KEYNOTE SPEECH

Patient Safety Events Prediction and Prevention using Clinical Big Data from Electronic Health Records

Speaker

Hyeoun-Ae Park

• Professor, College of Nursing, Seoul National University, Korea
• President, International Medical Informatics Association

Abstract

The study aims to develop and test a patient safety event prediction and prevention system integrated with electronic health records (EHR) system. Specific objectives are to develop risk prediction models for patient safety events such as fall, pressure ulcer and adverse drug events of chemotherapy using big data analytics and integrate the models into EHR, and to develop an evidence-based CDSS providing healthcare providers with tailored recommendations to prevent and manage these patient safety events.

We used a database containing EHRs of the patients discharged from Jan 1, 2015 to May 31, 2016 in a tertiary teaching hospital in Korea. The case group consists of patients with a safety event and the control group without any safety event during hospital stay. Features were extracted from 10 different data sources and normalized based on criteria such as completeness, currency, structure and granularity of documentation. Significant features were selected using data visualization techniques and univariate statistical analyses. For fall and pressure ulcer, Cox-hazard model was used to predict the time of patient safety event and logistic regression to predict the risk of patient safety event. For side effect of chemotherapy, neural network architectures were used.

The performance of prediction models for measures of binary and survival outcomes was assessed using a variety of different methods and metrics such as goodness of fit statistics, area under the ROC curve, sensitivity and specificity, and true positives.

Bio

Hyeoun-Ae Park is the President of International Medical Informatics Association (IMIA). She is also a professor of College of Nursing at Seoul National University (SNU) teaching Nursing and Medical Informatics at SNU. Prior to joining SNU in 1992, she was a research fellow at Korean Institute of Health and Social Affairs. Dr. Park received her BS in Nursing from SNU, her MS and PhD in Biostatistics and Health Informatics from the University of Minnesota. She had been a visiting scholar at health informatics department at the University of Minnesota, and SNOMED CT International at the College of American Pathologists. Her areas of research are healthcare vocabulary and terminology, especially around ICNP and SNOMED CT. Her recent research focuses on ontology as a framework for social big data, and data mining and big data analytics in healthcare. She served as a vice president of the IMIA in charge of Working Groups and Special Interest Groups for over 6 years from 2007 to 2013. She also served as the chair of NI Special Interest Group of the IMIA from 2012 to 2015. She received 2004 Distinguished Leadership Award for Internationals and School of Nursing 100 Distinguished Alumni Award in 2008 from University of Minnesota. She is a Fellow of American Academy of Nursing. She received Lael Cranmer Gatewood Distinguished Lectureship Award for her leadership in Health Informatics at the 50th anniversary of Health Informatics Program at University of Minnesota.