What question did this study set out to answer?

This research aims to predict emergency department patient arrivals diagnosed with upper respiratory infections.

March 27, 2026Open Access

Forecasting Patients Admitted to Emergency Departments with the Diagnosis of Upper Respiratory Tract Infection Using Time Series and Artificial Neural Network Modelling

DCDilber CoşkunMunzur University SOSumeyye OzbeyMunzur University MYMelih YücesanMunzur University

Key Points

This research aims to predict emergency department patient arrivals diagnosed with upper respiratory infections.
Analyzed patient arrival data from a training and research hospital
Used meteorological data including temperature, humidity, and wind for predictions
Employed forecasting methods such as SARIMAX, ANN, and NARX
Compared the performance of different predicting models
Hourly forecasting results using various models were obtained
The SARIMAX model demonstrated the best performance compared to others
Predictions can assist in effective planning of emergency department resources

Abstract

Emergency departments are vital units that operate 24/7, care for critically ill patients, and provide immediate emergency care in accordance with the triage code for admitted patients. The efficient operation of emergency services depends on adequate human and medical resources and on early planning. The increase in the intensity of emergency services due to COVID-19, a biological disaster, has limited the effective use of resources and planning efforts. Overcrowding in emergency rooms can disrupt services and endanger patients' lives. For this reason, it is essential to organize emergency service units according to patient estimates, provide service at an optimal level, facilitate planning and management, use medical and human resources effectively, and ensure patient satisfaction. This study was conducted to predict emergency department patient arrivals using meteorological data. In this study, hourly forecasting results are obtained using estimation methods, including the seasonal autoregressive integrated moving average (SARIMAX), an artificial neural network (ANN), and a nonlinear autoregressive with exogenous inputs (NARX) model. For the study, patient arrival data from a training and research hospital for December 2021 and meteorological data, including temperature, humidity, and wind, were used. Methods for predicting emergency department patient admissions were compared, and the SARIMAX model performed best. It is thought that predictions based on meteorological data will contribute to emergency department planning.

Ask AI

Helpful

Bookmark

View Full Paper