Background: Inflammatory arthritis (IA) causes work disability, absenteeism (sick leave) and presenteeism (reduced productivity) at high cost to individuals, employers and society. A trial of job retention vocational rehabilitation (VR) amongst people with IA in the US showed that VR reduced work disability. However, it is unknown whether this approach transfers to the United Kingdom (UK) with a different social and welfare structure. Previously, we modified the VR for the UK, funded by Arthritis Research UK (the WORKWELL programme) and demonstrated it to be deliverable and acceptable in a feasibility trial. Our aim now is to move to the definitive UK trial testing the effectiveness and cost-effectiveness of WORKWELL.

Methods: A multicentre randomized controlled trial will be conducted. Employed people with rheumatoid, psoriatic or inflammatory arthritis (n=240), with concerns about continuing working due to arthritis, will be randomized to receive WORKWELL or control (written advice). WORKWELL includes individualised VR (maximum 4.5 hours over several months): assessing work problems; encouraging arthritis self-management in the workplace; addressing ergonomics; considering fatigue and stress management; providing orthoses and educating on employment rights and support services, assistive technology and work modifications. It also includes psychological and disclosure support, workplace visits and employer liaison (as applicable). Outcomes will be assessed at 0, 6 and 12 months by questionnaire. The primary outcome is the Work Limitations Questionnaire-25 (measuring presenteeism: summed score) at 12 and 36 months, with cost-effectiveness analysis at 12 months.

Discussion: If effective and cost-effective, WORKWELL can be rolled out in Rheumatology services to help improve the quality and duration of people with arthritis' working lives.

Background Work problems are common in people with inflammatory arthritis (IA). In rheumatoid arthritis (RA) 33% stop working within two years and 50% within 10 years. Work difficulties include: pain (particularly hand pain); fatigue; unadapted work environments, equipment and work schedules; poor work self-efficacy (i.e. lack of confidence about working); job strain; limited use of self-management strategies; and limited support from employers and work colleagues. Two USA trials of brief (3 hours) job retention vocational rehabilitation (VR) identified job loss was postponed by modifying such factors. A UK proof-of-principle trial (n=32) identified 10 hours (minimum) of VR reduced work instability. VR is not routinely available in the NHS and Rheumatology departments. It is unclear from the evidence if brief VR provided by therapists is effective and cost-effective in the UK.

Aims

1. Assess if there is a difference in work presenteeism between people with RA, undifferentiated IA (UIA) or psoriatic arthritis (PsA) who receive either: usual care, written work self-help information plus brief VR (WORKWELL) provided by a therapist trained in providing VR; or usual care and written work self-help information only (WP1).
2. To assess if there are differences in self-reported work instability, work, activity limitations, work productivity, absenteeism, work status, work self-efficacy, health status, NHS and societal costs between people receiving WORKWELL or written work self-help information only (WP1)
3. To assess the cost-effectiveness of WORKWELL from the perspective of: the NHS using health-related quality of life as the primary outcome; the employer using presenteeism as the primary outcome; and the employer using health-related quality of life as the primary outcome and including presenteeism as a cost. (WP5)
4. To update and evaluate a VR programme for occupational therapists and physiotherapists to help them keep people with RA, IA or PsA in work (WP2).
5. To measure fidelity to the WORKWELL intervention (WP3).
6. To understand the social and structural context in which the intervention is delivered and to identify factors which may influence the quality of implementation (WP3).
7. Investigate contextual factors influencing participants' presenteeism (WP4).

Design A pragmatic, multi-centre individually randomised controlled trial (RCT) comparing the effects of brief VR (intervention) with written self-help advice only (control). Cost-effectiveness and process evaluations will also be conducted.

Methodology Employed people with RA/ UIA/ PsA who have concerns about being able to continue to work in future due to arthritis, will be consented into the study. They will complete a baseline questionnaire (hard copy or online according to participant preference) and will then be randomly allocated in equal numbers, stratified by participants' job skill levels, to either intervention or control groups. Participants (in both groups) will be mailed a self-help work information pack. Participants' in the intervention group will be referred to receive the WORKWELL intervention from a trained occupational therapist/ physiotherapist. The Lancashire Clinical Trials Unit (CTU: University of Central Lancashire) will contact participants monthly by text/ e-mail/ telephone to identify numbers of days on sick leave. Six months after completing the baseline questionnaire, participants will complete a short postal/online questionnaire including presenteeism, productivity health status measures and resource data. At 12 months, participants will complete a third postal/online questionnaire with most of the measures collected at baseline. Two sub-samples from the intervention and control groups will be interviewed: employed participants about their views of VR; and any unemployed/ retired due to ill-health participants about factors contributing to job loss. At 36 months, participants will complete a fourth postal/online questionnaire with some of the key work status measures collected at baseline. Two sub-samples from the intervention and control groups will be interviewed: employed participants about their views of VR; and any unemployed/ retired due to ill-health participants about factors contributing to job loss.

Planned Sample Size 240 people will be recruited from 18 sites. On completion of the trial, qualitative interviews will be completed with 15 employed participants and up to 7 (if any) unemployed / early retired through ill-health from the intervention group. WORKWELL participants' line managers/employers will also be interviewed (where possible: n=10). WORKWELL Therapists and their line managers will also be interviewed (one each from each site, or less if data saturation is reached (n=18 for each).

Analyses Statistical analyses will investigate differences in outcomes at 12 and 36 months after baseline between intervention and control groups using mixed effects linear, logistic or ordinal logistic regression modelling, adjusted for baseline values and stratification variables (as applicable). The primary outcome is the summed score of the Combined Work Activities Limitations Scale -Work Limitations Questionnaire-25 (a measure of work presenteeism) at 12 and 36 months after baseline assessment. The health economic analysis will be conducted from the UK National Health Service (NHS) and employer perspectives. Interviews will be thematically analysed. Records will be analysed using content analysis.

Inclusion Criteria:

• Adults (i.e. aged ≥ 18 years)
• diagnosed with RA, UIA or PsA by a Rheumatology Consultant. (Undifferentiated Inflammatory Arthritis is defined as: persistent synovitis in more than one small joint of the hand, without any other known cause, but the person does not yet meet all the diagnostic criteria for RA. Participants can have co-morbidities (which may also be related to having RA, UIA or PsA, for example, osteoarthritis, fibromyalgia, heart condition, mild to moderate anxiety or depression; or are unrelated, e.g. diabetes.
• In paid work (full or part-time, self-employed or contractual work) for at least 15 hours per week.
• Participants may be on sick leave at the time of screening BUT this must be for less than 4 weeks duration and not planned or likely to extend for longer than 4 weeks.
• Able to read and understand English.
• Score ≥10 on the RA-Work Instability Scale (RA-WIS), a measure of mismatch between the person's abilities and their job demands.
• Able to attend the participating site for WORKWELL appointments, if allocated to that group
• Able to provide informed consent.

Exclusion Criteria:

• on extended sick leave (i.e. > 4 weeks). The WORKWELL intervention is designed for people currently in work. Long-term sick leave is defined as being > 4 weeks [NICE, 2009b].
• Already planning to retire due to age or to take early retirement (for any health or non-health reason) within the next 12 months (i.e. within the trial follow-up period)
• planning to move out of area within the next 3 months and therefore would be unable to attend for WORKWELL sessions if allocated to that group (as treatment may not be completed in time).
• already receiving or awaiting VR services from Access to Work or a Vocational Rehabilitation company. These services conduct work assessments as part of VR provision. (Those receiving work services from other sources (e.g. occupational health or human resources in their organisation, online advice from Fit for Work) may still be recruited as VR provision can be of varied quality and this will be considered as "usual care."
• Employed in the armed forces (as they could be stationed overseas during the trial period. The armed services also have their own Vocational Rehabilitation service.

UK Work Environment Survey (work assessment: already open access). Workwell Solutions Manual to be shared at end of trial (open access).

At end of trial.

Application to study team

The Investigators will remain blinded to group allocation until the participant has completed and returned their baseline questionnaire, and the data verified at the Lancashire CTU.

Data co-ordination staff at the CTU, who will be responsible for contacting patients to obtain missing data from 6- and 12-month questionnaires will be blinded to group allocation. The Data Manager and Information Systems team will be unblinded.

Due to the nature of the intervention it will not be possible for therapists to be blinded to group allocation. Participants cannot be blinded to the intervention they receive.

The statistician will be blinded to group allocation by using a dummy variable for group allocation until analysis is complete. The health economist will not be blinded as the costs of WORKWELL will need to be included into the analysis.