ADULTS NEWLY INFECTED WITH HIV IN BOTSWANA: A BOX-JENKINS ARIMA APPROACH

Abstract

Using annual time series data on the number of adults (ages 15 and above) newly infected with HIV in Botswana from 1990 – 2018, the study predicts the annual number of adults who will be newly infected with HIV over the period 2019 – 2025. The paper applies the Box-Jenkins ARIMA technique. The diagnostic ADF tests show that the K series under consideration is an I (1) variable. Based on the AIC, the study presents the ARIMA (4, 1, 0) model as the optimal model. The residual correlogram shows that the presented model is very stable and suitable for forecasting new HIV infections in adults in Botswana. The findings of the study reveal that the numbers of new HIV infections in adults in the country are likely to go up from about 7907 in 2019 to almost 8262 new infections by 2025. This means that Botswana’s quest for an AIDS free society will remain a pipeline dream at least from the next 5 years. This is not a desirable public health outcome but rather a warning signal for policy makers so that they make the necessary decisions now before the situations goes out of hand. If nothing is done now, Botswana’s HIV/AIDS burden will definitely rise and possibly overwhelm the country’s limited resources. We basically encourage the government of Botswana to intensify its HIV prevention and control programs in order to possibly reserve the projected trajectory and save precious lives.