Testing an AI Large Language Model Tool for Cognitive Debiasing in Musculoskeletal Care

"As part of the intervention, patients first respond to a series of questions about their beliefs regarding their symptoms (e.g., What's usually behind these symptoms?), with responses transcribed verbatim via tablet. These responses are input into a Large Language Model (LLM), which generates a brief, supportive summary of the patient's beliefs, shared back with the patient to encourage self-awareness and reflection. Patients are then invited to consider prompts such as, What emotions or circumstances might be influencing your thinking? with their reflections again transcribed. The LLM analyzes these reflections to identify potential signs of emotional distress or maladaptive beliefs, and this output is again provided to the patient. The LLM summary of identified maladaptive beliefs is then also shown to the clinician ahead of the consultation to support more tailored, empathetic communication."