Human-centered Perspectives on a Clinical Decision Support System for Intensive Outpatient Veteran PTSD Care

Paper Title

Human-centered Perspectives on a Clinical Decision Support System for Intensive Outpatient Veteran PTSD Care

Publication Info

• Topic area: Clinical decision support systems (CDSS) for psychotherapy in veteran PTSD care.
• Keywords: PTSD, veterans, clinical decision support system, Prolonged Exposure therapy, cognitive behavioral therapy, human-centered design, sensor-captured data, psychotherapy, sociotechnical systems, veteran mental health.

Background and Problem

• Problem / challenge: Psychotherapy for PTSD relies on balancing patient self-reports and clinician intuition, but there is limited exploration of how technology can mediate this tension while maintaining clinical fidelity and contextual fit.
• Significance: PTSD affects a significant portion of veterans, yet evidence-based psychotherapies like Prolonged Exposure (PE) therapy are underutilized. A CDSS could enhance therapy delivery, reduce dropout rates, and improve patient outcomes.
• Motivation and related work: Prior research has explored technological support for mental health, including mobile health and sensor-captured patient-generated data (sPGD). However, challenges remain in designing CDSSs for sensitive clinical contexts, particularly for veteran care, where institutional and cultural factors (e.g., Veterans Affairs, military identity) add complexity.

Solution

• Proposed approach: A clinician-facing CDSS prototype designed to support the delivery of Prolonged Exposure (PE) therapy for veterans with PTSD, focusing on protocol adherence, homework review, and patient engagement.
• Novelty:
  1. Integration of sensor-captured patient-generated data (sPGD) to mediate between patient self-report and clinical intuition.
  2. Human-centered design considerations informed by interviews with PE clinicians and former PE patients.
  3. Application of three sociotechnical perspectives (distributed cognition, situated learning, infrastructural inversion) to analyze and guide CDSS design.
• Procedure and key techniques:
  • Developed a Figma prototype of a clinician dashboard with features like session checklists, homework data visualization, and sPGD integration.
  • Conducted a two-phase interview study with 9 PE clinicians and 7 former PE patients to gather insights on the prototype and therapy context.
  • Analyzed interview data using thematic analysis to identify opportunities and challenges for CDSS design.

Results

• Concrete findings:
  • Identified eight opportunities for a CDSS, including streamlined homework review, aiding patient conceptualization, and supporting clinician workflows.
  • Highlighted three contextual challenges: the role of the CDSS in patient-clinician interactions, skepticism about psychophysiology, and institutional barriers within Veterans Affairs.
• Advantage over baselines:
  • Enhanced data integration and visualization compared to existing tools like PE Coach, which lacks clinician-facing features.
  • Facilitated clinician-patient communication and reduced memory burden for patients.
• Experiments / evaluation:
  • Semi-structured interviews with 9 clinicians (varying experience levels) and 7 former PE patients (all veterans).
  • Prototype evaluation focused on usability, workflow integration, and perceived utility in therapy delivery.
• Limitations and future work:
  • Limited diversity in clinician and patient samples; all clinicians had doctoral degrees, and all patients were veterans from a single program.
  • Future work should explore broader clinician and patient demographics, patient-facing interfaces, and ethical considerations like data privacy and trust.
  • Additional research needed on integrating perspectives from patients' social networks and addressing institutional challenges like VA deployment.

Summary

This study explores the design and implications of a clinician-facing CDSS for Prolonged Exposure (PE) therapy in veteran PTSD care. By integrating sensor-captured patient-generated data (sPGD) and adhering to the PE protocol, the prototype aims to enhance therapy delivery, streamline clinician workflows, and support patient engagement. Interviews with clinicians and former patients revealed opportunities for data-driven insights and challenges related to contextual fit, psychophysiology, and institutional barriers. The findings were reframed through three sociotechnical perspectives—distributed cognition, situated learning, and infrastructural inversion—offering design considerations and future research directions for CDSS deployment in sensitive clinical contexts.