MODELING DIABETES PATIENTS' ATTENDANCE AT BAGAMOYO DISTRICT HOSPITAL USING AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (ARIMA) MODELS.

Abstract

This study presented the use of Autoregressive Integrated Moving Average (ARIMA) technique to modeling the diabetes patients' attendance at Bagamoyo district hospital using monthly time series data. The data used in the analysis of this study are monthly reading of diabetes patients data covered the period from January 2014 to December 2021. The data were retrieved from the hospital electronic health management information system. The Autoregressive Integrated Moving Average (ARIMA) approach was applied to the diabetes patients' data through the model identication, estimation, diagnostic checking, stationary and forecasting in R statistical software. The study identied Autoregressive Integrated Moving Average ARIMA (0, 1, 1) model to be the best one to t for the monthly total number of diabetes patients' attendance hospital outpatient department for seven years of monthly data. This was veried by Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC) of the model selection criteria, Autoregressive Integrated Moving Average ARIMA (0, 1, 1) model shown the smallest values, hence this was selected as an adequate model to represent the Bagamoyo district hospital diabetes patients data. The forecast values indicate clearly that diabetes patients' attendance at Bagamoyo district hospital would be on an increase rate per month during the year 2022. Therefore, the government needs to put in place more and to engage necessary requirements for a satisfying healthcare system by increasing the medical supplies to the Bagamoyo district hospital.