Application and comparison of time series methods on tuberculosis incidence data: A case study of Zimbabwe 1990-2013
Tuberculosis remains a major global public health problem, especially in countries that are considered as high burden countries. Zimbabwe is considered by WHO as one of the high burden countries and tuberculosis incidence continues to be very high, and therefore, there is continued need for monitoring and predicting tuberculosis incidence in an effort to make the control of tuberculosis more effective. The Box-Jenkins approach, specifically the autoregressive integrated moving average (ARIMA) model, is typically applied to predict the incidence of infectious diseases. This method takes into account changing trends, periodic changes, and random disturbances in time series. Autoregressive conditional heteroscedasticity (ARCH) models are the prevalent tools used to deal with time series heteroscedasticity. Holt Winters (HW) methods also play a significant role in time series forecasting and are especially effective for short term forecasting. In this study, based on the data of the tuberculosis incidence from 1990 -2003 in Zimbabwe, we establish the single ARIMA (2, 2, 1)model, the combined ARIMA (2, 2, 1)-ARCH (1) model, and the HW model, which can be used to predict the tuberculosis incidence successfully in Zimbabwe. Comparative analyses show that the ARIMA and ARIMA-ARCH models perform reasonably well, with the ARIMA model being the best in our case. To the best of our knowledge, this is the first study to establish the ARIMA model and ARIMA-ARCH model for prediction and monitoring the yearly incidence of tuberculosis in Zimbabwe. Based on the results of this study, the ARIMA (2, 2, 1)and ARIMA(2, 2, 1)-ARCH (1) models are suggested to give tuberculosis surveillance by providing estimates on tuberculosis incidence trends in Zimbabwe.