Undisclosed HIV infection among MSM in a behavioral surveillance study.

Objective: As a proxy for undiagnosed HIV, the Centers for Disease Control and Prevention's National HIV Behavioral Surveillance (NHBS) monitors participants who report being unaware of their infection, defined as self-reporting an HIV-negative or unknown status during the interview but testing positive for HIV infection. We validated the NHBS measure of awareness among MSM in 2014.

Design: We tested dried blood spots from MSM who reported being unaware of their infection for seven antiretrovirals (ARVs). MSM unaware with at least one ARV detected were defined as misreporters.

Methods: Weighted percentages and 95% confidence intervals were calculated to compare characteristics among misreporters, nonmisreporters, and those who self-reported as HIV-positive. Viral load was quantified with a validated assay using dried blood spots.

Results: Of 1818 HIV-positive MSM, 299 (16%) self-reported as HIV-negative or unknown infection status. Of these 299, 145 (49%) were considered misreporters based on ARV detection. Among the unaware, misreporters were more likely than nonmisreporters to be older and have health insurance. Compared with self-reported HIV-positive MSM, misreporters were more likely to be black, be bisexual, and have perceived discrimination. Of 138 misreporters with viral load data, 116 (84%) had an undetectable viral load.

Conclusion: ARV testing revealed that half of MSM classified as unaware of their infection misreported their status. Although off-label preexposure prophylaxis use might explain the presence of ARVs, it is unlikely as many misreporters were virally suppressed, suggesting they were on HIV therapy. Biomarker validation of behavioral data can improve data quality and usefulness in NHBS and other studies.