Probabilistic forecasting of hourly emergency department arrivals

An accurate forecast of Emergency Department (ED) arrivals by an hour the day is critical to meet patients’ demand. It enables planners match ED staff number arrivals, redeploy staff, and reconfigure units. In this study, we develop a model based on Generalised Additive Models advanced dynamic exponential smoothing generate hourly probabilistic for prediction window 48 hours. We compare accuracy these models against appropriate benchmarks, including TBATS, Poisson Regression, Prophet, simple empirical distribution. use Root Mean Squared Error examine point assess distribution using Quantile Bias, PinBall Score Pinball Skill Score. Our results indicate that proposed outperform their benchmarks. developed can also be generalised other services, such as hospitals, ambulances or clinical desk services.