Time Series Forecasts of Emergency Department Patient Volume, Length of Stay, and Acuity

摘要: Abstract Study hypothesis: Time series analysis can provide accurate predictions of emergency department volume, length stay, and acuity. Design: Prospective stochastic time modeling. Setting: A university teaching hospital. Interventions: All patients seen during two sequential years had arrival, discharge, acuity recorded in a computer database. variables were formed for arriving per hour, Prediction models developed from the year 1 data included five types: raw observations, moving averages, mean values with seasonal indicators auto-regressive integrated averages. Forecasts each model compared observations first 25 weeks 2. Model accuracy was tested on residuals by autocorrelation functions, periodograms, linear regression, confidence intervals variance. Results: There 42,428 44,926 Large periodic variations patient volume day found ( P Conclusion : powerful, short-range forecasts future ED volume. Simpler performed best this study. stay are not likely to contribute additional useful information staffing resource allocation decisions. [Tandberg D, Qualls C: Ann Emerg Med February 1994;23:299-306.]