A Study of Asthma Cases among Under-Five Children using Mixture Poisson Ar(Q) Model

Abstract

This study investigates asthma cases among children under five in Mubi North LGA of Adamawa State. Secondary data was obtained from National Health Management Information System as recorded in the Mubi Zonal center of Adamawa State for a period of 3 years. Through the use of Mixture Poison Autoregressive AR(q) process, a time series model was estimated. The results revealed through descriptive statistics analysis a multimodal distribution in the cases of asthma in the study area, indicating several peaks that may be attributed to distinct risk factors or subpopulations affecting the incidence of asthma. In the case of the Mixture Poisson Autoregressive model, the results also show that a three-component Mixture Poisson Autoregressive model offers the best fit to the observed asthma data. The estimated parameters of the model show considerable temporal correlations among the components and fluctuating rates of asthma cases. The significant (p<5%) mixing proportions suggests that the third component plays an important role in the model, indicating a higher average rate of asthma cases. The results for the out-of-sample forecasts demonstrated considerable variability in predicted asthma cases for the next 6 days, reflecting the influence of external factors