Spatial visualization of multivariate datasets: an analysis of STD and HIV/AIDS diagnosis rates and socioeconomic context using ring maps.

OBJECTIVES

We used existing data systems to examine sexually transmitted disease (STD) and HIV/AIDS diagnosis rates and explore potential county-level associations between HIV/AIDS diagnosis rates and socioeconomic disadvantage.

METHODS

Using South Carolina county data, we constructed multivariate ring maps to spatially visualize syphilis, gonorrhea, chlamydia, and HIV/AIDS diagnosis rates; gender- and race-specific HIV/AIDS diagnosis rates; and three measures of socioeconomic disadvantage-an unemployment index, a poverty index, and the Townsend index of social deprivation. Statistical analyses were performed to quantitatively assess potential county-level associations between HIV/AIDS diagnosis rates and each of the three indexes of socioeconomic disadvantage.

RESULTS

Ring maps revealed substantial spatial association in STD and HIV/AIDS diagnosis rates and highlighted large gender and racial disparities in HIV/AIDS across the state. The mean county-level HIV/AIDS diagnosis rate (per 100,000 population) was 24.2 for males vs. 11.2 for females, and 34.8 for African Americans vs. 5.2 for white people. In addition, ring map visualization suggested a county-level association between HIV/AIDS diagnosis rates and socioeconomic disadvantage. Significant positive bivariate relationships were found between HIV/AIDS rate categories and each increase in poverty index category (odds ratio [OR] = 2.03; p=0.006), as well as each increase in Townsend index of social deprivation category (OR=4.98; p<0.001). A multivariate ordered logistic regression model in which all three socioeconomic disadvantage indexes were included showed a significant positive association between HIV/AIDS and Townsend index categories (adjusted OR=6.10; p<0.001).

CONCLUSIONS

Ring maps graphically depicted the spatial coincidence of STD and HIV/AIDS and revealed large gender and racial disparities in HIV/AIDS across South Carolina counties. This spatial visualization method used existing data systems to highlight the importance of social determinants of health in program planning and decision-making processes.