Hospital demand variations: suggested instruments for hospital managers

Hospitals worldwide face variations in demand for inpatient care. The accurate forecasting of future demand assists hospitals in programming short-term needs such as staff and supplies, and long-term needs such as beds and buildings. The existence of the appropriate methodological instruments, applied by hospital managers, could help them smooth down upcoming patient flows. This work presents two such instruments: first, a univariate autoregressive moving average method, and second, a multivariate model. By applying these to the Greek National Health System, we have found significant demand variations. The univariate method provides more accurate forecasting of future unexpected demand.