Longitudinal Mediation Analysis to Identify Effective Intervention Components in Clustered Trial of RAISE-ETP (Recovery after an Initial Schizophrenia Episode-Early Treatment Program)

Project Details


Project Summary/Abstract The goal of this project is to develop and adapt methods to conduct longitudinal mediation analysis to identify the most effective components in combined pharmacological and psychosocial therapy (including supported education and employment) for improving functional outcomes among individuals in their first episode of psychosis (FEP). We will use the data from a two-year clustered randomized clinical trial: the NIMH-funded the Recovery After an Initial Schizophrenia Episode (RAISE) Early Treatment Program (ETP). This trial that showed NAVIGATE (a multifaceted intervention) resulted in significantly greater improvement in functional outcome of quality of life and the symptom severity over the two year period than Community Care at 34 sites representing real-world settings. There were four intervention components in NAVIGATE - personalized medication management, family psychoeducation; individual resilience therapy; and supported education and employment (SEE). It is critical to understand the pathways the intervention led to the improvement of the functional outcome and the symptoms. The intervention may operate through the enhanced engagement of patients with the treatments as well as increased therapeutic alliance which has been widely recognized as a critical factor for treatment success across different treatments and patient groups. This project examines the longitudinal adherence of the four intervention components in NAVIGATE as well as therapeutic alliance for their mediation effects on improving the functional outcome and symptom severity. The mediation analysis will help us to understand the mechanisms yielding functional recovery through coordinated specialty care for first- episode psychosis and will identify effective and more efficient bundles of services to enhance treatment delivery. Because treatment adherence at patient level for the various intervention components and therapeutic alliance are variable processes during the clustered trial and some of these components were measured repeatedly in discrete fashion, the traditional mediation analysis would not be adequate to validly assess the mediation effects of these intervention components. To address the more complex mediation questions in longitudinal trial for psychiatric disorder, we will use a general framework that handles longitudinal and discrete mediators for the analysis of the clustered trial. Additional sensitivity analysis to examine the sensitivity to the analytic assumptions in the mediation analysis and simulation study to investigate the bias in the estimation due to clustering will also be conducted. Once a component of the intervention is identified as more effective in our proposed mediation analysis, it could be used to guide for a subsequent study of enhanced and more efficient treatment delivery.
Effective start/end date5/1/174/30/19


  • National Institute of Mental Health: $83,750.00


  • Psychiatry and Mental health


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.