Can we use Large Language Models (LLMs) to assess the chronic pain experience? (Preprint)

BACKGROUND Chronic pain is a frequent problem in society, having an enormous impact. Tools that enhance the assessment to understand better people with pain experiences are essential to provide the care people need. In this line, a qualitative approach based in written narratives (WNs) from the people suffering chronic pain can be quite useful as supported in different studies. However, the assessment from this perspective can be time-consuming. OBJECTIVE This study explores the feasibility of employing LLMs to assess WNs of people with chronic pain. At the end, we want to evaluate the potential of applying this LLMs to assist clinicians in assessing patients’ pain. METHODS We performed an experiment based on a list of pain narratives made by people with fibromyalgia and qualitatively evaluated in Serrat et al.[17]. Focusing on pain severity and disability, we prompt GPT-4 to assign scores and scores' explanations, to these narratives. Then we quantitatively compare GPT-4 scores with experts' scores of the same narratives, employing statistical measures such as Pearson correlations, Root Mean Squared Error (RMSE), Gwet's AC2 and Krippendorff's α. Additionally, experts specialized in chronic pain conducted a qualitative analysis of the scores' explanation to assess their accuracy and potential applicability of GPT's analysis for future pain narrative evaluations. RESULTS Our analysis reveals that GPT-4's performance in assessing pain narratives yielded promising results. GPT-4 was comparable in terms of agreement with experts, correlations with standardized measurements, and error rates. Moreover, experts generally deemed the ratings provided by GPT-4, as well as the scores' explanation, to be adequate. CONCLUSIONS These findings underline the potential of LLMs in facilitating the assessment of pain narratives, offering a novel approach to understanding and evaluating patient pain experiences. The integration of automated assessments through LLMs presents opportunities for streamlining and enhancing the evaluation process, paving the way for improved patient care and tailored interventions in the realm of chronic pain management.