Time Series Analysis on Malaria Disease and Contorl in Bauchi State

Abstract

In many developing countries in Africa, the control of non-infectious diseases can bequite challenging due to a combination of factors such as poor housing, inadequate health care facilities, and poor sanitation. Nigeria has been recorded as the second highest country facing such challenges. Despite the measures that were taken by individuals, government and N.G.Os through vector control interventions with the use of insecticide-treated bed nets (ITN) and indoor residual spraying (IRS), but seasonal cases of Malaria couldn’t  be eradicated completely. Furthermore, 120 samples (monthly) data were captured and analyzed for 10year using R-Statistical  Software from the Bauchi State Agency for the control of HIV/AIDS, Tuberculosis, Leprosy and Malaria’s(BACATMA) register and different  Time Series Model(TSM) were estimated and compared to explain the scenario of malaria cases in Bauchi. ARIMA, SARIMA, and  ARFIMAmodels were fitted separately, model parameters were estimated and the model with the best fit was selectedusing AIC in each of  the separate models, and model stationarity and independence were analyzed using the moving average method. Among these good  models, an optimal model was chosen using both Likelihood test and AIC statistically. The ARFIMA(Phi) model performsbetter than  ARIMA(1,0,1) and SARIMA(1,1,1). Unit root tests were conducted using ADFandPP tests. Results confirmed that the data series is  stationary since the p-values 0.01 are less than alpha (0.05) for ADF test statistics and test statistics -5.3369 and -4.2434 which is less  thancritical-value (0.148) at 5% level of significance for PP test respectively and variables in the data set were tested for normality using the  Jargue-Bera test. The result showed that none of the variables is normally distributed as indicated by the p-values for all variables being  less than selected level of significance for the study (0.05). The fore casting model suggests that the numberof malaria cases in Bauchi  state is decreasing probably to extinction. The model has used the moving average method and trend analysis for time series analysis to  predict the future rate of malaria cases in the state.