Towards Adapting Open-Source Large Language Models for Expert-Level Clinical Note Generation

TL;DR

This paper adapts open-source LLaMA-2 models for expert-level clinical note generation, combining domain-specific training, reinforcement learning, and a new distillation approach, achieving physician-level quality in generated notes.

Contribution

It introduces a comprehensive adaptation process for open-source LLMs to produce high-quality clinical notes, including a novel reinforcement learning method called DistillDirect.

Findings

LLaMA-Clinic generates clinically acceptable notes in 92.8% of evaluations.

The model matches physician notes in the 'Assessment and Plan' section.

Physician ratings show high agreement with expert standards.

Abstract

Proprietary Large Language Models (LLMs) such as GPT-4 and Gemini have demonstrated promising capabilities in clinical text summarization tasks. However, due to patient data privacy concerns and computational costs, many healthcare providers prefer using small, locally-hosted models over external generic LLMs. This study presents a comprehensive domain- and task-specific adaptation process for the open-source LLaMA-2 13 billion parameter model, enabling it to generate high-quality clinical notes from outpatient patient-doctor dialogues. Our process incorporates continued pretraining, supervised fine-tuning, and reinforcement learning from both AI and human feedback. We introduced a new approach, DistillDirect, for performing on-policy reinforcement learning with Gemini 1.0 Pro as the teacher model. Our resulting model, LLaMA-Clinic, can generate clinical notes comparable in quality to those authored by physicians. In a blinded physician reader study, the majority (92.8%) of individual evaluations rated the notes generated by LLaMA-Clinic as "acceptable" or higher across three criteria: real-world readiness, completeness, and accuracy. In the more challenging "Assessment and Plan" section, LLaMA-Clinic matched physician-authored notes in real-world readiness score. We highlight key considerations for future clinical note-generation tasks, emphasizing the importance of pre-defining a "best practice" note format, rather than relying on LLMs to determine this for clinical practice.