Megan Shepherd-Banigan, Ph.D., obtained her doctorate in Health Services Research from the University of Washington in 2014 and her M.P.H from the University of Miami in 2006. From July 2015-November 2017, Dr. Shepherd-Banigan was a post-doctoral research fellow at the Durham VA Health Services Research & Development (HSR&D) Center of Innovation (COIN) and is now a VA Health Research Scientist in the COIN. Her long-term professional goal is to grow her career as an independent VA health services researcher who applies a diverse array of methods to evaluate and improve the effectiveness of mental health and family service delivery systems. The short-term objective of this application is to describe and strengthen Veteran and family determinants of mental health service utilization. The proposed CDA research seeks to address scientific gaps related to the low use of psychotherapy among Veterans with PTSD:
Aim 1 a? Identify Veteran and family determinants that predict initiation and completion of psycho-therapy for PTSD;
Aim 1 b? Explore mechanisms linking enabling resources, [treatment preparedness], and initiation and completion of PTSD psychotherapy;
Aim 2 ? Refine and test the feasibility of a family-involved treatment preparedness intervention to increase initiation and completion of EBP psychotherapy for PTSD;
Aim 3 ? Analyze whether use of VA supportive services predicts initiation and completion of PTSD psychotherapy by strengthening economic enabling resources. This application is highly innovative and will advance clinical practice by contributing evidence about 1) the role of complex Veteran and family risk factors that inhibit PTSD treatment use and 2) promising, strategies to strengthen these factors, including a family-involved treatment preparedness intervention and VA education and vocational rehabilitation services to increase uptake of PTSD evidenced-based practices (EBPs).21 This project is consistent with VA priorities, such as increased access to care and uptake of EBPs. The research proposed in Aim 1 will employ novel machine learning causal forest methods93 to Caregiver Support Program survey data and VA administrative data to identify Veteran and family predisposing, enabling, and need factors and elucidate how enabling resources (i.e. perceived financial strain, income, education level, family strain) moderate the association between need for treatment and use/completion of psychotherapy. Subsequent qualitative interviews with Veterans and family members will seek to identify the underlying mechanisms between Veteran/family determinants and service use. The goal of Aim 2 is to refine and test the feasibility and acceptability of an existing family engagement strategy74 using a successive cohort design (SCD) among Veterans who were referred to EBP psychotherapy for PTSD.
For Aim 3, VHSC, Veteran Benefits Administration, and Caregiver Support Program data will be combined to examine the effect of VA supportive service use on time to use/completion of psychotherapy treatment using Cox proportional hazard models; sequential stratified matching will be applied to minimize baseline differences between treatment groups. By the end of the 3rd year of CDA funding, a merit award application will be submitted to fund a randomized trial of the Aim 2 intervention. To complete the proposed research, the applicant requires additional skills and content expertise in four areas: [1) intervention and implementation science; 2) advanced methods in qualitative and quantitative research; 3) mental health clinical research and context; and 4) professional development and networking.] Specific topics and activities will include courses in machine learning and qualitative research; workshops about randomized clinical trials and implementation methods; clinical shadowing; and academic conferences and workshops.
Posttraumatic stress disorder (PTSD) is a disabling, chronic, and costly psychiatric condition that is estimated to occur in 23% of Veterans with recent service in Iraq and/or Afghanistan.1 Although there are available treatments for PTSD, studies suggest that fewer than 10% of Veterans receive the recommended dose of psychotherapy;2-4 moreover, the use of efficacious evidenced-based psychotherapy (EBP) is even lower.5 The role of Veteran and family-level determinants in the initiation and completion of an adequate dose of psychotherapy for PTSD is unclear and without robust evidence, efforts to develop and test strategies to strengthen these determinants will remain limited. The proposed research is highly relevant to Veteran health because it will use novel machine learning approaches to identify these risk-factors and test two innovative approaches to strengthen Veterans determinants of PTSD psychotherapy engagement.
