How Much Would a Clinician Edit This Draft? Evaluating LLM Alignment for Patient Message Response Drafting

TL;DR

This study evaluates how well large language models can generate patient message responses aligned with individual clinicians' preferences, highlighting the challenges and potential strategies for improving their integration into clinical workflows.

Contribution

It introduces a new taxonomy and evaluation framework for assessing LLMs in clinician response drafting, along with an expert-annotated dataset and comprehensive evaluation of adaptation techniques.

Findings

LLMs show variable ability to generate clinician-aligned responses across themes.

Theme-driven adaptation improves LLM performance in most response themes.

Significant epistemic uncertainty exists in aligning LLM outputs with clinician preferences.

Abstract

Large language models (LLMs) show promise in drafting responses to patient portal messages, yet their integration into clinical workflows raises various concerns, including whether they would actually save clinicians time and effort in their portal workload. We investigate LLM alignment with individual clinicians through a comprehensive evaluation of the patient message response drafting task. We develop a novel taxonomy of thematic elements in clinician responses and propose a novel evaluation framework for assessing clinician editing load of LLM-drafted responses at both content and theme levels. We release an expert-annotated dataset and conduct large-scale evaluations of local and commercial LLMs using various adaptation techniques including thematic prompting, retrieval-augmented generation, supervised fine-tuning, and direct preference optimization. Our results reveal substantial epistemic uncertainty in aligning LLM drafts with clinician responses. While LLMs demonstrate capability in drafting certain thematic elements, they struggle with clinician-aligned generation in other themes, particularly question asking to elicit further information from patients. Theme-driven adaptation strategies yield improvements across most themes. Our findings underscore the necessity of adapting LLMs to individual clinician preferences to enable reliable and responsible use in patient-clinician communication workflows.