Comparative Spatial Distribution of Five Infectious Diseases in Ghana

Abstract

Aims: To use exploratory geo-health informatics in probing if there are spatial homogeneities in the incidence rates and spatial co-distribution or interactions within the five commonly reported infectious diseases in Ghana. Study Design: Ecological study design was employed. Place and Duration of Study: District-level reported cases of the five commonly reported infectious diseases in Ghana for the periods 2010 to 2014 inclusive. Methodology: A geo-relational database for at-risk population and disease morbidities was created using ArGIS 10.1 for the 170 districts in Ghana. The incidence rates were computed and spatially standardised with empirical Bayesian smoothening. The spatial analyses of disease mapping, followed by spatial pattern and cluster-outlier analyses with global and local spatial autocorrelations were performed. Disease interactions were also evaluated with Pearson correlation and Moran's I bivariate analyses. Results: The results from the Bayesian smoothed maps, global and local Moran's indices suggested existence of statistically significant spatial variations in the incidences of Malaria (P =0.00), Diarrhoea (P =0.04), Typhoid fever (P = 0.00) and Intestinal worms infestation (P = 0.00) with positive Moran's I in the country with pockets of elevated rates in some locations. The Upper Respiratory Tract Infection (URTI), however, was not significant (P = 0.35). The magnitude of the z-statistics, indicating the increasing intensity of spatial clustering among the five diseases was URTI (z = -0.94), Diarrhoea (z = 2.10), Typhoid fever (z = 3.03), Malaria (z = 3.60) and Intestinal worms (z = 6.91). The number of disease-specific hotspot areas detected were 4 for Malaria, 8 each for URTI and Diarrhoea, 9 for Typhoid fever and 12 for Intestinal worms. Bivariate analyses of Pearson correlation and Moran's index established significant positive disease interactions between the incidences over space. The spatial disease interactions or co-distributions with positive Moran's indices were very strong in Malaria-URTI (r=0.53), Malaria-Diarrhoea (r= 0.56), URTI-Intestinal worms (r= 0.57), and strongest co-distribution was observed in Diarrhoea-URTI (r=0.83). Conclusion: Among the five infections, Malaria was receiving the most spatially homogeneous health interventions but serious spatial inequalities in the other infections, especially Intestinal worms with the highest number of hotspots and greatest clustering intensity. The identified hotspot districts critically need further strengthening or improving disease-specific control strategies. The significant positive disease interactions implies possible common risk factors and that disease-and-location targeted interventions could be strategically developed to yield simultaneous or multiple disease intervention outcomes while considering the impact of the environment and neighbourhood in relation to disease clusters. Public health planners should include spatial components in health policy and practices to optimise public health administration.