Development of a methodological and statistical framework for studying mental health interventions with micro-randomised trials

A central question in clinical psychology is: which treatment is effective for whom and under what circumstances? The primary method to determine this treatment effectiveness has been the group randomised controlled trial (RCT). However, RCTs are strongly limited in providing fine-grained insights into what works for whom and under which circumstances. To overcome this, this proposal introduces a within-person approach using micro randomized trials (MRTs), specifically designed for time-varying interventions, particularly in mobile health (mHealth). MRTs involve randomizing participants to intervention or control conditions multiple times, allowing for the assessment of immediate (proximal) and long-term (distal) outcomes of interventions.

This method allows for studying the effects of specific interventions on individuals by computing person-specific models and could be extend to studying the contexts and mechanisms of intervention effectiveness. However, it remains unclear a study that utilised this method can ideally be designed and what (statistical) analyses can optimally support this design to draw robust and valid conclusions.

The main aim of this BDSI grant is to develop a methodological and statistical framework to conduct such MRT studies and analyse data from this novel type of research. The proposed project plan spans six months, divided into three milestones: developing a methodological framework, creating a statistical framework, and integrating these into a comprehensive framework with accessible deliverables. The project aims to further research in personalized mental health treatments, with an emphasis on securing additional funding and disseminating results through various channels including social media, scientific conferences, and open-access articles.