Background: Underreporting of prior HIV diagnosis and antiretroviral therapy (ART) use based on self-report is well-documented in national surveys. Antiretroviral (ARV) testing has been used to improve survey estimates, by reclassifying respondents with ARVs detected in blood as previously-diagnosed and on ART. Viral load testing is more affordable and more routinely available than ARV testing. Undetectable viral load (UVL) is a potential proxy for ARV use but may be confounded by elite controllers who, while believed to be rare, have UVL in absence of ART.
We examined impact of adjusting the Kenya HIV cascade using UVL with and without ARV detection.
Methods: The 2012 Kenya AIDS Indicator Survey (KAIS) included questions on HIV diagnosis and ARV use, and collected dried blood spots for centralized viral load and ARV testing. We defined UVL as viral load < 550 copies/milliliter; ARVs were present if efavirenz, nevirapine, lopinavir or lamivudine were detected in blood. We reclassified participants as previously-diagnosed and on ART if either ARVs detected or viral load was undetectable. We compare self-reported prior diagnosis, and ART coverage among previously-diagnosed, to indicators adjusted for ARV detection, UVL, or both, among respondents aged 15-64 years. Indicators were weighted to account for the complex survey design.
Results: Among 235 of 648 HIV-infected respondents with UVL, self-reported status was: 65 undiagnosed (27.7%), 25 previously-diagnosed but not on ART (10.6%), and 145 currently on ART (61.7%). Prior diagnosis increased from 46.9% for self-report to 56.2% (95% confidence interval [CI] 50.7-61.7) when ARV-adjusted, 57.5% (95% CI 52.0-63.1) when UVL-adjusted, and 59.8% (95% CI 54.3-65.3) for ARV-UVL-adjusted. Treatment coverage among those previously-diagnosed increased from 67.9% to 76.2% (95% CI 70.8-81.5) when ARV-adjusted, 80.2% (95% CI 75.7-84.8) when UVL-adjusted and 81.7% (95% CI 77.3-86.1) when adjusted for both markers. Sensitivity and specificity of UVL-adjusted prior diagnosis were 95.8% and 91.3%, and of UVL-adjusted ART use were 93.0% and 88.8% respectively, versus ARV-adjusted self-report.
Conclusions: In this survey, UVL-adjusted point estimates were similar to, but slightly greater than ARV-adjusted estimates of prior HIV diagnosis and ART coverage. Viral load may be useful for adjusting indicators of prior HIV diagnosis and treatment in surveys.

Download the e-Poster