Article Text

Download PDFPDF

Original article
Adjuvant occupational therapy improves long-term depression recovery and return-to-work in good health in sick-listed employees with major depression: results of a randomised controlled trial
  1. Hiske L Hees,
  2. Gabe de Vries,
  3. Maarten W J Koeter,
  4. Aart H Schene
  1. Program for Mood Disorders, Department of Psychiatry, Academic Medical Center (AMC), University of Amsterdam, Amsterdam, The Netherlands
  1. Correspondence to Hiske Leona Hees, Program for Mood Disorders, Department of Psychiatry, Academic Medical Center (AMC), University of Amsterdam, Meibergdreef 5, Amsterdam 1105 AZ, The Netherlands; H.L.Hees{at}


Objectives To evaluate whether adjuvant occupational therapy (OT) can improve the effectiveness of treatment-as-usual (TAU) in sick-listed employees with major depression.

Methods In total, 117 employees sick-listed for a median duration of 4.8 months (IQR=2.6 to 10.1 months) because of major depression were randomised to TAU (n=39) or adjuvant OT (TAU+OT; n=78). OT (18 sessions) focussed on a fast return to work (RTW) and improving work-related coping and self-efficacy. The primary outcome was work participation (hours of absenteeism+duration until partial/full RTW). Secondary outcomes were depression, at-work functioning, and health-related functioning. Intermediate outcomes were work-related, coping and self-efficacy. Blinded assessments occurred at baseline and 6, 12 and 18 months follow-up.

Results The groups did not significantly differ in their overall work participation (adjusted group difference=−1.9, 95% CI −19.9 to +16.2). However, those in TAU+OT did show greater improvement in depression symptoms (−2.8, −5.5 to −0.2), an increased probability of long-term symptom remission (+18%, +7% to +30%), and increased probability of long-term RTW in good health (GH) (+24%, 12% to 36%). There were no significant group differences in the remaining secondary/intermediate outcomes.

Conclusions In a highly impaired population, we could not demonstrate significant benefit of adjuvant OT for improving overall work participation. However, adjuvant OT did increase long-term depression recovery and long-term RTW in GH (ie, full RTW while being remitted, and with better work and role functioning).

Trial Registration Dutch Trial Register NTR2057.

View Full Text

Statistics from

What this paper adds

  • Depression is one of the most prevalent causes of disability and leads to substantial negative effects in the workplace, incurring major costs for the individual, employer and society.

  • Clinical treatments have proven only moderately effective for reducing disability and remission; additional treatments may be needed.

  • In a highly impaired population, we could not demonstrate that addition of occupational therapy (OT) to treatment-as-usual (TAU) (adjuvant OT) significantly improves overall work participation. However, we did find that adjuvant OT increases the probability of returning to work (RTW) in good health (ie, full RTW while being remitted and with better work and role functioning).

  • Adjuvant OT also increased long-term depression recovery, both in terms of symptom reduction and sustainable remission.


Depression is one of the most prevalent causes of disability1 and leads to substantial negative effects in the workplace: Depressed employees have more absenteeism and at-work productivity loss than other employees,2 including those with chronic medical conditions, such as heart disease and rheumatoid arthritis.3 Characterised by an early age of onset,4 a recurrent-chronic course,5 and pervasive effects on social and cognitive functioning,6 depression is associated with substantial costs for the individual, and also for society.7

Standard clinical treatments have not proven sufficient for full recovery from depression8 or work-related disability.9 Although work is a central part of a person's life, standard clinical treatments often pay little attention to depression-related problems in the workplace. However, not being able to (fully) participate in the workplace may lead to social and economic deprivation, which in turn can have an added negative impact on the course of depression,10 initiating a downward spiral. Thus, improving standard treatment by adding interventions that specifically target the workplace may have an additional effect on reducing work-related disability and/or depression symptoms in sick-listed employees with depression.11

Surprisingly, a recent review9 found only one study that has evaluated such an occupational intervention specifically designed for sick-listed employees with major depression. This study,12 conducted by our research group, showed that adjuvant occupational therapy (TAU+OT) resulted in a higher reduction of work-loss days and was more cost-effective than treatment-as-usual (TAU) only.

Based on these encouraging results, we have developed a shorter and improved version of our earlier OT.13 This new OT adopts an increased focus on three elements found to be essential for a successful return to work (RTW): an early return to the work situation according to the ‘place-then-train’ principle,14 an increased focus on work-related coping15 and self-efficacy,16 and enhanced communication among the various stakeholders involved.17 In the present study, we aimed to evaluate the effectiveness of this new adjuvant OT (TAU+OT), when compared with TAU only, in employees who were sick-listed because of major depression.


This study was approved by the medical ethics committee of the Academic Medical Center in Amsterdam, The Netherlands (MEC 06/285), and is registered with the Dutch Trial Register (NTR2057). Written informed consent was obtained from all participants in the study.


Participants were eligible if they were aged 18–65 years, diagnosed with a major depressive disorder according to DSM-IV criteria, and absent from work for at least 25% of their contract hours due to their depression. In addition, the duration of their depressive disorder had to be at least 3 months, or the duration of their sickness absence had to be at least 8 weeks. Finally, there had to be a relationship between the depressive disorder and the work situation, that is, work was one of the determinants of depressive disorder and contributed substantially (>25%), or the depressive symptoms reduced productivity or hindered RTW. Participants with severe alcohol or drug dependence, bipolar disorder, psychotic disorder, depression with psychotic characteristics, or an indication of inpatient treatment were excluded from the study. Participants were referred by occupational physicians from several occupational health services in the Amsterdam area. After a telephone screening by a senior psychiatrist, participants received a 3-hour psychiatric intake, including the structured psychiatric interview structured clinical interview for DSM-IV axis I disorders.18 Participants who met the inclusion criteria and were willing to participate in the study were asked for written informed consent. More details regarding the study design and procedure can be found elsewhere.13


Treatment as usual

TAU consisted of treatment by psychiatric residents in an outpatient university clinic according to a treatment protocol consistent with the APA guidelines.19 Visits consisted of clinical management, including psychoeducation, supportive therapy and cognitive behavioural interventions. Therapies were supervised on a weekly basis by an experienced senior psychiatrist specialised in depression. If needed, participants received pharmacotherapy according to a protocollised algorithm. If the participant's condition deteriorated and outpatient treatment was no longer deemed adequate, he/she was referred to day treatment or inpatient treatment.

Occupational therapy

OT consisted of 18 sessions (nine individual sessions, eight group sessions and a meeting with the employer), and was conducted by two experienced occupational therapists who had received extensive training in the intervention protocol. During the intervention, the occupational therapist frequently communicated with the occupational physician and the resident treating psychiatric. Employees were required to work at least 2 hours per week when starting OT, so that employees were able to directly practise the things learned (eg, new coping strategies) during therapy. The content of OT has been described in detail elsewhere.13

Outcome measures

Primary outcome

Our primary outcome was work participation, defined in terms of absenteeism and time until partial/full RTW. Absenteeism was operationalised as the average number of hours of absenteeism over each 6-month period. Time until partial/full RTW was operationalised as the duration of sick leave due to depression in calendar days from the start of treatment until partial (or full) RTW. Partial RTW was defined as working an increment of at least 5 hours (compared with hours worked at baseline), for at least 4 weeks without partial or full recurrence. Full RTW was defined as working the full number of contract hours in own or other work for at least 4 weeks, without partial or full recurrence.

Secondary outcomes

Severity of depression was assessed by the Hamilton Rating Scale for Depression (HRSD), a semistructured clinical interview. A score of ≤7 is qualified as ‘normal’, 8–13 as ‘mild’, 14–18 as ‘moderate’, 19–22 as ‘severe’ and ≥23 as ‘very severe’.20 Thus, depression remission was defined as having a HRSD≤7. In addition, we used the self-report questionnaire Inventory of Depressive Symptoms-Self-Report, the IDS-SR,21 where a score of ≤13 is qualified as ‘normal’, 14–22 as ‘mild’, 23–30 as ‘moderate’, 31–38 as ‘moderate to severe’ and ≥39 as ‘severe’.

At-work functioning was assessed through weekly self-report records of work efficiency on a scale of 1 (‘not productive at all’) to 10 (‘very productive’). In addition, we examined three subscales of the Work Limitations Questionnaire (WLQ): ‘Output’ (ie, difficulties in completing the required amount of work), ‘Time’ (ie, difficulties in handling the job's time and scheduling demands) and ‘Mental-Interpersonal’ (ie, difficulties in handling the job's cognitive and social demands). Each WLQ scale was scored from 0 to 100, reflecting the percentage of time during which a participant experienced work limitations during the past 4 weeks.22

Health-related functioning was assessed with three subscales of the Medical Outcomes Study-Short Form (MOS-SF 36): ‘Mental health’, ‘Role limitations due to emotional problems’ (Role Emotional), and ‘Role limitations due to physical problems’ (Role Physical). Each scale ranged from 0 to 100, with higher scores reflecting higher levels of functioning.23

Intermediate outcomes: mechanisms of change

Coping with work-related situations was measured by an adapted version of the Utrecht Coping List (UCL).24 We included three UCL-subscales: ‘Active problem focussing’ (seven items; eg, being goal-directed, thinking of several solutions to a problem), ‘Avoidance Behaviour’ (eight items; eg, withdrawing from problematic situations), and ‘Passive Reaction’ (seven items; eg, worrying about the past and taking refuge in fantasies). All items were rated on a four-point scale, ranging from 1 (‘seldom or never’) to 4 (‘very frequently’). Work-related self-efficacy was measured by the 11-item questionnaire ‘Expectations regarding work resumption’. Items were rated on a five-point scale, with higher scores reflecting higher self-efficacy.25

In the study protocol,13 seven subscales from the ‘Perception and Judgment of the Working Situation’ questionnaire Vragenlijst Beleving en Beoordeling van de Arbeid (VBBA) were also described as intermediate outcomes. Considering that adjuvant OT primarily aimed to target coping and self-efficacy, and in order to limit the number of statistical tests, we decided (before starting the analyses) to only include coping and self-efficacy as intermediate outcomes in our final analyses.

Additional analyses

In additional analyses, we examined the percentage of participants who achieved a RTW in good health (RTW-GH), defined as having achieved a full RTW while being remitted from depression (HRSD≤7).

Questionnaires and clinical interviews were administered to employees at baseline and after 6, 12 and 18 months. For the WLQ and UCL, the baseline measure consisted of a retrospective assessment during the period before start of sickness absence. For all other measures, baseline scores referred to the period during study entry. Sickness absence data were derived from diaries that employees kept on a weekly basis during the 18-month study period.

Process outcomes

We recorded the number of sessions of OT, visits to the psychiatric outpatient department, primary care contacts (only those related to mental health), use of psychotropic medications, and days of psychiatric hospitalisation.

Sample size

Based on previous results,12 we expected to find a difference of 25% in hours of absenteeism between both groups. To achieve a power of 0.80 given a one-sided α of 0.05, an estimated effect-size of 0.30, and a ratio of our control sample to our experimental sample of 1 : 2, we needed at least 35 participants in the control condition and 70 participants in the experimental condition. Considering loss to follow-up, we included an extra 10% of participants, leading to a required sample size of 116 participants. Power calculations were made with G-power.26

Randomisation and blinding

Participants were randomised to either TAU or TAU+OT after baseline data had been collected and informed consent was obtained. Randomisation was conducted by an independent research assistant, using software based on a minimisation randomisation procedure.27 Randomisation was stratified for baseline depression severity (≤17 or ≥18) and number of previous depressive episodes (≤2 or ≥3). Study assessments were conducted by a psychiatrist and a researcher who were blind to group allocation. Due to the nature of our intervention, neither patients nor therapists could be blinded to the patients’ allocation status.

Statistical analysis

Data were analysed according to the intention-to-treat principle. Independent samples t tests and χ2 tests were used to evaluate potential differences in baseline characteristics between the experimental and control condition. All baseline measures with an effect size of at least d=0.3 between TAU and TAU+OT were combined into a propensity score,28 which was entered as a covariate in the analyses. To take potential biased outcomes caused by selective loss to follow-up into account, we used multiple imputation (five imputed datasets), which, assuming missing at random for missing values, gives unbiased results with correct SEs. Results of each of the five imputation sets were combined using Rubin's rules.29

Treatment effect

The primary outcome measure was analysed twofold: first, a Cox proportional hazard model was used to estimate HR for partial/full RTW with bootstrapping to account for the large variance in the outcome measure. Kaplan–Meier curves were used to describe the duration until partial/full RTW. Second, the reduction in hours of absenteeism was analysed with random coefficient regression analysis, using change scores between each follow-up assessment and baseline as dependent variables (ΔT0–T1, ΔT0–T2, ΔT0–T3). Treatment, time and the treatment by time interaction were entered as independent variables, the baseline (T0) and propensity score were entered as covariates. A model with change scores is the only way to include the change between T0 and T1 in the analyses, while still correcting for potential baseline differences. If a non-linear effect of time was suspected after examining the data, we added a quadratic main effect of time, and a quadratic time by group interaction as independent variables to the model. In order to account for individual trajectories of the dependent variable, we allowed the intercept (and slope) to vary randomly among participants. A random slope was only included if this improved the model fit.30 In these analyses, both the treatment main effect and/or the treatment by time interaction describe the effect of adjuvant OT.

For all continuous secondary outcomes and intermediate outcomes, similar analyses were conducted, with the exception of the WLQ and UCL. For these variables, the baseline measure concerned at-work functioning (WLQ) and work-related coping (UCL) before the start of sickness absence. This results in (a) variable time periods between ‘baseline’ (T0) and the first follow-up measure (T1) and (b) the impossibility to attribute changes between ‘baseline’ (T0) and first follow-up measure (T1) to the effect of adjuvant OT. For this reason, we did not include T0 (or change between T0 and T1) as a dependent variable in our model and, consequently, no change scores were necessary in order to be able to correct for baseline differences on the outcome variable. For these analyses, a follow-up vector with T1, T2, and T3 assessments was used as dependent variables. In these ‘more standard’ longitudinal analyses, only the interaction effect describes the effect of adjuvant OT.

Dichotomous outcomes (% of remission, % of RTW in GH) were analysed using population-averaged logistic regression analysis (Generalised Estimating Equations, GEE), with unstructured covariance matrices to allow for correlation in outcomes across time within participants. Independent variables were treatment group, time and the treatment group by time interaction. The propensity score was entered as covariate. All analyses were conducted with SPSS V.18 for Windows.


Figure 1 shows the progression through the trial. Between December 2007 and October 2009, 224 participants were screened for participation in the study. Of these, 107 were excluded and 117 participants were randomised to receive either TAU+OT (n=78) or TAU (n=39). The treatment groups were comparable at baseline, except for the number of contract hours (d=0.43), WLQ output scale (d=0.34), and HRSD scores (d=0.39; table 1). These baseline variables were included as a covariate in the propensity score in all further analyses.

Table 1

Baseline characteristics of participants randomised to treatment-as-usual (TAU) or adjuvant occupational therapy (TAU+OT)

Figure 1

Participant flow (TAU, treatment-as-usual; OT, occupational therapy).

Overall, our sample consisted of a highly impaired population: At baseline, more than two-thirds of participants (69% in TAU+OT; 68% in TAU) were depressed for more than 6 months, and a high proportion of participants (67% in TAU+OT; 56% in TAU) were absent from work for more than 3 months. In addition, more than half the participants (53% in TAU+OT; 54% in TAU) had had at least one previous depressive episode. At baseline, both groups scored well below the Dutch population average on all three SF-36 scales: Mental Health (−2.5 SDs below the population average), Role Emotional (−2.2SD), and Role Physical (−1.4SD).

Service and medication use

Participants in TAU+OT received an average of 15.2 (SD=5.8) OT sessions. In all, 85% of participants in OT (n=66) completed the intervention. Of these participants, only seven (11%) finished OT before the 6-month follow-up. Of the remaining, 45 participants (68%) finished OT before the 12-month follow-up, and 14 participants (18%) finished OT before the 18-month follow-up.

Participants in TAU+OT had significantly less visits to a psychiatrist (M=10.6, SD=6.3) than those in TAU (M=14.5, SD=8.4; p=0.005). When corrected for baseline covariates, this difference remained statistically significant (p=0.01). There were no significant group differences regarding the number of visits to a psychologist (TAU+OT=4.87 (8.94), TAU=4.79 (8.04); p=0.92), general practitioner (TAU+OT=1.40 (2.51), TAU=1.69 (2.82); p=0.19), or occupational physician (TAU+ OT=2.79 (4.13), TAU=3.79 (5.61); p=0.12).

In total, 13 participants (17%) in TAU+OT and eight participants (21%) in TAU were admitted for day treatment or inpatient treatment, with a mean of 38 days (SD=14) in TAU+ OT and 58 days (SD=36) in TAU. Although we were not able to perform statistical tests due to the small number of participants being admitted, these findings seem to suggest that TAU+OT leads to fewer days of day treatment/inpatient treatment. Regarding medication, 78% of those in TAU (n=30) and 73% of those in TAU+OT (n=57) used antidepressants (p=0.55) during the 18-month study period, for a mean of 290 (SD=235) and 303 (SD=237) days, respectively (p=0.77).

Primary outcome: work participation

Both groups significantly decreased in their hours of absenteeism (p<0.001), with the highest decrease between six and 12 months (time2: p<0.001). However, there were no significant differences between groups (table 2). Similarly, two Cox proportional hazards models, adjusted for baseline covariates, indicated no significant group differences in time until partial RTW (HR=0.72; 95% CI 0.44 to 1.11; p=0.14) or time until full RTW (HR=0.93; 95% CI 0.57 to 1.53; p=0.79). The median number of days until partial RTW was 80 (IQR: 42–172 days) in TAU+OT and 166 (67–350) in TAU. For full RTW, the median number of days was 361 (193–653) in TAU+OT and 405 (189–613) in TAU. During the 18-month study period, 91% of participants achieved at least partial RTW (TAU+OT=92%; TAU=89%), and 63% of participants achieved full RTW (TAU+OT=66%; TAU=56%).

Table 2

Mixed-model analyses of work outcomes for participants in TAU+OT (n=78) and TAU (n=39)

Secondary outcomes

Over time, participants in TAU+OT showed greater improvement in depression symptoms than those in TAU, both in terms of severity (p=0.03) and long-term symptom remission (HRSD≤7; OR=1.8, 95% CI 1.0 to 3.3; p=0.05; see also table 3). In addition, the percentage of participants who attained sustainable remission—defined as remission for ≥6 months31—was higher in TAU+OT (92%) when compared with TAU (69%; p=0.04). Although participants in TAU+OT also improved more on the self-report measure, IDS-SR, this difference was not statistically significant (p=0.13; table 3).

Table 3

Mixed-model analyses of clinical outcomes for participants in TAU+OT (n=78) and TAU (n=39)

In addition, participants in TAU+OT showed greater improvement on the SF-36 Mental Health scale over time than participants in TAU (p=0.04; table 3). Although both groups decreased in SF-36 Emotional (p<0.001) and Physical (p<0.001) Role Limitations, with the highest decrease between six and 12 months (Emotional: time2 p<0.001, Physical: time2 p=0.01), no significant differences between groups were found regarding improvements in role functioning (table 3).

Both groups significantly decreased in their work limitations (all three WLQ scales p<0.001), with the highest decrease between six and 12 months (Output: time2 p=0.01, Time Management: time2 p=0.02, Mental/Interpersonal: time2 p=0.02). Similarly, both groups increased in work efficiency (p<0.001), with the highest increase between six and 12 months (time2 p=0.01). However, no significant group differences were found for these measures (table 2).

Intermediate outcomes: self-efficacy and coping

Both groups improved in their active coping (p<0.001) and self-efficacy (p<0.001). In addition, both groups showed a reduction in their passive coping strategies (Passive Reaction: p<0.001, Avoidance: p=0.05), with the highest reduction between six and 12 months (Passive Reaction: time2 p=0.004, Avoidance: time2 p=0.04). No significant group differences were found (table 4).

Table 4

Mixed-model analyses of intermediate outcomes for participants in TAU+OT (n=78) and TAU (n=39)

Additional analyses

Over time, the probability of RTW-GH increased more for participants in TAU+OT when compared with participants in TAU (OR=1.9, 95% CI 1.1 to 3.2, p=0.02; see table 2). Participants who had RTW-GH had significantly less work limitations (all three WLQ scales: p<0.001), higher work productivity (p<0.001), and better role functioning (Role Emotional: p<0.001; Role Physical: p<0.001) than those who achieved full RTW, but without being remitted (all analyses included correction for baseline differences).

Furthermore, participants who RTW-GH had comparable role functioning with the general Dutch population (Emotional: 0.03SD; Physical: 0.12SD), high average work productivity (M=8.01, SD=1.5), and work limitations for less than 30% of the time (WLQ Time: M=24.10, SD=20.5; WLQ Output: M=24.00, SD=18.8; WLQ Mental/Interpersonal: M=26.4, SD=15.3).


In a highly impaired population, we could not demonstrate significant benefit for adjuvant OT with regards to overall work participation. We did find, however, that adjuvant OT increased the probability of long-term RTW-GH (RTW-GH, ie, full RTW while being remitted from depression and with better work and role functioning). Adjuvant OT also increased long-term depression recovery, both in terms of symptom reduction and sustainable remission. Finally, our findings suggested that employees in adjuvant OT used less high-cost medical treatment than those in TAU only (ie, fewer sessions with a psychiatrist and fewer days of day treatment or inpatient treatment).

Our inability to detect significant differences with regards to overall work participation is in contrast with our previous study.12 One potential explanation may be that due to the high variability in the duration until partial/full RTW, our power was insufficient to detect significant differences between treatment groups. Second, as previous studies have indicated that functional improvement lags behind symptom reduction,32 it is possible that our follow-up period of 18 months was too short for measuring whether additional effects in depression recovery lead to further improvement in work outcomes after the 18-month follow-up. Finally, an important factor that may have contributed to the discrepancy in study findings is the rapid societal changes that have occurred in The Netherlands since our previous study. First, several legislative changes (ie, the new Disability Act and an improved version of the Gatekeeper's Act) have been implemented, stipulating more (financial) incentives for both the employer and employee to achieve a fast RTW. Second, guidelines for occupational physicians and mental health professionals have increasingly emphasised the importance of an early RTW before the recovery of symptoms. These societal changes probably have led to a different TAU, resulting in a reduced contrast with TAU+OT in the current study when compared with our previous study, especially with regards to work participation. Indeed, overall, the population in this study seemed to RTW much faster than the population in our previous study (for TAU+OT, mean of 149 days instead of 207 days; for TAU, mean of 213 days instead of 299 days).

Nevertheless, our finding that adjuvant OT provides additional benefit for achieving RTW-GH is advantageous for all stakeholders involved: Without (interfering) residual symptoms, the employee is able to be fully present at work while having improved work functioning. This outcome is not only beneficial for the employee's well-being, but also contributes to his/her economic productivity, which is of interest to the employer. As previous research has suggested, that employees who achieve a healthy RTW have lower chances of recurrent sickness absence,33 future research should investigate whether adjuvant OT also provides benefit for a long-term sustainable RTW.

Furthermore, the added benefit of adjuvant OT for long-term depression recovery is especially important in our highly impaired sample with long depression duration, as previous research suggests that treatment response decreases as a function of symptom duration.34 Our finding that adjuvant OT leads to lower residual symptoms is of clinical relevance since these symptoms have proven an important predictor for relapse35 and subsequent functional impairment.36 Indeed, in this study, we found that employees in adjuvant OT more frequently achieved sustainable remission than those in TAU. It should be noted that this positive effect of adjuvant OT was statistically significant for the clinical interview (HRSD), but not for the self-report measure (IDS-SR). However, previous findings have indicated that clinician-rated instruments may be more sensitive in measuring changes in depression symptoms than self-report instruments.37 Moreover, we found that patients in adjuvant OT showed significantly more improvement on the SF-36 Mental Health scale than patients in TAU. Given the considerable overlap of the SF-36 Mental Health scale with depression symptoms,38 this finding corroborates the beneficial effect of adjuvant OT on symptom recovery.

Although the effect of adjuvant OT did not occur through the working mechanisms as initially hypothesised (ie, improved work-related coping and self-efficacy), the beneficial effect of adjuvant OT may be explained by the provision of extra therapeutic support during the RTW process. This therapeutic support may increase the employee's capacity to positively use his/her RTW experiences as part of the recovery process. This explanation is supported by previous studies showing that on-the-job support is effective for health recovery and work outcomes.39 However, more research is needed in order to confirm the working mechanisms through which adjuvant OT achieves its added benefit on depression recovery and RTW-GH.

Strengths and limitations of the study

This study has the following strengths. First, we included a broad range of outcome measures in order to comprehensively assess the effect of adjuvant OT. Other strengths are the use of a ‘protocolised’ intervention, the inclusion of process variables, the long follow-up period with multiple measures, the low loss to follow-up, and the use of an up-to-date approach for missing data that can account for differential loss to follow-up.

A limitation of our study is the small sample size. Although this sample size was sufficient to meet the requirements of our power analyses, the wide variability in duration until partial/full RTW may have limited our power to detect differences in our primary outcome (work participation). Second, at-work functioning was measured with self-report questionnaires. Hence, we cannot rule out a discrepancy with employer-based ratings. However, previous studies have demonstrated substantial correlations between self-report measures and employer-based data.40 Second, our TAU consisted of highly specialised treatment at an academic department for mood disorders; this could have potentially reduced the contrast between the two groups, which may have lead to an underestimation of the potential effects of adjuvant OT. Third, an advantage of using random coefficient analyses is that these analyses result in more valid estimates by taking the dependency of the data into account. However, one has to pay a price. The calculation of effect estimates becomes very complicated, and there seems to be no consensus yet regarding how to calculate these effect sizes. Finally, the high proportion of employees with chronic sickness absence (69% were absent for more than 3 months) may have diluted a potential effect of adjuvant OT in those with shorter sickness absence spells. Although power was too low to reach significance, exploratory analyses indicated a beneficial trend of adjuvant OT for those with sickness absence spells of less than 3 months (data available from first author). Thus, our findings may only generalise to those with a substantial duration of impairments.


Although adjuvant OT was not significantly more effective than TAU for improving overall work participation, adjuvant OT did increase depression recovery and the probability of RTW-GH in a highly impaired population. Future research should (a) examine whether adjuvant OT provides added benefit for those with a shorter duration of impairments and (b) include a longer follow-up period in order to examine whether adjuvant OT also prevents the recurrence of sickness absence.


We would like to thank occupational therapists Ada Frans, Lea Stor and Sylvia Teunissen for conducting the occupational intervention and for their helpful comments. We would also like to thank Wendy Ooteman for her help with the study design and data collection. Finally, we are grateful to all patients and clinicians who contributed their time and effort to the study.


View Abstract


  • Contributors GV designed the occupational intervention. HH drafted the manuscript and was responsible for study coordination and collection of the data. AS is guarantor of the study. GV, MK and AS have provided critical revisions on the manuscript. All authors participated in the study design, and read and approved the final version of the manuscript. All authors had full access to the data.

  • Funding This study was funded by the Netherlands Foundation for Mental Health (grant no. 20035713) and the National Institute for Employee Benefit Schemes (grant no. 5002002) in Amsterdam, The Netherlands. The authors were independent of the funders, and the funders had no role in the project.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Medical ethics committee of the Academic Medical Center in Amsterdam, The Netherlands (MEC 06/285).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice This article has been corrected since it was first published Online First. The sentence “Indeed, overall, the population in this study seemed to RTW much faster than the population in our previous study (for TAU+OT, 66 days instead of 207 days; for TAU, 80 days instead of 299 days).” has been corrected to read “Indeed, overall, the population in this study seemed to RTW much faster than the population in our previous study (for TAU+OT, mean of 149 days instead of 207 days; for TAU, mean of 213 days instead of 299 days).”

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles