Expected symptom change trajectories for the early identification of probable treatment nonresponse in VA PTSD specialty care clinics: A proof-of-concept.

Abstract

The purpose of measurement-based care (MBC) is to detect treatment nonresponse sufficiently early in treatment to adjust treatment plans and prevent failure or dropout. Thus, the potential of MBC is to provide the infrastructure for a flexible, patient-centered approach to evidence-based care. However, MBC is underutilized across the Department of Veterans Affairs (VA) posttraumatic stress disorder (PTSD) specialty clinics, likely because no actionable, empirically determined guidelines for using repeated measurement effectively are currently available to clinicians. With data collected as part of routine care in VA PTSD specialty clinics across the United States in the year prior to COVID-19 (n = 2,182), we conducted a proof-of-concept for a method of generating session-by-session benchmarks of probable patient nonresponse to treatment, which can be visualized alongside individual patient data using the most common measure of PTSD symptoms used in VA specialty clinics, the PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (PCL-5). Using survival analysis, we first identified the probability of cases reaching clinically significant change at each session, as well as any significant moderators of treatment response. We then generated a multilevel model with initial symptom burden predicting the trajectory of PCL-5 scores across sessions. Finally, we determined the slowest changing 50% and 60% of all cases to generate benchmarks at each session for each level of the predictor(s) and then assessed the accuracy of these benchmarks at each session for classifying treatment responders and nonresponders. The final models were able to accurately identify nonresponders as early as the sixth session of treatment. (PsycInfo Database Record (c) 2024 APA, all rights reserved).