Cohort design and natural language processing to reduce bias in electronic health records research

Autor: Shaan Khurshid, Christopher Reeder, Lia X. Harrington, Pulkit Singh, Gopal Sarma, Samuel F. Friedman, Paolo Di Achille, Nathaniel Diamant, Jonathan W. Cunningham, Ashby C. Turner, Emily S. Lau, Julian S. Haimovich, Mostafa A. Al-Alusi, Xin Wang, Marcus D. R. Klarqvist, Jeffrey M. Ashburner, Christian Diedrich, Mercedeh Ghadessi, Johanna Mielke, Hanna M. Eilken, Alice McElhinney, Andrea Derix, Steven J. Atlas, Patrick T. Ellinor, Anthony A. Philippakis, Christopher D. Anderson, Jennifer E. Ho, Puneet Batra, Steven A. Lubitz
Jazyk: angličtina
Rok vydání: 2022
Předmět:
Zdroj: npj Digital Medicine, Vol 5, Iss 1, Pp 1-14 (2022)
Druh dokumentu: article
ISSN: 2398-6352
DOI: 10.1038/s41746-022-00590-0
Popis: Abstract Electronic health record (EHR) datasets are statistically powerful but are subject to ascertainment bias and missingness. Using the Mass General Brigham multi-institutional EHR, we approximated a community-based cohort by sampling patients receiving longitudinal primary care between 2001-2018 (Community Care Cohort Project [C3PO], n = 520,868). We utilized natural language processing (NLP) to recover vital signs from unstructured notes. We assessed the validity of C3PO by deploying established risk models for myocardial infarction/stroke and atrial fibrillation. We then compared C3PO to Convenience Samples including all individuals from the same EHR with complete data, but without a longitudinal primary care requirement. NLP reduced the missingness of vital signs by 31%. NLP-recovered vital signs were highly correlated with values derived from structured fields (Pearson r range 0.95–0.99). Atrial fibrillation and myocardial infarction/stroke incidence were lower and risk models were better calibrated in C3PO as opposed to the Convenience Samples (calibration error range for myocardial infarction/stroke: 0.012–0.030 in C3PO vs. 0.028–0.046 in Convenience Samples; calibration error for atrial fibrillation 0.028 in C3PO vs. 0.036 in Convenience Samples). Sampling patients receiving regular primary care and using NLP to recover missing data may reduce bias and maximize generalizability of EHR research.
Databáze: Directory of Open Access Journals
Externí odkaz: