Brief narrative interventions for adults with chronic illness or psychosocial distress: a scoping review protocol

Comprehensive Summary

Narrative interventions are an under-recognized therapeutic modality for individuals experiencing chronic medical conditions and/or psychosocial stressors. This review protocol provides a framework for understanding existing narrative interventions, which are defined as brief intervention sessions involving the patient expressing personal stories of distress. The goal of this therapeutic technique is to gradually reduce distress through narrative discussion and empathetic listening. The proposed study protocol plans to address the characteristics of narrative interventions including duration, content, and structure. It determines targeted patient populations by assessing patients' clinical diagnosis, psychosocial distress states, and cultural factors. The protocol also aims to determine how to measure and define targeted outcomes of narrative interventions. The protocol authors plan to use the U-M GPT (GPT-4.1, University of Michigan) large language model to extract narrative data, which will be followed by an assessment of the extracted data by a human reviewer.

Outcomes and Implications

The format of narrative interventions appears to be promising in patient populations that require brief forms of interventions such as patients with a short life expectancy or illness-related fatigue. Using artificial intelligence models to extract narrative data broadens our existing data acquisition capabilities allowing us to use narrative forms of data instead of being limited to quantitative data sources, but also introduces potential privacy concerns as the AI tool would have access to confidential patient data prone to data breach.