Study design
This is a feasibility randomised controlled trial with a 4-month intervention (called PROWORK) to be delivered to employees and managers in SMEs and LEs. Each of the organisations are the units of randomisation (the clusters), with data collected from individual workers (the participants). In addition, managers with a worker on LTSA will be eligible to participate. Although, dyads of managers and workers can participate, this is not essential. This design overcomes the problem of contamination between the intervention and control arms and the problems associated with individually consenting and randomising workers to a trial where the employers may be managing the sick leave of 2 or more workers. A repeated measures design will be adopted, whereby worker participants will complete outcome measures at baseline, 2 months and 4 months and managers will complete outcome measures at baseline and 4 months. Each organisation’s involvement in the trial is for 12 months. Observations and end-of-study interviews with workers and managers and informal monthly meetings with the organisation’s human resources representative will be conducted throughout the intervention period as part of a full process evaluation. Figure 1 shows the study flow diagram, and Fig 2 indicates the schedule of enrolment, intervention and outcome measures. The project has been granted ethical approval by the Loughborough University Ethical Advisory Committee (reference 2020-1889-2041). The trial was pre-registered in the ISRCTN registry on [15th December 2020] (https://doi.org/10.1186/ISRCTN90032009).
[Insert figure 1 about here]
Setting
PROWORK is part of the Mental Health and Productivity Pilot (MHPP; https://mhpp.me/), a larger research programme which has been funded by the Midlands Engine (CPU 2640) to develop cost effective and sustainable research to support good mental health at work, reduce stigma and increase productivity in businesses within the nine areas of the Midlands Region in the UK. This is the UK’s largest regional economy outside London (£246bn) and it is home to a sixth (11 million) of the UK’s population, concentrating a mixture of organisations ranging from advanced technology, to transport, research and manufacturing. Therefore, this study will take place in a wide range of organisational settings across the Midlands and will involve small (10 to 49 workers), medium (50 to 249 workers) and large (over 250 workers) enterprises.
Sample size
Recommendations for feasibility pilot RCT sample sizes were followed. Guidance for pilot clinical trials [16] recommends recruiting a minimum of 4 clusters per arm. Additionally, feasibility RCTs are recommended to recruit between 24 [17] and 50 [18] participants per arm, consistent with the median sample size found in pilot RCTs [19]. This is in line with the guidance from The National Institute for Health Research (NIHR), which indicates that a sample size of 30 is appropriate to answer the questions posed by a feasibility trial [20]. To match these recommendations, we aim to recruit at least 8 organisations of all sizes (small, medium, large) (4 control and intervention), and 30 participants, including workers and managers, in the intervention group and the 30 in the control group. Participating organisations will be recruited over a 6-month period; the recruitment of workers and their managers will continue for 18 months.
Organisation recruitment and inclusion criteria
Small, medium and large organisations from across all sectors within the Midlands region, UK, will be invited to participate in PROWORK. Organisations will be recruited via webinars, phone calls, newsletters, emails and social media. The study will also be advertised through the Mental Health Productivity Pilot (MHPP) programme’s network, where organisations can fill out an expression of interest form. Organisations who express an interest in the study, will be asked to provide data on the prevalence of their long-term sickness absence over the last 12-months along with a description of their RTW policy and procedure to cross-check for any overlaps or contradictions to PROWORK. Those organisations with conflicting policies that overlap PROWORK will not be eligible for the study. Eligible organisations are small, medium and large organisations with at least 2, 4 and 6 workers, respectively, on long-term sick leave over the past 12 months and with no contradictory or overlapping RTW polices.
Organisations will be asked to promote the study to potential participants (workers on long-term sick leave and their managers) through direct email and newsletters. To ensure the organisations’ compliance and timely recruitment of participants, monthly meetings will be arranged between the research team and the appointed organisation leads.
Participant recruitment, inclusion criteria and consent
Within participating organisations, workers (over 18 years of age) that go on sick leave for at least eight days (or from issue of fit note) and up to eight weeks (50 days) and/or the person managing their sickness absence are eligible to participate. Sickness absence must be associated to either poor mental wellbeing or where poor mental wellbeing may be a comorbidity [14]. Once the worker on long-term sick leave is identified by the organisation, both the worker, and their manager, will be sent a detailed information sheet about the study so that they can make an informed decision about their participation. Those interested in taking part will contact the research team directly either by phone or email, at which stage they will be consented by the research team.
Workers on long-term sick leave will not be eligible to participate in PROWORK if they are on sick leave a) with a psychotic episode such as schizophrenia, or with substance abuse; b) whilst under formal investigation for misconduct or in the formal process of disciplinary action; c) due to cancer and signed off work for at least 6 months; d) due to a neurological condition (e.g., multiple sclerosis, Parkinson, Dementia).
[Insert figure 2 about here]
Allocation to intervention
In this pilot RCT, organisations will be stratified by size. To avoid study contamination, organisations will be randomised into the intervention or active control group after baseline measures are taken using an allocation ratio of 1:1. Randomisation will be carried out by computer-generated randomisation stratified by organisational size (small, medium or large).
Consented workers will be allocated to either the intervention or active control group based on the randomisation of their participating organisation. Baseline data will be collected from each participant prior to either being given the RTW toolkit (intervention group participant) or supporting information (active control group participant). The participating workers and employers will not be blinded in group allocation.
Intervention
This RTW intervention uses a multicomponent approach grounded on behavioural change techniques (BCTs) to promote early and positive communication between the worker and the person managing their sick leave. The overall aim is to reduce the number of days on long term sick and enable a successful RTW. The intervention comprises of 2 toolkits – an employer and worker RTW toolkit. Both toolkits are self-led interventions used by the employer (or manager) and the worker themselves. The guidance and resources in the toolkits for the worker and the employer mirror each other to ensure both receive the same messages and to encourage transparency.
Once participants have been consented into the intervention, they will be able to access the toolkits through a secured website with a code provided by the research team. The toolkits include three step-by-step approaches to be used at different stages of the workers’ RTW process: step 1) managing initial sick leave, step 2) preparing to RTW, and step 3) managing back at work. Additionally, the worker toolkit is supported by a coaching component and the employer toolkit is supported by an upskilling training webinar session. Both toolkits have been co-created with workers with mental health and RTW experience and managers and employers from SMEs and LEs. Both toolkits have also been developed with input from the charity Mind as well as evidence from the scientific literature, and best practice guidelines for the UK (NICE, 2019). PROWORK toolkits are grounded in the Implementation Intentions Theory [21], Conservation of Resources (CoR) Theory [22], and Communication Accommodation Theory [23]. In addition, the worker toolkit is also grounded in the Implementation Intentions [24] Transtheoretical Model of Change [25] and the Socio-Cognitive Theory [26]. The cognitive behavioural elements (i.e., unhelpful thinking worksheet) in the toolkit are also informed by principles of problem-solving and cognitive behavioural approaches [27-29] Table 1 provides a description of each intervention components whilst Figure 3 outlines the Logic Model.
[insert table 1 about here]
[insert figure 3 about here]
Workers
Workers on sick leave will receive the intervention until RTW or 4 months (last follow-up) from when the worker has completed baseline measures. PROWORK includes a step by-step action-oriented toolkit that provides guidance and support from initial absence to post RTW. Three workplace coaching sessions based on goal-setting and problem-solving will be offered to worker by a trained project researcher over the phone. These will be delivered at three time points throughout the intervention: 1) At the start, to support making contact, to coach the participant in the use of the resources, to build a relationship with the participant and to direct the participant to external resources available in the toolkit; 2) at 2 months to coach them for preparation to RTW, or if they have returned to work, then to coach them in adjusting back to work, 3) at 3 months to either coach them for preparation to RTW, or if they have returned to work, then to coach them in adjusting back to work. The coaching sessions aim to motivate and support workers to use the RTW toolkit, to help them with any activities they might feel stuck with and to discuss their action plan to avoid any relapse. During these sessions, and as part of the process evaluation, the workplace coach will ask about what resources is the worker finding more useful and what are the benefits of using this toolkit.
To reinforce the message of having regular communication with the workplace whilst on sick leave, the worker will have the option to contact the workplace coach (i.e., project researcher) when needed either by email or phone call.
Employers
Employers will have access to a RTW training video at the start of the intervention study, prior to recruiting workers on sick leave. The key workplace person (e.g., the line manager) responsible for managing a worker on sick leave will receive an employer version of the toolkit and provided with advice on when and how to use it with a worker. For those individuals who are managing more than one worker on long-term sick leave, they will receive an email from the researcher at the start of each worker on long term sick leave, to remind them to use the toolkit (once they had done the webinar training).
Active control group
To compare the effects of the intervention against usual practice, organisations allocated into the control group will be asked to continue with their usual procedures and no training or guidance will be offered. Participants in the control group will be asked to complete the same measurements as those in the intervention group and at the same timepoints. At the end of the PROWORK study, organisations in the control group will receive a report evaluating their current RTW policies and procedures, access to the line manager training and hard copies of both toolkits.
Outcome measures
Primary and secondary outcome measures will be collected at baseline, 2 months and 4 months. Outcome measures will be collected for both workers and managers. Outcome measures have been clustered as process outcomes and research outcome measures (see below) and they will be collected using Qualtrics; data and personal details will be stored according to General Data Protection Regulation (GDPR, 2016/679) and research governance guidelines. A summary of the outcome measures at the various time points is provided in Table 2. Qualitative interview data will be collected either over the phone or using online conferencing facilities. Both full-return (defined for this study as working the same days or hours per week as before sickness absence in an identical or equivalent role for at least 4 weeks) and partial RTW (defined as working any number of hours in any role) data will be collected via text messages.
[insert table 2 about here]
Trial process-related outcome:
Organisational data collection pre-randomisation
- Summary of long-term sickness absence data for the past 12 months (only total numbers and % by reasons)
- Organisation size and sector
- Copies of sickness absence policy and frameworks
- Copies of the work policy and frameworks
- Details on mental health training and support available for managers
Organisational data collection post-randomisation and the duration of the study
- Number of workers on sick leave (≥ 8 days) and their reasons
- Number of workers that employers have contacted to take part in the study
- Number of workers consenting to take part
- Number of workers using the toolkit (data collected at interviews and website use)
- Number of persons responsible for managing the sick leave and RTW of workers consenting to take part
- Number of persons responsible for managing the sick leave and RTW of workers attending the training
- Number of persons responsible for managing the sick leave and RTW of workers using the employer toolkit (data collected at interviews and website use)
Coverage (recruitment and attrition)
Where possible, data on reach of study participation advertisement, expressions of interests, recruitment, participation and drop-out for all participants.
Toolkit use /adoption
The PROWORK website has an in-built GOOGLE analytics system, that will provide metrics information such as number of visits to the site per day, behaviour of viewers once on the website (i.e., which pages they visited), where most of the traffic to the toolkit website came from, the percentage of returning participants, etc.
Trial feasibility and acceptability-related outcomes:
We will collect data on:
- Willingness of organisations to take part (baseline) and retention through follow-up with reach, uptake and completion as primary endpoints (>30% green for main trial to go ahead, 10-30% amber for main trial with adjustments; <10% red, no main trial).
- Proportion of workers on sick leave referred to pilot study (>50% green for main trial to go ahead, 30-50% amber; <30% red).
- Proportion of workers and line manager participating in both control and intervention groups (Worker: >50% green, 30-50% amber; <30% red).
- For intervention worker participants, short worker interviews will be conducted at each coaching session (at intervention start, 2 months and 3 months) to explore use of the toolkit.
- At 4 months (end of intervention), an interview will be conducted with both worker participants and line manager/RTW contacts to ask them about their engagement, usage and the effectiveness of the toolkit. Information regarding barriers and facilitators to implementation, functionality and interest in on-going usage will also be assessed
- At the end of 12-month study period, organisational stakeholder will be interviewed to explore their perceived benefits of the intervention (engagement, usage, functionality and the effectiveness of the toolkit), as well as barriers/facilitators to implementation, interest in on-going usage, and whether the intervention could form part of any workplace RTW policies.
- Costs associated with toolkit website build and delivery, training delivery, coaching sessions and other associated costs will be collected.
Trial research-related outcomes
Primary worker measures (at baseline, 2 months, and 4 months)
The primary research outcome of the intervention will be assessed monthly after randomisation at the worksite level using 1.1 ratio approach. For the primary research outcome, we will contact (text message) the workers from the control and intervention group on monthly basis and will ask them to report if they are still on sick leave or if they are planning to RTW. For those who have returned to work, we will ask them for the date of their first day back at work (and whether it is a partial or full return). The RTW date will also be collected from the organisational records. Secondary outcome measures of the intervention will be assessed at 3 timepoints (within 2 weeks of consenting, at 2 months and at 4 months).
Number of days of sick leave will be recorded from the employer and from the worker (self-report). Self-report data will be collected at baseline at 2 months and at 4 months (from control and intervention participants) using an online survey. Participants will be asked the following questions: a) Are you still on sick leave? If yes, do you have an RTW date? If not on sick leave, when did you go back to work? b) How many hours are you currently working? c) Is this the same as before your sick leave? Those who have returned to work will be asked to report their first day back at work (and whether it is a partial or full return). The last data collected will be at 4 months post randomisation.
Secondary worker measures at baseline, 2 months, and 4 months
- Self-report mental health: the 9-item Patient Health Questionnaire (PHQ-9) [30], and the 7-item General Anxiety Disorder (GAD-7) [31] will be used to measure depression and anxiety, respectively. The PHQ-9 is used by GPs and practitioners involved in the Improving Access to Psychological Therapies (IAPT) initiative, providing an opportunity to compare the outcomes of this study directly with routine care. Both measures accurately reflect improvement and worsening of symptoms of depression and anxiety.
- Return to work measures: Expectations about length of sick leave will be asked using one question from Aasdahl et al (2018) [32] “For how long do you believe you will be on sick leave from today?” with 6 response options “not at all”, “less than 1 month”, “1–2 months”, “2–4 months”, “4–10 months” and “more than 10 months”. The Lagerveld et al (2010) [33] 11-item Return-to-work Self-Efficacy Scale and the Franche et al (2007) [34] 13-item Readiness to Return-to-work will be used to assess confidence and readiness to RTW. For those who have returned to work the 9-item Readiness to Stay at Work Scale [34] will be used.
- Workplace support and communication: 6-item Workplace Health Communication Scale (Yarker et al, no date) will be used to assess quality of communication between the worker, employer and organisation.
- Work outcomes: For those who have returned to work at 2 months and 4 months, work productivity will be measured using the Work Productivity and Activity Impairment: General Health v2.0 (WPAI:GH) [35]. The WPAI:GH yields 4-type of scores: ‘Absenteeism’, ‘Presenteeism’, ‘Work productivity loss’ and ‘Activity Impairment’. A 1-item job satisfaction scale will be used to assess satisfaction [36].
- Intention to use toolkit: 4-item Toolkit Use (Yarker et al., no date), will be used to assess motivation and engagement for those in the intervention group. These questions will be asked at each coaching session.
- Quality of life: Health-related quality of life will be assessed using the EQ5D-5L [37].
- Demographics and other measures: basic demographic information for each participant including their date of birth, ethnicity, and highest level of education will be collected. The average wage for each worker will be identified using UK Standard Occupational Classification coding and annual earnings data for each job type. Workers will also be asked if they are the main wage earner. Information on medical diagnosis of health conditions, prescribed medication use and other current therapeutic treatments for mental health will be collected (adapted from Peveler et al, 2005; [38]).
- Participants will be asked what actions their workplace contact (i.e., person responsible for managing their return) carried out to support their RTW. The line manager behaviour questionnaire [39,40] will be adapted for this purpose.
Secondary employer measures at baseline
- Mental Health and RTW experience: Those with responsibility for RTW will be asked about their experience with mental health “how much experience you have with mental health either yourself or through a close friend or family member?” with five response options “None at all”, “A little”, “Some”, “Quite a lot”, “A great deal” (1-item, Yarker et al, no date) and 2-item training question “indicate how much formal training you have had in a) managing mental health of others and b) training in sickness absence management and RTW” with five response options “None at all”, “A little”, “Some”, “Quite a lot”, “A great deal” (1-item, Yarker et al, no date) and a 1-item question of long-term sickness absence and RTW management experience “We would like to know how much experience you have with managing long-term sick leave and RTW in the past 12 months” with five response options “None at all”, “1 worker only”, “2-3 workers”, “4-5 workers” and “More than 5 workers”
- Demographics: Data on age, gender, ethnicity, job role and tenure will be collected.
Secondary employer measures at 4 months
Participants will also be asked what actions they carried out to support the return of their worker. The Measure for Supervisors to Support Return to Work (SSRW) [39,40] will be adapted for this purpose.
Process evaluation analysis
A detailed process evaluation informed by the Implementation Outcome Framework (IOF) [41,42] will be conducted. This framework includes eight implementation outcomes (acceptability, adoption, appropriateness, feasibility, fidelity, implementation cost, coverage and sustainability). In addition, the Theoretical Domains Framework (TDF) [43,44], a widely used framework in behaviour change and implementation research, will also inform the process outcome data collection. Use of the TDF allows an in-depth exploration of the barriers and facilitators of implementing the trial. The data collection methods and its process outcomes outlined below capture the process outcome information for the IOF and TDF.
Overall, end of study semi-structured interview schedules will be developed to explore participants experiences using PROWORK during their sickness absence and RTW. Similar interview schedules will be created to explore the main barriers and facilitators to engaging and implementing PROWORK at the organisational level. These will be supplemented with researcher notes, including thoughts and observations about the organisation’s procedures and implementation approach and notes from the monthly calls between the research team and the organisation’s HR contact. Collectively, these will provide information on the acceptability of the trial procedures including randomisation, the measurement instruments, and the overall acceptability of the intervention. Interviews and monthly calls will be conducted by a conferencing platform, so they can be recorded and transcribed verbatim. Short employer interviews (e.g., with HR, Health and Safety Manager) with intervention and control sites at monthly intervals will be conducted to explore any changes to policies or processes that may impact the study. Questions around study participation (e.g., identifying workers on sick leave, sending out study information, etc) will also be explored and formal notes will be taken as part of the process evaluation. A detailed plan of the measures and methodologies used in the process evaluation is described below.
Process outcomes:
- The number of organisations agreeing to participate in the trial will be summarised in terms of their size, sector, sick leave and RTW polices and number of workers who were on long term sick leave in the past 12 months prior to the start of the study.
- The number of worker participants identified on long-term sick leave and the number recruited into the study will be reported, along with the number of participants followed up at each time-point. Withdrawals (and where possible, reasons for withdrawals) will be reported.
- Difference in recruitment uptake rate and follow-up rates at each time point will be compared between the intervention and control arms.
- As organisations of different sizes are taking part, it is likely there will be some imbalance between participants in each treatment arms on one or more baseline characteristics. Baseline comparisons will be carried out to detect any substantial differences between participants recruited from the control and intervention arms. This will be done by scrutinising the baseline data for any serious imbalances in observable baseline variables and the trends of the imbalance if any. The recruitment rates will also be estimated and compared between the control and intervention arms. The size of any imbalances will be examined, in addition to evidence of systematic selection bias in the types of patients being recruited in control versus intervention arms
- Key baseline characteristic will be compared between those participants followed up and those lost to follow-up at each timepoint.
- Intervention fidelity will be assessed by the log-in and downloads of the toolkits. Successful adherence is defined as at least 60% download of the total toolkit by workers and employers (i.e., those with a responsibility for managing RTW), and at least 60% completion of the activities/checklist in total.
- Following the guidance from the COREQ-32 checklist [45], qualitative data for the process evaluation will be recorded, transcribed verbatim and coded following the principles of thematic analysis [46], the Theoretical Domains Framework [43], and the Normalisation Process Theory [47]. These will be supplemented with observations made by the researchers throughout the implementation of the intervention. Collectively, these will provide information on the acceptability of the trial measurements and the intervention.
- Survey data for the process evaluation will be summarised using means, standard deviations, medians and ranges for continuous variables and counts and percentages for categorical variables.
- The pilot data will provide information on the parameters needed for a realistic sample size calculation (mean, standard deviation and treatment effects of the primary outcome for the 2 arms) for a future, main cluster RCT.
Participant appreciation
As a thank you for participating in the pilot trial, participants will have the chance to win a £50 gift voucher for every survey they complete during the baseline and follow-up measures to encourage participation.
Economic analysis
The economic analysis will be exploratory, with the aim to inform the design of a full cost utility analysis alongside a future main trial. Data on costs will be sought from all participants and results will be presented taking into account worker-incurred costs and productivity losses. Analyses will be mainly descriptive, and all costs and outcomes will be summarised using means and 95 % confidence intervals.
- Healthcare resource used will be collected using self-completed questionnaires at baseline, 2 and 4 months, with a recall period of 2 months in each. Questions will ask workers to recall GP consultations, visits to healthcare professionals, outpatient appointments, investigations or treatments and inpatient stays related to the index condition (adapted from Peveler et al, 2005 [38]). Participants will be asked to distinguish between National Health Service (NHS) and private practice visits.
- Resource used for the intervention will be directly recorded and costs attached, staff time (e.g., coaching sessions, training sessions), materials (posters, flyers, referral forms, website set-up and maintenance, toolkit printing) and training sessions.
- Costs will take into account both absenteeism and presenteeism and will utilise self-report data on employment status, occupation and time off work and reduced productivity at work (presenteeism).
- All workers will be asked to complete the 5-level version of the EuroQoL-5DL (EQ-5DL) (1990) [48] (questionnaire at baseline, 2 months and 4 months) in order for the quality-adjusted life years (QALYs) over the 6-month time period to be calculated for each participant. The QALYs combine information on health-related quality of life and survival.
- Productivity costs will be calculated using data collected on absence from the number of days taken to RTW.
- Using the human-capital approach (which assumes that the value of lost work is equal to the amount of resources an individual would have been paid to do that work) the self-reported days of absence will be multiplied by the respondent-specific wage rate.
Statistical data analysis
This study will be analysed according to the Consolidation Standards of Reporting Trials (CONSORT) statement for cluster RCTs [49]. As this is a feasibility study, main analysis will include descriptive statistics (mean, SDs and medians). Data will be analysed using IBM SPSS statistics 25. Data analysis will be performed after the last trial participant has completed their 4-month post-intervention. A baseline table (descriptive statistics and frequencies) will compare the demographic and clinical characteristics including gender, age, education, number of days on sick leave, mental health status, readiness, intention and self-efficacy to RTW, work support, communication, performance at work, sleep, and physical activity, between the 2 arms. As this is a pilot trial, no emphasis will be put on the p values for any inferential statistical tests conducted. Statistical analysis will be carried out on an intention to treat basis with missing outcome data being imputed using multiple imputation. To explore the extent and patterns of missing outcome data, we will report the proportion of missing values per item, proportion of participants who complete all items on the questionnaire and the proportion of respondents who answer at least 50 % of the items in a scale. The proportion of missing data will also be reported for the other key outcomes and compared between the participants from intervention and control practices. The point estimate of the proportion and its 95% confidence interval (CI) will be provided for the primary research outcome i.e., number of days taken to RTW (partial or full) for the intervention and the control arms. Key baseline characteristic will be compared between those participants followed up and those lost to follow-up at each timepoint.
Data management and research governance
All data will be anonymised and entered into a secure database and only accessible by trial staff and authorised personnel. The study will comply with the Data Protection Act which requires data to be anonymised as soon as it is practical to do so. Personal data will be processed on the public task basis under the General Data Protection Regulation (GDPR). The study will be sponsored by Loughborough University. An independent Trial Steering Committee (TSC) will be created to oversee the evolution of the study and they will meet every 4 months. Due to the low-risk nature of this study no adverse events are anticipated to occur; however, should any arise, the TSC and the Loughborough University guidelines will be consulted for the managing and reporting of any intervention-related adverse events.