Artificial Neural Networks and risk stratification models in Emergency Departments: The policy maker's perspective

tThe primary goal of Emergency Department (ED) physicians is to discriminate betweenindividuals at low risk, who can be safely discharged, and patients at high risk, who requireprompt hospitalization. The problem of correctly classifying patients is an issue involvingnot only clinical but also managerial aspects, since reducing the rate of admission of patientsto EDs could dramatically cut costs. Nevertheless, a trade-off might arise due to the needto find a balance between economic interests and the health conditions of patients.This work considers patients in EDs after a syncope event and presents a comparativeanalysis between two models: a multivariate logistic regression model, as proposed by thescientific community to stratify the expected risk of severe outcomes in the short and longrun, and Artificial Neural Networks (ANNs), an innovative model.The analysis highlights differences in correct classification of severe outcomes at 10 days(98.30% vs. 94.07%) and 1 year (97.67% vs. 96.40%), pointing to the superiority of NeuralNetworks. According to the results, there is also a significant superiority of ANNs in termsof false negatives both at 10 days (3.70% vs. 5.93%) and at 1 year (2.33% vs. 10.07%). However,considering the false positives, the adoption of ANNs would cause an increase in hospitalcosts, highlighting the potential trade-off which policy makers might face.