Generating SOAP Notes from Doctor-Patient Conversations Using Modular Summarization Techniques

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, pages 4958–4972 August 1–6, 2021. ©2021 Association for Computational Linguistics 4958 Generating SOAP Notes from Doctor-Patient Conversations Using Modular Summarization Techniques Kundan Krishna, Sopan Khosla, Jeffrey Bigham, Zachary C. Lipton Carnegie Mellon University 5000 Forbes Avenue Pittsburgh, PA {kundank,sopank,jbigham,zlipton}@andrew.cmu.edu Abstract Following each patient visit, physicians draft long semi-structured clinical summaries called SOAP notes. While invaluable to clini- cians and researchers, creating digital SOAP notes is burdensome, contributing to physician burnout. In this paper, we introduce the first complete pipelines to leverage deep summa- rization models to generate these notes based on transcripts of conversations between physi- cians and patients. After exploring a spectrum of methods across the extractive-abstractive spectrum, we propose CLUSTER2SENT, an al- gorithm that (i) extracts important utterances relevant to each summary section; (ii) clus- ters together related utterances; and then (iii) generates one summary sentence per cluster. CLUSTER2SENT outperforms its purely ab- stractive counterpart by 8 ROUGE-1 points, and produces significantly more factual and co- herent sentences as assessed by expert human evaluators. For reproducibility, we demon- strate similar benefits on the publicly available AMI dataset. Our results speak to the benefits of structuring summaries into sections and an- notating supporting evidence when construct- ing summarization corpora. 1 Introduction Electronic health records (EHR) play a crucial role in patient care. However, populating them can take as much time as attending to patients (Sinsky et al., 2016) and constitutes a major cause of physician burnout (Kumar and Mezoff, 2020). In particular, doctors document patient encounters with SOAP notes, semi-structured written accounts containing four sections: (S)ubjective information reported by the patient; (O)bjective observations, e.g., lab re- sults; (A)ssessments made by the doctor (typically, the diagnosis); and a (P)lan for future care, includ- ing diagnostic tests, medications, and treatments. Sections can be subdivided into 15 subsections. In a parallel development, patients increasingly record their doctor’s visits, either in lieu of taking notes or to share with a family member. A budding line of research has sought to leverage transcripts of these clinical conversations both to provide in- sights to patients and to extract structured data to be entered into EHRs (Liu et al., 2019b; Schloss and Konam, 2020; Krishna et al., 2021). In this paper, we introduce the first end-to-end methods for generating whole SOAP notes based on clinical conversations. Our work builds on a unique corpus, developed in collaboration with Abridge AI, Inc.1), that consists of thousands of transcripts of recorded clinical conversations to- gether with associated SOAP notes drafted by a work force trained in the official style of SOAP note documentation. On one hand, this task is much harder than traditional summarization bench- marks, in part, because SOAP notes are longer (320 words on average) than summaries in popu- lar datasets like CNN/Dailymail (Nallapati et al., 2016), Newsroom (Grusky et al., 2018), and Sam- Sum (Gliwa et al., 2019) (55, 27, and 24 words on …

Content truncated to 500 words