Abstract
Overcrowding is one of the most crucial issues confronting emergency departments (EDs) throughout the world. Efficient management of patient flows for ED services has become an urgent issue for most hospital administrations. Handling and detection of abnormal situations is a key challenge in EDs. Thus, the early detection of abnormal patient arrivals at EDs plays an important role from the point of view of improving management of the inspected EDs. It allows the EDs mangers to prepare for high levels of care activities, to optimize the internal resources and to predict enough hospitalization capacity in downstream care services. This study reports the development of statistical method for enhancing detection of abnormal daily patient arrivals at the ED, which able to provide early alert mechanisms in the event of abnormal situations. The autoregressive moving average (ARMA)-based exponentially weighted moving average (EWMA) anomaly detection scheme proposed was successfully applied to the practical data collected from the database of the pediatric emergency department (PED) at Lille regional hospital center, France.