Paper # 33|
Decreases in Community Viral Load Are Associated with a Reduction in New HIV Diagnoses in San Francisco
Moupali Das-Douglas*1,2, P Chu1,2, G-M Santos1,2, S Scheer1, W McFarland1,2, E Vittinghoff2, and G Colfax1,2
1San Francisco Dept of Publ Hlth, CA, US and 2Univ of California, San Francisco, US
Background: We hypothesized that the decrease in
community viral load (CVL) in San Francisco from 2002 to 2008 would be associated
with a reduction in new HIV infections.
Methods: We used San Francisco’s comprehensive
HIV/AIDS surveillance system, which includes mandatory laboratory reporting of
viral loads, to calculate 2 distinct measures of CVL (mean and total), and to
examine trends in CVL overall and within different populations. We defined mean
CVL as the mean of the most recent viral load of all reported HIV+
individuals in a particular population, divided by the number of reported HIV+
individuals in the population. Total CVL is the sum of the most recent viral
loads of all HIV+ individuals in a particular population. Outcomes
included annual numbers of new HIV diagnoses (from 2004 to 2008), changes in estimated
HIV incidence from 2006 to 2007. To assess the relationships of CVL with these
outcomes, we first plotted them against CVL, and then used Poisson models with
robust standard errors for inference.
Results: Both measures of CVL (mean and total) decreased
from 2002 to 2008 and were accompanied by decreases in new HIV infections, as
measured by both new diagnoses of HIV cases and HIV incidence. The trend in mean
CVL (Figure 1) was significantly associated with newly diagnosed HIV cases from
2004 to 2008 (P =0.003) and HIV incidence from 2006 to 2007 (P <0.0005).
Total CVL was also associated with newly diagnosed HIV cases (P =0.002)
and HIV incidence (P <0.0005).
Conclusions: Increased antiretroviral treatment
options and coverage, as well as increasing HIV status awareness, may have led
to decreases in CVL in San Francisco from 2002 to 2008 with subsequent decreases
in HIV diagnoses and, though limited by only 2 years of data, with fewer new HIV
infections. Findings support the hypothesis that wide-scale early ART can have
a preventive effect at a population-level. Because CVL is temporally upstream
of new HIV infections, public health departments should consider adding CVL to
routine HIV surveillance to track the epidemic and evaluate the effectiveness
of HIV prevention and treatment interventions.
Figure 1: Trends in Mean Community Viral Load with HIV
Incidence and New HIV diagnoses in San Francisco, 2002-2008