Socio-demographic factors associated with HIV infection

Abstract

Background: From the "Kenya Modes of Transmission" (KMOT) study (Haile G. 2008), Kenya has a mixed epidemic with (a) National prevalence ranging from 6.7% (KDHS2003) to 7.4% (KAIS 2007) 7.1% translated to approximately 1.33 million Kenyans aged 15-64 living with HIV in 2007 while casual heterosexual sex contributing two thirds new infections and (b) Great regional variations, ranging from almost 1% (Northeast Province) to above 12% (Nyanza Province, and up to 30% in some fishing communities of Districts adjacent to Lake Victoria area). The KMOT study recommended that more research on the behaviour and mapping of most-at-risk individual(s), cultural issues requiring behaviour change and uptake of HIV services be conducted to establish where, how and who gets the next new HIV infection. Such research would require the use of geospatial multilevel and multidimensional survey Methods: All 13,060 participants who consented to be counseled and tested for HIV drawn from a population under continuous demographical surveillance were linked to their homesteads and geolocated using a geographical information system (accuracy of <2 m). Individuals and/or families were counseled and tested as per national counseling and testing guidelines. In addition to capturing their behavioral, social and demographic variables, spatial coordinates of their homesteads were recorded as well. Point patterns were used to show geospatial variations while spatial auto-correlations were used to produce robust estimates of HIV prevalence that varied across continuous geographical space and discreet demographic factors (sex, age, marital status, education and occupation). Neighborhood spatial effects with a threshold of 0.135 degrees (15 km radius) were also applied to identify clusters of prevalence (P < 0.05). Results: The results reveal considerable geographical variation in local HIV prevalence (range = 0.001-2.16) within this relatively homogenous population and provide clear empirical evidence for the localized clustering of HIV infections. Four discreet spatial clusters with filled contours ranging from 0.02-0.97 were identified by using ordinary least squares autocorrelation after testing for residual auto-correlation (P<0.05), within the study area. Conclusions: The findings show the existence of several clustered HIV prevalence of varying intensity contained within the study community. Despite the overall low prevalence of HIV in Kalimoni sublocation, the results support the need for interventions that target socio-geographic spaces (clustered villages) at greatest risk to supplement measures aimed at the general population.