300 Section - Operationalizing the Biopsychosocial-Spiritual Model in Pain Management: A Narrative AI Approach

Pain remains one of the most prevalent yet least understood conditions worldwide, often assessed using numeric scores that overlook psychosocial and spiritual dimensions. The biopsychosocial-spiritual model emphasizes whole-person care but lacks scalable tools for clinical implementation. This study evaluated the feasibility of a narrative AI intake tool designed to capture emotional, social, and spiritual drivers of pain, amplify patient voice, and generate structured clinician summaries for rehabilitation practice across diverse contexts.

Design:

We conducted a feasibility study using GPT-4 and no-code platforms. Simulated patients (n=25) submitted narrative intakes via SMS or email, responding to five guided prompts exploring physical, emotional, and spiritual burdens. Natural language processing identified recurring themes and generated clinician-ready summaries with key quotes and discussion prompts. Outputs were assessed for usability, thematic accuracy, and linguistic fidelity.

Results:

The tool generated usable summaries in 23 of 25 cases (92%). Summaries averaged 182 words, preserved >90% of original patient phrasing, and consistently avoided anthropomorphic or diagnostic claims. Thematic analysis revealed frequent psychosocial and spiritual contributors to pain, including grief (52%), anxiety (48%), mistrust (44%), and resilience (32%). Failures (2/25, 8%) occurred due to incomplete entries. Outputs were concise, thematically accurate, and demonstrated structured extraction of clinically relevant insights not captured by numeric scores.

Conclusion:

This study provides early evidence that narrative AI, grounded in the biopsychosocial–spiritual model, can systematically capture and synthesize whole-person perspectives on pain. By amplifying patient voice and producing concise, clinician-ready summaries, the tool may strengthen empathy, communication, and trust in pain management. Its low-cost, text-based design enhances adaptability across diverse cultural and resource-limited contexts, supporting equitable access to whole-person pain assessment worldwide. Future research will validate clinical performance in real-world populations, assess patient and provider outcomes, and evaluate scalability as a standardized framework for integrating psychosocial–spiritual dimensions into rehabilitation systems globally.