Abstract:
A time series is a sequence of data points, typically measured at uniform time intervals. Time series analysis comprises of methods for analyzing time series data in order to extract meaningful characteristics of the data and forecast future values. Box Jenkins approach was used to identify the best Autoregressive Integrated Moving Average (ARIMA) model for secondary data of malnourished HIV positive children under 5 years for the past 10 years i.e from 2006 through 2015.The data source was County health directorate in Narok town. Narok - level 5 hospital is one of the largest hospital in Narok County which is believed to serve the whole population in the county. In developing countries like Kenya severe acute malnutrition is associated with high morbidity and mortality rates in the under 5 year olds especially those from low socio-economic class. The objective of this research work was to fit a time series ARIMA model which can aid in forecasting the level of malnutrition in HIV positive children under 5 years old and to make forecasts of the case in Narok County. The series non-stationarity was corrected through single differencing. The models selected were ARIMA (2, 1, 0) and ARIMA (1, 1, 0). ARIMA (2, 1, 0) had both low Akaike’s information criterion (AIC) and Bayesian information criterion (BIC) values hence being selected to be the best fitted model. The fitted model was used to make forecasts for 5 periods. The mean absolute percentage error (MAPE) was 15.256 using the selected model. This model should be updated continuously incorporation of recent data. We believe that the research findings will form the basis for technical and policy advice to health sector in eliminating the levels of malnutrition in HIV positive children hence reducing morbidity levels leading to improved quality of life in patients from health care services. We made recommendations based on the results. SPSS and Minitab software were used in the analysis of the data.