Monitoring patient flow in a hospital emergency department: ARMA-based nonparametric GLRT scheme

Overcrowding in emergency departments (EDs) is a primary concern for hospital administration. They aim to efficiently manage patient demands and reducing stress the ED. Detection of abnormal ED (patient flows) systems aids managers obtain appropriate decisions by optimally allocating available resources following attendance. This paper presents monitoring strategy that provides an early alert when abnormally high influx occurs. Anomaly detection using this involves amalgamation autoregressive-moving-average (ARMA) time series models with generalized likelihood ratio (GLR) chart. A nonparametric procedure based on kernel density estimation employed determine threshold ARMA-GLR The developed ARMA-based GLR has been validated through practical data from at Lille Hospital, France. Then, method’s performance was compared other commonly used charts, including Shewhart chart exponentially weighted moving average chart; it proved more accurate.