Large Language Models are Few-Shot Clinical Information Extractors

TL;DR

This paper demonstrates that large language models like InstructGPT can effectively perform zero- and few-shot clinical information extraction tasks, even without domain-specific training, and introduces new datasets for benchmarking these tasks.

Contribution

It shows how large language models can be applied to structured clinical NLP tasks and provides new datasets for evaluation in this domain.

Findings

GPT-3 outperforms existing zero- and few-shot baselines.

Models effectively handle span identification, sequence classification, and relation extraction.

New datasets enable benchmarking of clinical information extraction.

Abstract

A long-running goal of the clinical NLP community is the extraction of important variables trapped in clinical notes. However, roadblocks have included dataset shift from the general domain and a lack of public clinical corpora and annotations. In this work, we show that large language models, such as InstructGPT, perform well at zero- and few-shot information extraction from clinical text despite not being trained specifically for the clinical domain. Whereas text classification and generation performance have already been studied extensively in such models, here we additionally demonstrate how to leverage them to tackle a diverse set of NLP tasks which require more structured outputs, including span identification, token-level sequence classification, and relation extraction. Further, due to the dearth of available data to evaluate these systems, we introduce new datasets for benchmarking few-shot clinical information extraction based on a manual re-annotation of the CASI dataset for new tasks. On the clinical extraction tasks we studied, the GPT-3 systems significantly outperform existing zero- and few-shot baselines.