Natural language processing system identifies care quality through EHRs

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - EHR Round Up

Researchers have developed a natural language processing (NLP) system capable of measuring the quality of heart failure inpatient care by analyzing data collected from electronic health records (EHRs), according to a study published in JMIR Medical Informatics.

Developed with the aim to accurately automate quality measures for patients with heart failure within the U.S. Department of Veterans Affairs (VA), the NLP Congestive Heart Failure Information Extraction Framework (CHIEF) identified heart failure patients with left ventricular ejection fraction (LVEF) less than 40 percent and where an angiotensin-converting enzyme inhibitor or angiotensin-receptor blocker was prescribed when patients were discharged.

“Congestive heart failure is a prevalent condition and CHIEF is an application that could provide an automated first review for heart failure patients to assess guideline-concordant care,” wrote first author Jennifer Hornung Garvin, MBA, PhD and colleagues.

In this study, researchers collected data from 1,083 inpatients form eight VA medical centers to train and test CHIEF. Additionally, 15 interviews were conducted for stakeholder feedback on whether the system could be implemented elsewhere.

Results showed CHIEF classified hospitalizations with 98.9 percent sensitivity and a positive prediction at 98.7 percent, in increase from the sensitivity and reference standard of 98.5 percent for External Peer Review Program evaluations. Additionally, the system was able identify mentions and values of LVEF and medications with high to fair recall. Medications were examined with recall at 97.8 to 99.7 percent and precision at 96 to 97.8 percent. Mentions of LVEF were examined with recall at 97.8 to 98.6 percent and precision of 98.6 to 99.4 percent.

“Our results demonstrate that automated methods using NLP can improve the efficiency and accuracy of data collection and facilitate more complete and timely data capture at the time of discharge, at a potentially reduced cost,” concluded Hornung and colleagues. “The next step is to transform health care big data into actionable knowledge for quality improvement and research that helps to improve patient care, and potentially limit health care costs, with the aim of developing infrastructure with real-time data to support decision making.”