NLP based congestive heart failure case finding: A prospective analysis on statewide electronic medical records

Yue Wang; Jin Luo; Shiying Hao; Haihua Xu; Andrew Young Shin; Bo Jin; Rui Liu; Xiaohong Deng; Lijuan Wang; Le Zheng; Yifan Zhao; Chunqing Zhu; Zhongkai Hu; Changlin Fu; Yanpeng Hao; Yingzhen Zhao; Yunliang Jiang; Dorothy Dai; Devore S Culver; Shaun T Alfreds; Rogow Todd; Frank Stearns; Karl G Sylvester; Eric Widen; Xuefeng B Ling

doi:10.1016/j.ijmedinf.2015.06.007

NLP based congestive heart failure case finding: A prospective analysis on statewide electronic medical records

Int J Med Inform. 2015 Dec;84(12):1039-47. doi: 10.1016/j.ijmedinf.2015.06.007. Epub 2015 Jul 2.

Authors

Yue Wang¹, Jin Luo², Shiying Hao², Haihua Xu³, Andrew Young Shin⁴, Bo Jin³, Rui Liu², Xiaohong Deng⁵, Lijuan Wang³, Le Zheng², Yifan Zhao³, Chunqing Zhu³, Zhongkai Hu³, Changlin Fu³, Yanpeng Hao², Yingzhen Zhao², Yunliang Jiang², Dorothy Dai³, Devore S Culver⁶, Shaun T Alfreds⁶, Rogow Todd⁶, Frank Stearns³, Karl G Sylvester², Eric Widen³, Xuefeng B Ling⁷

Affiliations

¹ State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou 310027, PR China; Departments of Surgery, Stanford University, Stanford, CA 94305, USA.
² Departments of Surgery, Stanford University, Stanford, CA 94305, USA.
³ HBISolutions Inc., Palo Alto, CA 94301, USA.
⁴ Departments of Pediatrics, Stanford University, Stanford, CA 94305, USA.
⁵ Chongqing Key Lab of Catalysis & Functional Organic Molecules, Chongqing Technology and Business University, Chongqing, China.
⁶ HealthInfoNet, Portland, ME 04103, USA.
⁷ Departments of Surgery, Stanford University, Stanford, CA 94305, USA. Electronic address: bxling@stanford.edu.

PMID: 26254876
DOI: 10.1016/j.ijmedinf.2015.06.007

Abstract

Background: In order to proactively manage congestive heart failure (CHF) patients, an effective CHF case finding algorithm is required to process both structured and unstructured electronic medical records (EMR) to allow complementary and cost-efficient identification of CHF patients.

Methods and results: We set to identify CHF cases from both EMR codified and natural language processing (NLP) found cases. Using narrative clinical notes from all Maine Health Information Exchange (HIE) patients, the NLP case finding algorithm was retrospectively (July 1, 2012-June 30, 2013) developed with a random subset of HIE associated facilities, and blind-tested with the remaining facilities. The NLP based method was integrated into a live HIE population exploration system and validated prospectively (July 1, 2013-June 30, 2014). Total of 18,295 codified CHF patients were included in Maine HIE. Among the 253,803 subjects without CHF codings, our case finding algorithm prospectively identified 2411 uncodified CHF cases. The positive predictive value (PPV) is 0.914, and 70.1% of these 2411 cases were found to be with CHF histories in the clinical notes.

Conclusions: A CHF case finding algorithm was developed, tested and prospectively validated. The successful integration of the CHF case findings algorithm into the Maine HIE live system is expected to improve the Maine CHF care.

Keywords: Congestive heart failure; Electronic Medical record; Natural language processing; Prospective validation; Random forests.

Publication types

Evaluation Study
Validation Study

MeSH terms

Algorithms*
Data Mining / methods*
Decision Support Systems, Clinical / organization & administration
Electronic Health Records / statistics & numerical data*
Heart Failure / epidemiology*
Humans
Maine / epidemiology
Natural Language Processing*
Pattern Recognition, Automated / methods*
Prevalence
Prospective Studies
Reproducibility of Results
Sensitivity and Specificity
Vocabulary, Controlled