A re-assessment: forecasting unplanned care activity

Abstract

A previously published article on the forecasting of unplanned care (accident and emergency attendances and emergency hospital admissions) detailed three approaches to forecast activity. One of these, a set numerical approach attempted to use the differences between the numbers of emergency admissions (EA) and non-admissions (NA) to forecast them. Further work has shown that the use of two variables, a and b cannot reduce the number of possible values from an infinite to a finite number, without at least one assumption being made on how limit them. The limitations of the approach are shown to link with weaknesses inherent with the available unplanned care data. A new and combined approach with the availability of hospital beds for emergency admissions is suggested.