CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

VL less than 1000 copies/mL. No significant difference in viral suppression at their last visit was detected between patients with baseline CD4 count less than 50 cells/μL and those with baseline CD4 counts of 50-99.9, 100-199.9, 200-349.9, or ≥350 cells/μL. Conclusion: These findings showed that patients with very advanced immunodeficiency have equivalent capacity for immunologic recovery and viral suppression compared to those with higher baseline CD4 cell counts, if they are retained in care.

particularly for newly-diagnosed cases, are needed to increase initiation of HIV care.

1035 HIV TREATMENT CASCADE AMONG MEN WHO HAVE SEX WITH MEN IN KIGALI, RWANDA Jean Olivier Twahirwa Rwema 1 , Benjamin Liestman 1 , Julien Nyombayire 2 , Sosthenes Ketende 1 , Amelia Mazzei 2 , Carrie E. Lyons 1 , Oluwasolape M. Olawore 1 , Mugwaneza Placidie 3 , Stefan Baral 1 , Patrick S. Sullivan 2 , Susan Allen 2 , Etienne Karita 2 1 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 2 Emory University, Atlanta, GA, USA, 3 Rwanda Biomedical Centre, Kigali, Rwanda Background: Men who have sex with men (MSM) have high HIV acquisition and transmission risk globally and are defined as a key population in the Rwanda national strategic plan. However, there are no published HIV epidemiological data among MSM in Rwanda. In this study, we characterize MSM engagement in HIV treatment cascade in Kigali, Rwanda. Methods: MSM> 18 years were recruited in a cross-sectional behavioral and biological survey using respondent driven sampling (RDS) between March – July 2018 in Kigali, Rwanda. Data on socio-demographic characteristics, sexual behavior and engagement in HIV services were collected using an interviewer-administered structured questionnaire. HIV infection and viral load were biologically assessed. We used a cascade framework to characterize engagement in HIV care continuum. Results: Overall, 736 eligible MSM were recruited in the study. The median age was 27 [range:18-68]. The HIV prevalence was 10.1% (74/736) [RDS adjusted prevalence: 9.2%; 95% CI: (6.4-12.1)]. Of the participants found to be living with HIV, only 61% (45/74) reported that they knew their HIV status before enrollment. Higher age (> 35 years) was significantly associated with both HIV positive status (p < 0.01) and knowing HIV diagnosis prior to enrollment (p< 0.05). Of MSM who knew their HIV positive status, 98% (44/45) reported to be on ART and 75% (33/44) were virally suppressed. Overall, we estimated that among the total population of MSM living with HIV in Kigali, 61% know their status, 59% are on ART and 59 % are virally suppressed. This represents gaps of 29%, 22% and 14% respectively to reach the 90-90-90 target in the MSM population in Kigali Conclusion: Taken together, these data demonstrate high HIV prevalence with suboptimal engagement in HIV treatment services among particularly young MSM in Rwanda. A quarter of those reporting ART were viremic suggesting the need for improved retention and adherence programing in addition to screening for HIV-drug resistance. Given the challenges in addressing the needs of young MSM, interventions leveraging emerging technologies and social media in addressing engagement and retention may be particularly effective in Rwanda. 1036 LONGITUDINAL HIV CARE TRAJECTORIES IN THE CNICS COHORT: A RETROSPECTIVE COHORT STUDY Kimberly A. Powers 1 , W. C. Mathews 2 , Kenneth H. Mayer 3 , Ellen F. Eaton 4 , Elvin Geng 5 , Richard D. Moore 6 , Michael J. Mugavero 4 , Joseph J. Eron 1 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA, 2 University of California San Diego, San Diego, CA, USA, 3 Fenway Health, Boston, MA, USA, 4 University of Alabama at Birmingham, Birmingham, AL, USA, 5 University of California San Francisco, San Francisco, CA, USA, 6 Johns Hopkins University, Baltimore, MD, USA Background: Long-term HIV care engagement is required for optimal clinical and prevention outcomes, but longitudinal patterns of HIV care attendance are poorly understood. Identification of distinct longitudinal trajectories of

Poster Abstracts

1034 PREVALENCE AND CORRELATES OF NONENROLLMENT IN HIV CARE, CHÓKWÈ DISTRICT, MOZAMBIQUE Kristen Heitzinger 1 , Anne F. McIntyre 2 , Isabelle Casavant 1 , Noela Chicuecue 3 , Victor Chivurre 4 , Aleny Couto 3 , Keydra Oladapo 1 1 CDC Mozambique, Maputo, Mozambique, 2 CDC, Atlanta, GA, USA, 3 Ministry of Health, Maputo, Mozambique, 4 Gaza Provincial Directorate of Health, Xai-Xai, Mozambique Background: HIV care and treatment is expanding in Mozambique and implementation of a national ‘Test and Start’ treatment program began in 2016. However, non-enrollment remains a barrier to epidemic control. To inform interventions to increase enrollment, we evaluated the prevalence and sociodemographic and behavioral correlates of non-enrollment in Chókwè District, Mozambique. Methods: Data were sourced from a cross-sectional survey conducted annually in Chókwè District during 2014–2017, with participants aged 15–59 identified via a household demographic surveillance system (HDSS). We analyzed data from participants who reported ever having received a positive HIV test. If surveyed in multiple years, data from first survey were analyzed. The 2013 HDSS census estimates were used to weight by age, sex, and urban residence distributions. We calculated the prevalence of and reasons for non-enrollment. Logistic regression was used to calculate adjusted odds ratios (aOR) and 95% confidence intervals (CI) of the association between sociodemographic and behavioral variables and non-enrollment in HIV care, adjusting for intra- household sampling. Results: Of 2,654 participants who reported ever having received a positive HIV test, 127 (5.3%) had not enrolled in HIV care. There was no difference in non-enrollment after district-wide implementation of Test and Start. Most frequently cited reasons for non-enrollment were did not need care due to good health status (33/126; 27.1%) and did not believe they had HIV (7/126; 7.6%). Participants who first tested positive since 2013 and those who received their first positive test <2 years had increased odds of non-enrollment (p=0.02 and p=0.002, respectively). Compared to testing at a district healthcare facility, testing positive at home had increased odds of non-enrollment (aOR: 3.92, 95% CI: 2.39, 6.45). HIV status nondisclosure was associated with non-enrollment (aOR: 6.15, 95% CI: 3.79, 10.00). Among participants who first tested positive ≤1 year , those who did not meet with someone to help them enroll in care had increased odds of non-enrollment (aOR: 4.60, 95% CI: 1.94, 10.93). Conclusion: Obstacles to enrollment reflect the importance of accurate health messaging, strong social support, and prompt clinical linkage to care, regardless where HIV testing occurs. Enhanced patient advocacy and case management,

CROI 2019 406