Forecasting Pediatric Emergency Department Arrivals: Evaluating the Role of Exogenous Variables Using Deep Learning Models

Authors

Egbe Etu Etu, San Jose State UniversityFollow
Jordan Larot, San Jose State University
Kindness Etu, Ingram School of Engineering
Joshua Emakhu, Henry Ford Hospital
Sara Masoud, College of Engineering
Imokhai Tenebe, The University of Chicago Booth School of Business
Gaojian Huang, San Jose State UniversityFollow
Satheesh Gunaga, Henry Ford Wyandotte Hospital
Joseph Miller, Henry Ford Hospital

Publication Date

11-15-2025

Document Type

Article

Publication Title

Intelligence Based Medicine

Volume

12

DOI

10.1016/j.ibmed.2025.100313

Abstract

Background Forecasting pediatric emergency department (ED) demand remains a critical challenge in healthcare operations. This study aimed to identify exogenous variables influencing pediatric ED visits and evaluate the performance of different forecasting models. Method Using a retrospective observational design, we analyzed 192,347 pediatric ED visits across nine hospitals in Southeast Michigan between 2017 and 2019. Patient data were aggregated into daily arrival counts and enriched with exogenous variables such as weather, air quality, pollen, calendar, Google search trends, and chief complaints. Feature selection was performed using XGBoost and SHapley Additive exPlanations to identify the most influential predictors. Three forecasting models were developed: a Naïve baseline, Long Short-Term Memory (LSTM), and an attention-based neural network. The models were evaluated across 1-day, 7-day, and 14-day forecasting horizons using mean absolute percentage error (MAPE) and R² metrics. Results LSTM and attention-based model significantly outperformed the Naïve baseline across all horizons. The LSTM model incorporating calendar data achieved the best 1-day forecast (MAPE: 8.71 %, R²: 0.67). For 7-day forecasts, the attention-based model using chief complaint data performed best (MAPE: 9.18 %, R²: 0.57). At 14 days, the attention-based model without exogenous inputs outperformed most LSTM variants, reflecting superior performance in long-range forecasting. Among exogenous variables, calendar and chief complaint data added the most predictive value, while Google Trends and pollen data introduced noise and diminished model performance. Conclusion Combining deep learning architectures with selected external data improves pediatric ED arrival forecasting. From an operational perspective, such forecasts can support more efficient staffing, reduce wait times, and mitigate ED crowding.

Keywords

Attention-based neural network, Children, Deep learning, Forecasting, Hospital, Long short-term memory, Pediatric emergency department

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 License.

Department

Marketing and Business Analytics

Recommended Citation

Egbe Etu Etu, Jordan Larot, Kindness Etu, Joshua Emakhu, Sara Masoud, Imokhai Tenebe, Gaojian Huang, Satheesh Gunaga, and Joseph Miller. "Forecasting Pediatric Emergency Department Arrivals: Evaluating the Role of Exogenous Variables Using Deep Learning Models" Intelligence Based Medicine (2025). https://doi.org/10.1016/j.ibmed.2025.100313