Abstract Body

Background: Linkage and retention in care are central to reducing morbidity and mortality for HIV-infected persons. The Centers for Disease Control and Prevention recommends linkage to care within three months of diagnosis, but 33% fail to link within this period. An HIV-infected person’s residence may be critical to his potential for linkage and retention in care after diagnosis. We hypothesize that within metro Atlanta, low “community linkage to care” and “community viral suppression” may correlate with neighborhood-level variables at the zip code level, including high levels of poverty and income inequality, low educational attainment, high proportion vacant houses, and poor public transportation access. Methodology: We used data from the Georgia Department of Public Health, U.S. Census Bureau American Community Survey, and the Atlanta Regional Commission to examine distributions of neighborhood-level variables and statistical correlations with community linkage to care (average percent linked to care within three months) and community viral suppression (average percent with reported viral load below 200 cells/mL), by zip code tabulation areas (ZCTAs). The analysis included persons living in six Georgia counties comprising the metro Atlanta area with a new HIV diagnosis reported through Georgia’s Electronic HIV/AIDS Reporting System from January 1, 2006 through December 31, 2010. With zip codes as observations, we evaluated correlations between neighborhood-level variables and two primary HIV outcomes, community linkage to care and community viral suppression. Pearson’s and Spearman’s correlations were used to evaluate associations between zip code level variables and primary outcomes. Results: In the 123 included ZCTAs, median HIV incidence was 203 per 100,000 persons. Overall, 53% of people diagnosed with HIV linked to care within three months of diagnosis and 45% achieved viral suppression. 16% of the population was living below the poverty line, and 88% had at least a high school diploma equivalent. Income inequality was not prominent (mean Gini coefficient 0.44). On average, 3% of occupied houses lacked telephone service, and 13% of houses were vacant. There was a median of 54 public bus stops in each zip code. Areas of high poverty and greater number of bus stops correlated with poor linkage to care and lower levels of viral suppression. Higher educational attainment and low proportion vacant houses correlated with greater community virologic suppression. Conclusions: Certain neighborhood-level variables may correlate with community-level linkage to care and viral suppression for HIV-infected persons. Bus stop density may be a marker of high poverty, in addition to an indicator of transportation access, in the linkage and retention in care paradigm.