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

Background: Natural language processing allows efficient extraction of clinical variables and outcomes from electronic health records (EHRs). However, measuring pragmatic clinical trial outcomes may demand accuracy that exceeds natural language processing performance. Combining natural language proc...

Descripción completa

Detalles Bibliográficos

Publicado en:	Clinical Trials Vol. 23; no. 2; pp. 145 - 155
Autores principales:	Lee, Robert Y, Li, Kevin S, Sibley, James, Cohen, Trevor, Lober, William B, O'Brien, Janaki, LeDuc, Nicole, Mallon Andrews, Kasey, Ungar, Anna, Walsh, Jessica, Nielsen, Elizabeth L, Dotolo, Danae G, Kross, Erin K
Formato:	research tables/charts Journal Article
Publicado:	Sage Publications, Ltd. Apr2026
Acceso en línea:	Ver este registro en EBSCOhost