A modular pipeline for natural language processing-screened human abstraction of a pragmatic trial outcome from electronic health records.

ABSTRACT:

Background

Natural language processing (NLP) allows efficient extraction of clinical variables and outcomes from electronic health records (EHR). However, measuring pragmatic clinical trial outcomes may demand accuracy that exceeds NLP performance. Combining NLP with human adjudication can address this gap, yet few software solutions support such workflows. We developed a modular, scalable system for NLP-screened human abstraction to measure the primary outcomes of two clinical trials.

Methods

In two clinical trials of hospitalized patients with serious illness, a deep-learning NLP model screened EHR passages for documented goals-of-care discussions. Screen-positive passages were referred for human adjudication using a REDCap-based system to measure the trial outcomes. Dynam

SUBMITTER: Lee RY 

PROVIDER: S-EPMC12262768 | biostudies-literature | 2025 Sep

REPOSITORIES: biostudies-literature