Neural networks, bayesian networks, and logistic regression for ESI prediction

Seifu Chonde, David Nembhard, Gül E. Okudan Kremer, Omar Ashour

Research output: Contribution to conferencePaperpeer-review

Abstract

Emergency Department (ED) triage is a process of determining illness severity and accordingly assigning patient priority. The Emergency Severity Index (ESI) is a 5-level acuity categorization system that aides in triage by considering patient complaints and resource needs. The objective of this paper is to compare the capabilities of predicting ESI level using ordinal logistic regression, neural networks and naïve Bayesian networks in predicting ESI levels 2 through 5. Data was obtained from Susquehanna Williamsport Hospital for 947 patients over a one month period in 2008. It contained the assigned ESI level, chief complaint, systolic blood pressure, pulse, respiration rate, temperature, oxygen saturation (SaO2) level, age, gender, and pain. A multivariate ordinal logistic regression model was fit using significant univariate logistic regression predictors. Naïve Bayesian Network (NBNs) and artificial Neural Networks (NNs), two machine learning techniques, are proposed to relax the inherent assumptions of linearity and independence in logistic regression. These three techniques are compared using incremental training dataset sizes between 50% and 100% of given data. Considerations of speed, accuracy, data utilization, model flexibility, and interpretability suggest NBNs and NNs for further research.

Original languageEnglish (US)
Pages1478-1487
Number of pages10
StatePublished - 2013
EventIIE Annual Conference and Expo 2013 - San Juan, Puerto Rico
Duration: May 18 2013May 22 2013

Other

OtherIIE Annual Conference and Expo 2013
Country/TerritoryPuerto Rico
CitySan Juan
Period5/18/135/22/13

All Science Journal Classification (ASJC) codes

  • Industrial and Manufacturing Engineering

Fingerprint

Dive into the research topics of 'Neural networks, bayesian networks, and logistic regression for ESI prediction'. Together they form a unique fingerprint.

Cite this