Ensuring participant safety in research studies evaluating digital mental health interventions delivered remotely - insights from the PIPA trial of an online parenting intervention to prevent affective disorders in high-risk adolescents.

backgroundMonitoring of adverse events to protect clinical trial participants is imperative but can be challenging for interventions delivered remotely via digital channels. We report our experience of monitoring participant safety in a trial of a web-based parenting intervention to prevent affective disorders in high-risk adolescents.

methodsIn the PIPA trial, parents/carers of at-risk adolescents (aged 11–15 years) randomly received either a web-based personalised parenting programme or online parenting factsheets. Interventions were delivered electronically, and study data were collected remotely. To address challenges of monitoring adverse events, we defined any adolescent participant as requiring a duty of care response if they met specific criteria in their responses to the Short Mood and Feelings Questionnaire (SMFQ) or the Development and Well-Being Assessment (DAWBA) scale. Where a duty of care incident was identified, the family or school were contacted to communicate safety concerns and suggest sources of support. Duty of care responses were compared across trial arms and follow-points using Z-tests to assess whether the intervention, or study participation, were causing participant harms.

resultsTwo hundred thirty-one (45%) of respondents met the duty of care criteria at baseline, 84 of whom met two or more criteria. This fell to 52 (20%) at six months, of whom 41 (78%) had met duty of care criteria at baseline. The corresponding figures at fifteen months were 72 (29%) and 52 (72%) respectively. Duty of care responses did not differ significantly across arms at any time point. Approximately 100 h of researcher time was spent overall on engaging with those who indicated a duty of care need. DISCUSSION: More participants required a duty of care response from the study team at baseline than anticipated, reflecting lack of mental health service capacity and probable after-effects of COVID-19 on adolescent mental health. Duty of care data provided reassurance that neither the intervention nor study participation caused harm. It also guided targeted direct contact from study researchers, which was welcomed by participants, but added to study team workload.

conclusionStudies evaluating digital mental health interventions delivered remotely should consider how they identify those who might require direct support and ensure study processes and resources are able to accommodate this need. In such studies, a duty of care management-based approach can be a useful way to monitor potential harms.

trial registrationISRCTN63358736. Registered 18 September 2019.