PREDICTING THE PREVALENCE OF CHILDHOOD DISEASES IN SOUTH WEST NIGERIA USING TIME SERIES MODEL

Abstract

Despite the global decline in mortality rate in southwestern part of Nigeria, the country has failed to record any substantial improvement. Childhood diseases have been discovered in a way that exhibits an unusual pattern of occurrence. In order to predict the prevalence of the diseases, ARIMA/SARIMA models were applied; however the model that best fit the childhood disease is being recommended for future forecasts. Based on the analysis, ARIMA(2,1,4)(1,0,0)12 provided the best fit for children with cough cases in south west region of Nigeria. The trend shows that the prevalence of cough cases were increasing and decreasing and the data was also affected by seasonal variation. The predicted values shows that there will be an increase of whopping cough cases in children of south West region Nigeria from September 2017 to February 2018 and later decrease. Also ARIMA(2,1,3)(1,0,0)12 was considered to be the best model for prevalence of fever cases in children of south west Nigeria because it satisfied the Box-Jenkins approach of modelling time series data. Lastly, ARIMA(2,1,2) was used to model children measles cases in south west of Nigeria and it was observed that measles cases are no longer strong. It is important to periodically evaluate the prevalence of these diseases in children so that appropriate preventive strategy can be instituted.