INTERVENTION ANALYSIS OF MALARIA CONTROL PROGRAMME IN NORTHERN REGION OF GHANA

Abstract

Malaria has been a major public health problem in sub-Saharan Africa for a long period with Ghana not an exception. In this study, we investigated the impact of various malaria control interventions in the Northern region of Ghana through intervention analysis. Data on monthly Out-patient malaria cases, In-patient cases and Deaths obtained from the Northern Regional Health Directorate were modelled using SARIMAX and ARIMAX models. The results revealed that SARIMAX (1, 1, 1)(1, 1, 1)12 was the best model for the Out-patient cases since it had the least AIC, AICc and BIC values of -53.2994, -52.9515 and -39.9352 respectively. Also, SARIMAX (1, 1, 1)(2, 1, 1)12 emerged as the best model for the In-patient cases with AIC, AICc and BIC values of37.9810, 38.3288 and 52.0181 respectively. Furthermore, ARIMAX (1, 1, 1) was the best model for the Deaths with the least AIC, AICc and BIC values of 171.4890, 171.5916 and 179.8264 respectively. Diagnostic checks of the three models with the Ljung-Box test and ARCH-LM test revealed that all the models were free from higher-order serial correlation and conditional heteroscedasticity respectively. A CUSUM test on the residuals of the models all revealed that the models were structurally stable over time. A chi-square goodness of fit test revealed that there was no significant difference between the predicted values from the models and the observed values for the year 2014. The study further revealed that the coefficients of the intervention variable for the Out-patient, In-patient and Death cases were all negative. These were -0.1467 for the Out-patient, -0.0262 for the In-patient and -1.8809 for the Deaths. It was therefore concluded that the intervention brought about a decline in the number of Out-patient, In-patient and Death cases