Integrated mental health care and vocational rehabilitation to improve return to work rates for people on sick leave because of depression and anxiety (the Danish IBBIS trial): study protocol for a randomized controlled trial

The aim of this randomized, three-arm, investigator-initiated, multicenter, parallel-group superiority trial is to compare the effect on return to work of the following interventions: (1) IBBIS mental health care integrated with IBBIS vocational rehabilitation, (2) IBBIS mental health care and standard vocational rehabilitation, and (3) standard mental health care and standard vocational rehabilitation. The primary hypothesis is that participants allocated to the IBBIS mental health care integrated with IBBIS vocational rehabilitation will have significantly faster return to work rates than people who are allocated to standard mental health care together with standard vocational rehabilitation. The secondary hypothesis is that IBBIS mental health care together with standard vocational rehabilitation will have a lesser but significant effect on return to work compared with standard mental health care together with standard vocational rehabilitation. The superiority of the IBBIS mental health care integrated with IBBIS vocational rehabilitation will also be tested by comparison with the IBBIS mental health care together with standard vocational rehabilitation intervention. The IBBIS vocational rehabilitation intervention alone will not be tested in this trial. We hypothesize that the superiority of the IBBIS interventions will be applicable for secondary outcomes and exploratory measures at 6-, 12-, and 24-months follow-up so that (1) symptom level and presenteeism will be lower in participants allocated to IBBIS interventions; (2) improvement in self-efficacy, quality of life, and functioning will be better among participants allocated to IBBIS interventions; and (3) satisfaction with services and number of weeks worked will be higher for participants allocated to IBBIS interventions.

The IBBIS trial is designed and reported in this article according to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2013 statement (see Additional file 1) [40], and the final results will be published according to the Consolidated Standards of Reporting Trials (CONSORT) criteria for randomized trials of nonpharmacological treatment [41]. There is a parallel trial in the IBBIS project with an identical research design for participants with adjustment disorder, exhaustion disorder, or psychological stress.

The interventions will be delivered by a cross-sector and multidisciplinary IBBIS team that is organized in collaboration between Mental Health Services in the Capital Region of Denmark and the following four municipalities: the City of Copenhagen and the three suburban municipalities Gentofte, Gladsaxe, and Lyngby-Taarbæk. Participants are referred to the study by social security officers from job centers in the four municipalities, and the interventions are provided in locations other than the job centers and the mental health centers.

Eligible participants in this trial meet the following inclusion criteria:

Eligibility and diagnosis are assessed by an IBBIS team member (nurse, physiotherapist, social worker, occupational therapist, psychologist, or psychiatrist) who is trained to perform the assessment.

The IBBIS interventions are not designed to accommodate people in need of acute or highly specialized care. Thus, potential participants will not be eligible if they meet the exclusion criteria:

We wish to compare the IBBIS mental health care intervention alone with standard treatment. Thus, a seventh exclusion criterion is that a potential participant will not be included in the study if he or she does not accept the need to abstain from taking part in any psychotherapy or psychotherapy-like treatment outside the IBBIS intervention if he or she is randomized to IBBIS mental health care. Exclusion criteria 2–5 are applied only if the medical doctor on the IBBIS team confirms the assessment and anticipates that the patient cannot benefit from the IBBIS interventions. In cases with high suicide risk, the participant will be referred to acute care services. People with alcohol and substance abuse problems will be offered referral to treatment if relevant.

Case managers from the four job centers can refer adult Danish-speaking citizens on sick leave from either work or unemployment to a psychiatric assessment if either the case manager, the citizen, or the individual’s general practitioner (GP) suspects a mental health condition to have caused the sick leave. The referral and assessment are voluntary. The results of the psychiatric assessment will be used in the treatment plan if the participant is allocated to treatment in IBBIS. The results of the assessment will be shared with the individual’s GP and the job center. The psychiatric assessment is based on three sources of information about the participant:

Assessors are IBBIS team members who are specially trained to use the above-mentioned instruments. Trial eligibility will be evaluated after the psychiatric assessment, and subsequently, assessment data will constitute baseline data at the time −1 (see Table 1 and Additional file 1: Extended SPIRIT figure). The assessment process should not take more than 2 weeks but can be prolonged if one or more of the three types of information are missing.

Individuals who meet the aforementioned trial criteria at assessment and subsequently consent to participate will be randomly allocated to an intervention by the assessor at t₀. Participants will be followed at 6, 12, and 24 months after allocation (see Fig. 1). Participants will be prompted to fulfill self-assessment questionnaires at each follow-up through up to five personal contacts.

All electronic data (self-assessment, interview, and register data) are stored on secured servers at closed networks, and access to data is logged through a unique login for an assigned list of IBBIS personnel. Physical data material (case report forms with selected interview data) is stored in locked spaces in locked facilities. Transfer of electronic data between staff members and other approved data-managing institutions is carried out using only tunnel encrypted email or encrypted USB sticks.

The allocation ratio between the three arms is 1:1:1. Centralized randomization will take place according to a web-based, computer-generated allocation sequence with varying block sizes kept unknown to the assessors. The Odense Patient Data Explorative Network (OPEN) is responsible for the randomization; administrative personnel in the IBBIS team perform the online randomization; and the IBBIS team leader will assign the participant to interventions and professionals.

We expect that service delivery can vary from municipality to municipality and that the process of gaining a new job from unemployment will take a longer time than returning to an existing job. Previous research has shown that diagnosis is a possible predictor of return to work [47]. Thus, the randomization is stratified according to (1) municipality, (2) employment status (on sick leave from work vs. on sick leave from unemployment), and (3) diagnosis (depression vs. anxiety).

Owing to the psychosocial nature of the IBBIS interventions, the participants and the professionals delivering the IBBIS interventions cannot be blinded to the group allocation. All outcomes are based on registries or self-assessed questionnaire data, and no assessor-based follow-up data will be obtained. Register data on employment status and income is created automatically through the national registries. Information on the participant’s benefit status is created through the job center management system, and benefits are granted and registered by the employment consultants in the IBBIS team.

Blinding of assessors is relevant only at the baseline interview, which takes place before group allocation. Referring personnel will likewise be blinded to the allocation sequence and block size to prevent them from anticipating the next group allocation. The researchers will be blinded to group allocation during the process of data analysis. Group allocation will be coded with names like X, Y, and Z to conceal the given intervention. The researcher will draw up conclusions at 6- and 12-month follow-up on the basis of six scenarios where each group (X, Y, or Z) has received integrated IBBIS mental health care and IBBIS vocational rehabilitation, IBBIS mental health care and standard vocational rehabilitation  and standard mental health care and vocational rehabilitation. After this, the blinding will be broken. The researcher performing analysis at 24-month follow-up will not be blind to group allocation, because any possible differences between groups will be revealed after 12-month follow-up.

The IBBIS intervention team is constituted by (1) three and a half full-time care managers who are nurses, occupational therapists, physiotherapists, or social workers with mental health care experience and a minimum of 1 year of certified training in cognitive behavioral therapy (CBT), (2) three and a half full-time employment consultants who are social insurance officers from the job centers, and (3) equivalent to e.g. 0,75 psychiatrist? (alternatively, GPs or psychologists). Care managers, psychiatrists, and psychologists are employed in the mental health services, and employment consultants are employed in the job centers of the four municipalities. Care managers have a maximum momentary caseload of 25, and employment consultants have a maximum momentary caseload of 20. The IBBIS mental health care intervention is expected to have an average duration of 4 months, and the duration of the IBBIS vocational rehabilitation is expected to average 7 months.

The team delivers two separate interventions: (1) IBBIS mental health care alone and (2) IBBIS mental health care integrated with IBBIS vocational rehabilitation. Both interventions are carried out with emphasis on participant involvement through shared decision-making to improve participant satisfaction [48] and involvement of the participants’ relatives. A fidelity scale (Rie Poulsen, unpublished data) is developed and used for biannual fidelity reviews to ensure program adherence and continuous focus on program implementation and improvement. Once program fidelity is achieved, future fidelity reviews will be conducted annually.

IBBIS mental health care is delivered as manualized stepped care according to national and international guidelines for treatment of depression and anxiety [19, 20, 22, 23]. The participant will be offered treatment options according to a stepped care plan, with the least invasive and least resource-demanding effective treatment offered first. One or more of the following treatment options will be provided by the care manager:

Intervention modalities will be offered according to the stepped care plan (see Tables 2 and 3).

The psychiatrist of the IBBIS team is responsible for the following:

Participants will receive standard vocational rehabilitation services from the local job center along with continuous control of the grounds for receiving sick leave benefit. The IBBIS team will not collaborate with job center personnel.

The mental health care in the integrated intervention is identical to that described in the IBBIS mental health care above. The concurrent vocational rehabilitation in IBBIS is composed of the following elements, which are delivered to meet the participant’s individual needs for vocational recovery:

Consistency between goals in treatment and vocational rehabilitation is crucial [34, 37]. Several integrational elements ensure coherence in the participant’s process of returning to work and recovering from mental problems in the integrated IBBIS intervention:

The integrated services are based on the theoretical framework of relational coordination by Jody Hoffer Gittell in which timely problem-solving communication between different professionals is created by focusing on shared goals, shared knowledge, and mutual respect [54]. Unfortunately, it has not been possible to establish a shared electronic folder for IBBIS staff from different organizations to use, owing to separate, secure information technology systems, and hence written communication across sectors and municipalities can be shared only through emails.

Employment consultants and care managers all attended a 4-week training course in April 2016, comprising 1 week of joint training and 3 weeks of training in their monodisciplinary groups. Care managers are trained in all aspects of the IBBIS mental health care intervention, with a special focus on psychiatric assessments and CBT. Likewise, employment consultants are trained in all aspects of the IBBIS vocational rehabilitation intervention, with a special focus on the problem-solving method and job development. Care managers have weekly case-based supervision and CBT supervision every 2 weeks; the employment consultants have weekly supervision; and the team has monthly case-based cross-disciplinary supervision.

Participants who will be allocated to the control group of the trial will receive standard health care by their GP and standard services in the job center. GPs can, with supervision, offer up to seven therapy sessions to patients or refer eligible patients to a private psychologist. People with depression (all age groups) and anxiety (ages 18–38) can receive payment for partial reimbursement of up to 12 sessions of individual psychological therapy. People who qualify for secondary mental health services at baseline are excluded from the IBBIS trial, but some participants will experience deteriorating symptoms during the follow-up period and eventually meet the criteria for treatment by the secondary mental health services. Mental health services offer outpatient treatment in structured treatment packages for people with depression, generalized anxiety disorder, panic disorder, and social phobia. These packages include specialized medical treatment and 6–18 sessions of therapy according to diagnosis. Therapy is most often provided in groups supplemented by individual sessions.

The job centers offer a variety of courses and support and manage the sick leave benefit case according to government legislation, which requires regular follow-up every 4 weeks, reassessment of the sick leave diagnosis after 22 weeks, self-management courses, and support to gradual return to work (in paid or unpaid jobs). Collaboration between the job center and the health care system is normally conducted through standardized sick leave certificates from the GP to the job center. Representatives from the health care system other than the individuals’ own health care providers can be used in reassessment of the individuals’ sick leave case. See Additional file 2 for an overview of the intervention-components in the three arms.

The primary outcome is time from baseline to the event return to work within 12 months after baseline. Work is defined as having 4 consecutive weeks of working with a salary and with no concurrent vocational benefits. Benefit and income status is retrieved from the Danish DREAM database and the electronic income register [55]. The DREAM database is administered by Danish Agency for Labour Market and Recruitment and can be linked to a range of different registers, including the Danish income register. Returning to or achieving a flex job, a type of subsidized work, is also defined as returning to work for participants who are entitled to flex job when they enter the trial. The work-related, symptom-based, and functional secondary outcomes are presented in Table 4. All exploratory and safety measures are presented in Table 5.

The Beck Depression Inventory (BDI-II) consists of 21 items to assess the intensity of depression in clinical and normal patients. Each item comprises a list of four statements (scored 0–3) arranged in increasing severity about a particular symptom of depression [56]. The Beck Anxiety Inventory (BAI) is a 21-item general questionnaire for anxiety, measuring symptoms during the last week as rated on a 4-point Likert scale from 0 to 3 [57]. The BDI-II and BAI have been shown to have excellent psychometric properties, with internal consistency around 0.9 [58, 59]. The Perceived Stress Scale (PSS) of Cohen and colleagues is a global measure of perceived stress. The scale was originally a 14-item questionnaire, and it was later modified to a 10-item questionnaire that has improved and satisfactory psychometric properties [60]. The Work and Social Adjustment Scale (WSAS) is a simple, reliable, five-item scale that measures functional impairment related to an identified problem [61], which is defined in this trial as psychological symptoms.

The 4DSQ is a 50-item questionnaire designed to assess common psychological symptoms in the last week and has a special focus on distinguishing general distress from depression, anxiety, and somatization [46]. The Karolinska Exhaustion Scale 26-item version measures the degree of exhaustion disorder and the 4 interrelated dimensions of exhaustion disorder according to the Swedish National Board of Health and Welfare: lack of recovery, cognitive exhaustion, somatic symptoms, and emotional distress [62, 63]. The EQ-5D-5L is a measure of health status in five domains: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression and also includes a visual analogue scale from 0 (worst imaginable health status) to 100 (best imaginable health status) [64]. The Flanagan QOLS is a 16-item instrument that measures 5 conceptual domains of quality of life: material and physical well-being; relationships with other people; social, community, and civic activities; personal development and fulfillment; recreation; and independence [65]. The six-item personal control subscale from the revised version of the Illness Perception Questionnaire (IPQ-R) is used to evaluate the participant’s self-efficacy regarding symptom management [66]. Return to work self-efficacy (RTW-SE) is an 11-item measure for self-efficacy beliefs regarding return to work where respondents are asked to respond to statements about their jobs, imagining that they would start working tomorrow in their present emotional state [67]. The General Self-Efficacy Scale is a ten-item psychometric scale that is designed to assess optimistic self-beliefs to cope with a variety of difficult demands in life [68]. The Client Satisfaction Questionnaire (CSQ-8) is an eight-item questionnaire that is used to measure participants’ satisfaction with mental health care services and vocational rehabilitation [69]. Presenteeism refers to the state where a person attends work while sick [70] and is used as a proxy measure for returning to work while having reduced workability.

The sample size of this trial is based on the primary outcome return to work rate (HR). There are, to our knowledge, no comparable Danish studies, and the sample size estimates are based on Dutch studies of comparable interventions for populations on sick leave with common mental disorders. The desired type II error risk is set at 10% (power of 90%). The mean number of days from baseline to return to work in the control group is conservatively estimated to be 210 days [71‐73]. Owing to multiple testing because we will compare the three study arms, we use the Bonferroni correction for the type I error risk (α) and set it to 0.0167. An HR of 1.5 is estimated to be clinically relevant [30, 74, 75], and participants will be recruited through 639 days and followed for 365 additional days. With an allocation ratio of 1:1:1, we need 201 participants in each of the 3 arms to reject the null hypothesis that the return to work rate is equal in the control group, the IBBIS mental health care intervention, and the integrated IBBIS mental health care and IBBIS vocational rehabilitation intervention. If we fail to include 603 participants, the statistical power can be lowered to 80%, and thus only 468 participants would be needed. Power calculations (Tables 6 and 7) indicate that a sample size of 201 participants per group will be adequate to detect relevant significant differences in the secondary outcome measures with a minimum power of 80%.

All sample size and power calculations will be conducted using PS Power and Sample Size Calculations software [76].

The primary objective of this trial is to test the null hypothesis that there is no difference in time from baseline to the event return to work between the three groups at 12-month follow-up. The alternative hypothesis is that one intervention is superior to another intervention. Because the primary outcome data are collected as register data, the data are expected to be complete. Kaplan-Meier survival curves will be computed, and the differences between the three intervention groups will be tested with Cox proportional hazards regression analysis to estimate the treatment effect as HR with 95% CI. Cox regression analysis will also be used for the secondary outcomes at 24-month follow-up: “time from return to work to recurrent sick leave” for the subpopulation of individuals who have started working and “time from baseline to return to work.”

The four continuous outcomes BDI, BAI, PSS, and WSAS are used with a repeated measures design, and the difference in the individuals’ scores between measurements will be analyzed by using linear mixed models with repeated measures and an unstructured covariance matrix if possible. All participants will be included in the analysis according to the intention-to-treat principle. Missing data from the questionnaire-based instruments will be imputed with multiple imputations if we can assume that data are missing at random or missing completely at random. The effect of missing data from all questionnaire-based outcomes will further be assessed by sensitivity analyses. All statistical tests are two-sided. All exploratory continuous outcomes will be analyzed by the same method.

A nonparametric model will be used in situations where the scores are not normally distributed. The binary outcome proportion in ordinary work will be analyzed with logistic regression. All models will be adjusted for the stratification variables. We will assess the potential interaction between time and intervention for continuous secondary outcomes. All statistical tests are two-sided.

The IBBIS depression and anxiety trial was initiated in April 2016, and as of 7 November 2017, 395 participants had been recruited. This protocol is version 2.0. Trial recruitment is expected to end on 31 December 2017.