Modelling malaria treatment practices in Bangladesh using spatial statistics

Abstract

BackgroundMalaria treatment-seeking practices vary worldwide and Bangladesh is no exception. Individuals from 88 villages in Rajasthali were asked about their treatment-seeking practices. A portion of these households preferred malaria treatment from the National Control Programme, but still a large number of households continued to use drug vendors and approximately one fourth of the individuals surveyed relied exclusively on non-control programme treatments. The risks of low-control programme usage include incomplete malaria treatment, possible misuse of anti-malarial drugs, and an increased potential for drug resistance.MethodsThe spatial patterns of treatment-seeking practices were first examined using hot-spot analysis (Local Getis-Ord Gi statistic) and then modelled using regression. Ordinary least squares (OLS) regression identified key factors explaining more than 80% of the variation in control programme and vendor treatment preferences. Geographically weighted regression (GWR) was then used to assess where each factor was a strong predictor of treatment-seeking preferences.ResultsSeveral factors including tribal affiliation, housing materials, household densities, education levels, and proximity to the regional urban centre, were found to be effective predictors of malaria treatment-seeking preferences. The predictive strength of each of these factors, however, varied across the study area. While education, for example, was a strong predictor in some villages, it was less important for predicting treatment-seeking outcomes in other villages.ConclusionUnderstanding where each factor is a strong predictor of treatment-seeking outcomes may help in planning targeted interventions aimed at increasing control programme usage. Suggested strategies include providing additional training for the Building Resources across Communities (BRAC) health workers, implementing educational programmes, and addressing economic factors.

Modelling the Influence of Covariates of Typhoid Patients at the Tamale Teaching Hospital Using Generalised Linear Models

“…Largely, these concept has been applied mostly to diarrhoea, cholera and malaria [43][44][45], and [46][47][48]. Although these studies have shown great potential in identifying the covariates for this disease, no study has been found in the literature that applied the same analytical tool to the study of length of stay of typhoid patients in Ghana.…”

“…The conditional variance of the Negative Binomial distribution exceeds the conditional mean. A general class of Negative Binomial is given in [42], and [43]. The variance function is…”

Diagnostic approaches to malaria in Zambia, 2009-2014

“…Regularly updated malaria risk maps (Haque et al, 2010b;Reid et al, 2010Reid et al, , 2012Haque et al, 2011;Wang et al, 2014) can help the NMCC to monitor progress and mobilize resources. Distance, population density, treatment seeking behavior and other health facilities should be considered (Ahmed et al, 2009;Haque et al, 2012;Simon et al, 2013). Accessibility (road network) proved to be a significant risk factor in this study for clinical diagnosis.…”

Mapping Urban Expansion and Exploring Its Driving Forces in the City of Praia, Cape Verde, from 1969 to 2015

“…A properly specified OLS regression can predict the variability in the dependent variable [47] since it meets all of the following requirements: (i) coefficients for model explanatory variables, which are statistically significant and have the expected sign; (ii) explanatory variables free of multicollinearity; (iii) model not biased (heteroscedasticity); (iv) residuals normally distributed; (v) no missing key explanatory variables; (vi) residuals free of spatial autocorrelation; and (vii) enough adjusted R-Squared to explain the variability of the dependent variable [48][49][50][51]. The exploratory regression tool (ArcGIS 10.1 (Esri, Redlands, CA, USA) helped us to find a good OLS model by attempting every possible combination for a set of explanatory variables; we selected the one that met all of the requirements of the OLS method and, notably, had a high-adjusted R 2 [52].…”