Natural Language Processing Methods to Identify Oncology Patients at High Risk for Acute Care with Clinical Notes.

PubMed, vol. 2023, pp. 138–147

Abstract

-penalised logistic regression with SHD achieved a C-statistic of 0.748 (95%-CI: 0.735, 0.762), while the same model with language features achieved 0.730 (95%-CI: 0.717, 0.745) and a transformer-based model achieved 0.702 (95%-CI: 0.688, 0.717). This paper shows how language models can be used in clinical applications and underlines how risk bias is different for diverse patient groups, even using only free-text data.

Authors 3

1. Claudio Fanconi

   ETH Zurich · Stanford University

   Affiliation as printed

   Stanford University, Stanford, California, United States

   ETH Zürich, Zürich, Switzerland

2. Marieke Meija van Buchem Aachen

   Stanford University · Leiden University Medical Center

   Affiliation as printed

   Stanford University, Stanford, California, United States

   Leiden University Medical Center, Leiden, The Netherlands

3. Tina Hernandez‐Boussard

   Stanford University

   Affiliation as printed

   Stanford University, Stanford, California, United States