Exploring patients' treatment journeys following randomisation in mental health trials to improve future trial conduct: a synthesis of multiple qualitative data sets

Background: The way in which pragmatic trials are designed suggests there are differences between the experiences of participants randomised to usual care and intervention arms. These potential differences relate not only to what treatment participants receive but also how they access and engage with their allocated treatment. Such differences could affect trial results. The aim of this study was to assess whether such differences exist and, if they do, to consider their implications for the design of future trials.

Methods: Interview transcripts were sampled from data sets gathered during three qualitative studies, all of which had been nested within large, primary care depression trials. Each study had explored trial participants' views and experiences of treatments received following randomisation. Transcripts from 37 participants were purposefully sampled, 20 of which were from interviews held with individuals allocated to receive usual GP care. Data were analysed thematically.

Results: There was evidence of differences between trial arms across all 3 data sets. Intervention participants were willing and able to engage with the treatment they had been allocated to. Randomisation had led to them embarking upon a clear treatment pathway and receiving care in a context where they felt comfortable discussing their mental health and had sufficient time to do so. Intervention participants also had continuity with and confidence in the practitioner they saw. A few usual care participants talked about having continuity with and confidence in their GP. However, most of the usual care participants reported a reluctance to consult GPs about mental health, difficulties in securing treatment appointments, and little or no changes in care following randomisation. Additionally, most reported a lack of continuity of care and a lack confidence in the treatment available to them.

Conclusions: There are important differences between usual care and intervention arms that go beyond treatment received and relate to how participants experience accessing and engaging with their allocated care. As these differences could affect trial results, researchers may want to measure or reduce them in order to fully appreciate or control for the range of factors that might affect treatment outcomes.