseases at the level of job-loss prevention. In contrast to our study, Allaire and coworkers (17) found that job loss was significantly delayed and reduced among study participants who received a job-retention vocational rehabilitation intervention.

There may be several possible explanations for this discrepancy. First, there may have been differences regarding the components and execution of the intervention. Although job accommodation, vocational counseling and guidance, education, and self advocacy were elements of the interventions in both the study by Allaire et al and the present study, the focus and intensity may have varied. Moreover, the program as provided in the study by Allaire et al was conducted in connection with an ongoing state vocational rehabilitation program, whereas in the present study the intervention was delivered in a health care setting. In the Dutch health care and social security system, the occupational physician plays an important role in the process of vocational rehabilitation. The occupational physician is linked to occupational health services, with which all companies have been legally obliged to have a contract since January 1998. However, the cooperation between occupational physicians and other health professionals, including our multidisciplinary vocational rehabilitation team, has previously been found to be an important but often troublesome element in the vocational guidance of patients with a health-related problem at work (15, 23, 30–32).

A second explanation for the difference in the results of the study by Allaire et al and our study may be that differences in the contrast between the vocational rehabilitation program and usual care may have occurred between the 2 studies. In our study, patients were directly referred for participation in the trial by the rheumatologist, who was thus aware of the working problems the patient encountered. Moreover, the rheumatologist was informed about the treatment allocation in a later stage, another factor that could have induced enhanced treatment or referrals in connection with the work problem in the UC group. Indeed, patients in the UC group initially paid more visits to the rheumatologist than patients in the VR group. The patients' participation in the trial could have made rheumatologists aware of their patients' problems at work, and if a patient was allocated to the UC group rheumatologists might have thought they needed to act on account of their patients. In addition, it is possible that patients who were allocated to the control group made an extra appointment with their rheumatologist to discuss their working problem and potential solutions. In the study by Allaire et al, the connection between regular rheumatologic care and the trial appeared to be less close (17).

A third explanation for the discrepancy between the results of the 2 trials may be that the populations studied were different. In general, patients in the study by Allaire et al were ∼5 years older, were more often women, and had better functional status as measured with the HAQ than the patients in the present study. Moreover, there may have been differences in the severity of the working problems. In our study, >40% of the patients in both groups were on sick leave at baseline, many of them longer than 6 weeks. Long-term sick leave usually indicates substantial limitations in work capacity and often precedes permanent work disability. At the time the study was conducted, the genuine setting of vocational rehabilitation plans by the occupational physician in collaboration with the patient and the employer was often postponed until the medical examination for a work disability pension was imminent at 12 months of sick leave, making job loss unavoidable. Although patients in our study were motivated to stay in the work force, with a relatively long period of sick leave, individuals may have lost confidence in their own capacity for employment and accepted their inability to work. A relatively long duration of sick leave may also have played a role in the initial excess job loss in the VR group. In our study, there were 6 patients in the VR group and 2 patients in the UC group with a sick-leave duration >40 weeks. Although this difference did not reach statistical significance, it is conceivable that the few extra patients in the VR group with a relatively long duration of sick leave at the start of the study could explain the initial excess job loss in the VR group. Overall, it could be that for those patients with a relatively long duration of sick leave, the intervention was provided too late to make a difference. Only recently the Dutch occupational health law has changed, and employees on sick leave are now seen by the occupational physician in the first 6 weeks of sick leave.

Despite the fact that our study did not demonstrate a quantitative effect regarding the prevention of job loss, a beneficial effect of the vocational rehabilitation program on fatigue and mental health was found. Fatigue has been described as a persistent disease-related threat to employment (23, 33, 34). To cope with fatigue, patients can make a number of job accommodations such as altering work hours, taking more and shorter breaks, working at home, delegating specific tasks, or making adaptations aimed at conserving energy in their personal lives to save themselves for the job. These changes take time and may not have a direct effect on the short-term working situation. Two studies (35, 36) demonstrated the relationship between fatigue and health-related quality of life as measured with the SF-36 in patients with chronic arthritis. Fatigue, general health, physical health, and mental health went hand in hand with diminished work productivity and work quality.

In addition to the beneficial effect on fatigue and mental health, a trend towards greater satisfaction with the job for patients who remained in the work force was seen. This positive trend might have reached statistical significance if the study sample had been larger. However, the considerable dropout rate in the present study, which was larger than anticipated, has negatively affected the statistical power of this study.

Although the results of the present study did not confirm the positive effect of the previous study by Allaire et al, there is ample rationale for the future development and evaluation of vocational rehabilitation programs. First, work disability remains a major problem in patients with rheumatic diseases, and second, there are a number of starting points for the design of effective interventions. For the effectiveness of job-retention vocational rehabilitation programs, it is important that patients at risk for work disability are identified in an early stage. It has been found, however, that rheumatologists often do not recognize the working problems (15, 23, 33, 37), and the same might apply to other health professionals. Currently, a number of instruments to measure work disability have become available, such as the Work Limitations Questionnaire or the Work Instability Scale (38–40). The broad implementation of such instruments in the clinical setting of rheumatologic care, especially in connection with early arthritis clinics, deserves consideration.

Apart from its timing, the connection between the health care system and vocational rehabilitation systems needs to be further developed. With respect to the Dutch situation, the role of the occupational physician as a potential participant in the vocational rehabilitation process should be explained more clearly and more communication should take place in earlier phases of vocational guidance (23).

In conclusion, a job-retention vocational rehabilitation program did not reduce the risk of job loss but improved fatigue and mental health in patients with rheumatic diseases. With the development of vocational rehabilitation interventions, the provision of these services in early phases of the work problems and the collaboration between various health care professionals including occupational physicians, employers, and the patient/employer themselves deserve special attention.


The authors would like to thank research nurse Jacqueline Boon and all participating rheumatologists: A. J. Peeters and D. van Zeben, Reinier de Graaff hospital, Delft; M. L. Westedt and I. Speijer, Bronovo Hospital, The Hague; H. K. Ronday, Hospital Leyenburg, The Hague; J. A. P. M. Ewals, Rode Kruis Hospital, The Hague; G. Collée and W. M. de Beus, Medical Center Haaglanden, The Hague; J. Ph. Terwiel, A. Linssen, and C. Mallee, Spaarne Hospital and Kennemer Hospital, Haarlem; and J. M. Bok, Groene Hart Hospital, Gouda.