GIS and correlation analysis of geo-environmental variables influencing malaria prevalence

Analysing the significance of geo-environmental variables influencing malaria incidence will help
decision makers design area-specific interventions for tackling the menace, particularly in high risk
areas. This study applied geocoding and raster extraction functionalities in GIS (ArcMap) and
Pearson correlation in SPSS to identify the relationship between five geo-environmental variables
and malaria incidence. It first geocoded malaria incidence data and extracted the corresponding
values for five geo-environmental variables in ArcMap 10.1. The five geo-environmental variables
are: distance to marshy areas, distance to watercourses (rivers and streams), soil water retention
capacity, elevation and population. Pearson correlation was then used to find the relationship
between the variables and malaria incidence. The study also applied spline interpolation technique
to map malaria prevalence in the district using standardised malaria incidence. The result indicates
that distance to marshy areas are inversely and significantly (at 1% level) related to malaria
incidence. This means that malaria incidence decreases as distance to marshy areas increases. The
distance to watercourses and elevation are also inversely related to malaria incidence in the study
area. This means that as distance to watercourses increases and elevation rises, malaria incidence
decreases. However, these relationships are not statistically significant at any of the conventional
levels of significance (p<0.01 and p<0.05). The result also indicates that water retention capacity
of different soils and population are positively related to malaria incidence. This means that malaria
incidence rises with increases in the two variables, but the relationships are not statistically
significant at any of the conventional levels. The study concluded that the null hypothesis (H0) that
there is no significant relationship between distance to marshy areas and malaria incidence may be
rejected at 1% (P<0.01) level of significance. However, there is not enough evidence to reject the
null hypothesis (H0) that there is no significant relationship between distance to watercourses,
different soil retention capacity, elevation and population, on the one hand, and malaria incidence
on the other. It is, therefore, recommended that much broader settlement planning policies be
adopted to curb building in those areas that are malaria prone.