Evaluation of Mental Health First Aid From the Perspective of Workplace End Users

Help-Seeking Behavior

Mental health first aid, Help-seeking behaviours, Workplace, Process evaluation, Economic evaluation

This trial will evaluate the effectiveness of Mental Health First Aid in the workplace. Half of the organisations will receive Mental Health First Aid training and half will receive treatment as usual (a brief consultation from Mental health First Aid England on mental health and well-being in the workplace).

The EMPOWER study will implement a cluster randomised controlled trial (cRCT) to examine the effectiveness and cost-effectiveness of MHFA in the workplace on help-seeking behaviours of employees, with embedded qualitative, process and social impact evaluations . In order to best answer questions regarding effectiveness, mechanisms and acceptability/feasibility, a design combining a RCT with a qualitative component will be adopted. The qualitative component will assess the efficiency, utility, usability, feasibility, acceptability as well as the mechanisms underlying the intervention. This registration details the trial element of EMPOWER only. Data will be collected from employees working in eligible UK organisations expressing an interest to MHFAE in undertaking MHFA training. Organisations expressing an interest (directly to MHFAE), who are willing to take part in the study will be asked to contact Graham Durcan (Centre for Mental Health) for detailed information about the study. If willing to take part, organisations will sign a written agreement and their contact details will be sent to the research team. The research team will email the lead contact, detailing what will happen next, asking for company demographics and requesting organisations send a standard email to ALL employees providing information about the study with a link to the survey for completion of baseline measures. Consent will be sought, using Qualtrics, before completing the online survey. One, two and three weeks later the research team will send the organisation lead another standard email, to be sent to all employees, as a gentle reminder to complete the survey. After this period, participants who have not completed the survey will be treated as lost to follow-up. Organisations will be assigned randomly to the intervention or control arm after completing baseline measures, using computer-generated random numbers through Random.org: https://www.random.org/randomness/. An independent researcher not associated with the study will generate the random sequence and communicate this via email to the research team. The unit of randomisation is organisation and randomisation of organisations will be stratified by size of organisation, with three strata being defined22, small (\< 50 employees), medium (50-249 employees) and large (\> 250 employees). To optimize completion of follow-up measures, at each follow-up point the research team will send organisations up to four weekly email reminders to ask employees to complete measures. Participant data will be exported from Qualtrics into SPSS version 22 by the research team (and any hard copies entered manually), stored on a password-protected network drive, will be accessible only to research team members and will only be linked directly with their participant ID code. Any hard copies of data will be destroyed by confidential waste disposal fifteen years after the research findings have been published. Electronic copies of data will be stored in two archives. In both cases, only anonymous data will be archived at London South Bank University archive and a national data repository. Data monitoring will be conducted by the Trial Steering Committee (TSC) comprising a trialist, statistician and health economist and will act as Data Safety Monitoring Board (DSMB). The TSC will oversee the conduct of the trial independently of and on behalf of the project funders and sponsors and ensure it is carried out with reference to good practice. The TSC will have ultimate responsibility for deciding if the trial should be stopped on grounds of efficacy. The TSC is a sub-group of the project Expert Reference Group (ERG) set up by the funders, and will audit the trial conduct through reports submitted to it by the researchers on a three-monthly basis' We anticipate little risk to participants in the study. The Senior Manager in each company will be asked to report to the CI in writing any adverse or untoward incidents that occur because of the study. The CI will assess each adverse event and decide on an appropriate course of action. In case of Serious Adverse Events (SAEs), i.e. those causing serious injury or death, the CI will notify the Ethics Committee and the TSC as soon as possible after being notified. A TSC within the ERG consisting of members independent of the investigators, their employing organisations, funders and sponsor will monitor trial progress and conduct and will have ultimate responsibility for deciding if the trial should be stopped on the grounds of safety. The sample size calculation is based on a simplification of the proposed analysis model. To be able to detect a change of two additional help seeking resources11, with 80% power at the 0.05 alpha level, assuming that the mean in the control group is 1 and in the intervention group 3 with a common standard deviation of 8, 506 participants are required, increasing to 596 allowing for 15% attrition at 6 months. Pilot data indicates a response rate of 40% across organisations and an average cluster size of 72 employees (calculated from organisations currently signed up) therefore approximately 29 employees per organisation (cluster size). With 24 clusters, 12 organisations per condition, and assuming an intraclass correlation coefficient (ICC) equal to 0.01 (ICC's are lower for participant outcomes as opposed to process variables, when cluster size is large and when estimates are adjusted by participant baseline characteristics21) a minimum target sample size is 763, rounded up to 800, 400 per condition All outcomes will be described with the appropriate descriptive statistics: mean and standard deviation for continuous outcomes (or medians and interquartile range for skewed data), and counts and percentages for dichotomous and categorical outcomes. The analysis of the primary outcome will estimate the mean difference (with 95% confidence intervals) in the Actual Help-Seeking score at 6-month follow-up between the intervention (MHFA) and standard care groups using a mixed effects repeated measures model (which assumes incomplete outcome data to be missing at random). The model will incorporate demographic and other baseline covariates as fixed effects. The dependent variable will be the count of help sources on the Actual Help-Seeking questionnaire. The independent variable will be 'time point'. The effect size of the intervention will be estimated as the exponentiated coefficient for the interaction term between time-point (baseline versus 6-months follow-up) and intervention status (MHFA versus control group). A random effect of participant will be included in the model to adjust for the repeated measures on participants. 'Employing organisation' will be included as a higher level random effect (with participants nested in employers). Statistical significance will be at the 5% level. Secondary outcomes will be analysed using an appropriate generalised linear model, for example binary logistic regression for dichotomous outcomes and ordinal logistic regression for ordered categorical outcomes. All models will be adjusted for employing organisation and baseline score (where applicable). Age, gender, level of education, ethnicity, nature and frequency of and MHFA interventions directly experienced will be included in the primary analysis as covariates. The missing at random assumption for primary outcome data will be assessed further in sensitivity analyses. Treatment effects will be estimated under varying assumptions of data being missing not at random using pattern-mixture models. A complete case analysis will also be conducted. Graham Durcan from The Centre for Mental Health (CMH) has oversight of the project management on behalf of the funders. The Chief Investigator \[CI\] (Callaghan) has overall responsibility for the research. A Research Management Group (RMG), which the CI will chair, comprising all authors, will advise and assist on the project's management. The RMG meets every 6 weeks. An Expert Reference Group (ERG), a group of independent research experts and lay people and the CI, will provide subject matter expertise to the funders and work with the CMH and London South Bank University's representatives to guide and oversee the impact of the research. The ERG meets quarterly for the entirety of the project's duration.

United Kingdom