Managing demand efficiently in emergency departments (ED) has become an important task for decision makers of hospitals. Currently, decision makers focus on improving strategies for optimally managing flow of patients and overcrowding in EDs. Since time is very critical for emergency situations, and can generally mean the difference between life and death, EDs need substantial amount of resources which are indeed limited. In this context, forecasting demand in ED with a minimum error, has noticeable significance for hospitals in planning and managing operations. The objective of this paper was to develop time series models for forecasting demand at the ED of a large scaled training hospital in Izmir, Turkey. Since in winter periods, a significant increase is expected in demand, forecasting demand during winter period is focused. By using Electronic Health Record (EHR) of this hospital, demand in ED during 1st of December, 2016 to 28th of February, 2017 were obtained. First 76 days data (1st December to 14th February) were used to test appropriateness and accuracy of different autoregressive integrated moving average (ARIMA), and seasonal autoregressive integrated moving average (SARIMA) models, where remaining 14 days were used to test the performance of them. Daily and periodical (8-hour lengths) forecasts were evaluated and compared. This study shows how time series models are proper in forecasting patient volumes in EDs.
Alan : Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|