Forecasts of Performance Indicators in the Health System Using the Arima Method

Abstract

This paper presents quarterly forecasts on several performance indicators from the Romanian health system, from a county emergency hospital. Using data from the period 2010-2022, forecasts are made for the period 2023-2025 of the average duration of hospitalization, the rate of bed utilization, the index of complexity of cases, the number of cases and the average cost of hospitalization. The method used is that of the auto-regressive integrated moving average (ARIMA) applied to time series. The Dickey-Fuller test is used to check the stationarity of the time series, as well as other tests for the validation of prediction models.