Generating medically-accurate summaries of patient-provider dialogue: A multi-stage approach using large language models

TL;DR

This paper presents a multi-stage approach using GPT-3 to generate accurate medical summaries of patient-provider dialogues by identifying key medical entities and affirmations, improving clinical correctness over baseline methods.

Contribution

The authors introduce a multi-stage, prompt-based method leveraging GPT-3 for medical dialogue summarization, emphasizing entity recognition and affirmation detection for improved accuracy.

Findings

Summaries are clinically accurate and outperform baseline zero-shot methods.

The multi-stage approach improves medical correctness of summaries.

GPT-derived metrics effectively evaluate summarization quality.

Abstract

A medical provider's summary of a patient visit serves several critical purposes, including clinical decision-making, facilitating hand-offs between providers, and as a reference for the patient. An effective summary is required to be coherent and accurately capture all the medically relevant information in the dialogue, despite the complexity of patient-generated language. Even minor inaccuracies in visit summaries (for example, summarizing "patient does not have a fever" when a fever is present) can be detrimental to the outcome of care for the patient. This paper tackles the problem of medical conversation summarization by discretizing the task into several smaller dialogue-understanding tasks that are sequentially built upon. First, we identify medical entities and their affirmations within the conversation to serve as building blocks. We study dynamically constructing few-shot prompts for tasks by conditioning on relevant patient information and use GPT-3 as the backbone for our experiments. We also develop GPT-derived summarization metrics to measure performance against reference summaries quantitatively. Both our human evaluation study and metrics for medical correctness show that summaries generated using this approach are clinically accurate and outperform the baseline approach of summarizing the dialog in a zero-shot, single-prompt setting.