CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Conclusions: Older HIV+ adults in a high HIV prevalence community in rural South Africa report sexual behaviors that are consistent with HIV transmission; similarly, older HIV- adults report sexual behaviors that are consistent with HIV acquisition. HIV prevention inititatives that target older adults are urgently needed to reduce both HIV transmission and acquisition in this and similar communities in sub-Saharan Africa.

906 Factors AssociatedWith Misreporting HIV Status Among MSM From Baltimore Danielle German 1 ; Kate Shearer 1 ; Ju Nyeung Park 1 ; Colin Flynn 2 ; Carl Latkin 3 ; Oliver Laeyendecker 4 ;Thomas C. Quinn 5 ;William Clarke 5 1 Johns Hopkins Bloomberg Sch of PH, Baltimore, MD, USA; 2 Maryland DHMH, Baltimore, MD, USA; 3 Johns Hopkins Univ, Baltimore, MD, USA; 4 NIH, Bethesda, MD, USA; 5 Johns Hopkins Univ Sch of Med, Baltimore, MD, USA Background: Unrecognized HIV infection and antiretroviral (ARV) drug use are critical indicators for monitoring the HIV epidemic, but frequently rely on self-report. We tested for presence of ARVs in HIV+men who have sex with men (MSM) and analyzed socio-demographic factors associated with misreporting their HIV status. Methods: From the 2008 MSM cycle of National HIV Behavioral Surveillance in Baltimore, sera from 147 HIV+MSM were tested for the presence of ARVs using liquid chromatography-tandemmass spectrometry (API 4000 Triple Quadruple Mass Analyzer) which detects 20 antiretroviral (ARV) drugs. Factors associated with misreported unrecognized HIV infection were analyzed. Results: The prevalence of unrecognized infection, calculated as those with no self-reported knowledge of prior HIV infection divided by the total that tested HIV positive, was 74%. No error in ARV reporting was detected among self-reported HIV-positive participants. Of 109 participants originally classified with unrecognized HIV infection, 33% (36/109) had at least one ARV drug detected and 31% (34/109) had detectable regimens consistent with ARV therapy. Compared to those with no ARV detected, misreporters were significantly older (mean 38.4 vs. 29.6, p<0.0001) and less likely to report binge drinking (31% vs. 55%, p=0.017). No socio-economic differences were significant in multivariate models. Compared to those who self-reported HIV positive status and ARV use, misreporters were significantly more likely to be unemployed (39% vs. 9%, p=0.032) and marginally more likely to be recently homeless (33% vs. 9%, p=0.058) and recently arrested (17% vs. 0%, p=0.073). No differences in sexual or drug use behaviors were observed. Recalculated prevalence of unrecognized HIV infection among MSMwith HIV was 49.7%; 27.8% among NH white (n=18) and 51.4% among NH black participants (n=111), with decreasing prevalence by age. Conclusions: These analyses provide new insight into the extent of under-reporting of known HIV infection and ARV status in behavioral surveillance. The high rate of unrecognized HIV infection among Baltimore MSM, with nearly half of HIV+ individuals who knew their status misreporting their status, is alarming. Further qualitative research may help to understand and contextualize misreported HIV status in future behavioral surveys. 907 Validation of Self-Reported HIV Status Among Older Adults in Rural South Africa Julia K. Rohr 1 ; Molly S. Rosenberg 2 ; Xavier Gomez-Olive 3 ; RyanWagner 3 ; Brian Houle 4 ; Joshua Salomon 1 ; StephenTollman 5 ; Kathleen Kahn 3 ;Till Bärnighausen 2 1 Harvard Univ Sch of PH, Boston, MA, USA; 2 Harvard Sch of PH, Boston, MA, USA; 3 Med Rsr Council/Wits Rural PH and Health Transition Unit, Johannesburg, South Africa; 4 Australian Natl Univ, Canberra, Australia; 5 Univ of the Witwatersrand, Johannesburg, South Africa Background: Little is known about willingness to disclose HIV status among the growing number of HIV positive older adults in South Africa. Estimates of sensitivity and specificity of self-reported status in this population are unavailable but important for researchers and policy makers. We validate self-reported HIV status and explore factors associated with accuracy of self-report. Methods: We analyzed data from the Health and Aging in Africa: Longitudinal Studies of INDEPTH Communities (HAALSI) baseline survey, an observational cohort of randomly sampled adults age 40+ years in a poor, rural community in Agincourt, South Africa. Self-reported HIV status and prior HIV testing was obtained through structured interviews by local field workers. Dried blood spots (DBS) were collected at time of interview and HIV enzyme-linked immunosorbent assays for HIV status were conducted. We calculated sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for self-reported HIV status compared to “gold standard” DBS results, and stratified findings by demographic characteristics. Logistic regression explored associations between demographic characteristics and sensitivity of self-report. Results: 3,426 individuals (94.4% of respondents) had DBS results available and were included in analysis. HIV prevalence was 21.8% from DBS and 12.5% from self-report. Sensitivity of self-report was 53% (95% CI: 49%-57%), specificity 99% (95% CI: 99%-99%), PPV 94% (95% CI: 91%-96%) and NPV 88% (95% CI: 87%-89%). Among those who reported knowing their HIV status, sensitivity increased to 67% (Table 1). The lowest sensitivity was found among illiterate participants (49%) and oldest ages (33% among age 80+ years). Correct report of being HIV positive was more likely among participants 50-59 years old compared to 70+ years old (OR = 1.69, 95% CI: 1.04-2.75) and literate

Poster Abstracts

381

CROI 2016