Building a Natural Language Processing Tool to Identify Patients With High Clinical Suspicion for Kawasaki Disease from Emergency Department Notes.

O RIGINAL C ONTRIBUTION Building a Natural Language Processing Tool to Identify Patients With High Clinical Suspicion for Kawasaki Disease from Emergency Department Notes Son Doan, PhD, Cleo K. Maehara, MD, MMSc, Juan D. Chaparro, MD, Sisi Lu, MS, Ruiling Liu, MS, Amanda Graham, MPH, Erika Berry, Chun-Nan Hsu, PhD, John T. Kanegaye, MD, David D. Lloyd, MD, Lucila Ohno-Machado, MD, Jane C. Burns, MD, Adriana H. Tremoulet, MD, MAS, and the Pediatric Emergency Medicine Kawasaki Disease Research Group Abstract Objective: Delayed diagnosis of Kawasaki disease (KD) may lead to serious cardiac complications. We sought to create and test the performance of a natural language processing (NLP) tool, the KD-NLP, in the identification of emergency department (ED) patients for whom the diagnosis of KD should be considered. Methods: We developed an NLP tool that recognizes the KD diagnostic criteria based on standard clinical terms and medical word usage using 22 pediatric ED notes augmented by Unified Medical Language System vocabulary. With high suspicion for KD defined as fever and three or more KD clinical signs, KD-NLP was applied to 253 ED notes from children ultimately diagnosed with either KD or another febrile illness. We evaluated KD-NLP performance against ED notes manually reviewed by clinicians and compared the results to a simple keyword search. Results: KD-NLP identified high-suspicion patients with a sensitivity of 93.6% and specificity of 77.5% compared to notes manually reviewed by clinicians. The tool outperformed a simple keyword search (sensitivity = 41.0%; specificity = 76.3%). Conclusions: KD-NLP showed comparable performance to clinician manual chart review for identification of pediatric ED patients with a high suspicion for KD. This tool could be incorporated into the ED electronic health record system to alert providers to consider the diagnosis of KD. KD-NLP could serve as a model for decision support for other conditions in the ED. ACADEMIC EMERGENCY MEDICINE 2016;23:628–636 © 2016 by the Society for Academic Emergency Medicine From the Department of Biomedical Informatics, University of California (SD, CKM, JDC, CNH, LOM), San Diego, CA; the Depart- ment of Computer Science, University of Pittsburgh (SL), Pittsburgh, PA; The University of Texas Health Science Center at Hous- ton (RL), Houston, TX; Children’s Healthcare of Atlanta (AG, DDL), Atlanta, GA; the Department of Pediatrics, University of California at San Diego (EB, JTK, JCB, AHT, PEMKDRG), La Jolla, CA; Rady Children’s Hospital San Diego (JTK, JCB, AHT, PEMKDRG), San Diego, CA; and the Emory University School of Medicine (DDL), Atlanta, GA. Members of the Pediatric Emergency Medicine Kawasaki Disease Research Group included Lindsay T. Grubensky, RN, MSN, CPNP-PC, Jim R. Harley, MD, MPH, Paul Ishimine, MD, Jamie Lien, MD, Simon J. Lucio, MD, Seema Shah, MD, and Stacey Ulrich, MD. Received August 4, 2015; revision received November 29, 2015; accepted December 30, 2015. Part of this work was presented in poster format at the Pediatric Academic Societies Annual Meeting in San Diego, CA, on April This work was supported in part by the Patient-Centered Outcomes Research Institute (PCORI), contract CDRN-1306-04819 and NIH grant U54HL108460. The authors have no potential conflicts to disclose. Supervising Editor: Damon Kuehl, MD. Address for correspondence and reprints: Adriana H. Tremoulet; e-mail: atremoulet@ucsd.edu. ISSN 1069-6563 PII ISSN 1069-6563583 © 2016 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12925