CROI 2020 Abstract eBook

Abstract eBook

Oral Abstracts

arms, controlling for potential imbalances between arms and specifying for clustering. Results: A total of 5336 participants were enrolled, 3MF (1898), 3MC (1558) and 6MCD (1880) arms. Retention in ART care was not different across the arms and achieved the noninferiority limit (-3.25%) with 3MC vs. 3MF 6MCD vs. 3MF (control) and 6MCD vs. 3MC, adjusted RD= -0.1% (95% CI: -1.6% to 1.5%), adjusted RD= -1.3% (95% CI: -3.0% to 0.5%), and adjusted RD= -1.2% (95% CI: -2.9% to 0.5%), respectively. Retention in the intervention arms for both 3MC and 6MCD arms did not differ vs. 3MF, adjusted RD=1.1% (95% CI: -0.6% to 2.8%) and adjusted RD= -0.6% (95% CI: -2.4% to 1.1%), respectively. However, there was a slight reduction in 6MCD vs 3MC, adjusted RD= -1.9% (95% CI: -3.6% to -0.2%). Amongst 1503, 1126 and 1285 participants with available viral load results after 12 months, 1482 (98.6%), 1104 (98.1%) and 1263 (98.3%) were virally suppressed in arms 3MF, 3MC and 6MCD, respectively. There were no differences in viral suppression between 3MC, or 6MCD vs. control, risk ratio (RR)=1.00 (95% CI: 0.98-1.01) and RR=1.00 (95% CI: 0.98-1.01), respectively. Conclusion: There is no difference in retention in care or viral load suppression for stable patients receiving 3 or 6 month dispensing of ART within community- based differentiated models of care when compared to the standard 3 month facility dispensing model 44 AN MHEALTH CHW INTERVENTION TRIAL IN AN HIV HYPERENDEMIC COMMUNITY IN RAKAI, UGANDA Larry W. Chang 1 , Ismail Mbabali 2 , Xiangrong Kong 1 , K. Rivet Amico 3 , Caitlin E. Kennedy 1 , Fred Nalugoda 2 , Aggrey Anok 2 , David Serwadda 2 , Joseph Ssekasanvu 1 , Thomas C. Quinn 4 , Steven J. Reynolds 4 , Ronald H. Gray 1 , Maria Wawer 1 , Heidi Hutton 1 , Gertrude Nakigozi 2 1 Johns Hopkins University, Baltimore, MD, USA, 2 Rakai Health Sciences Program, Kalisizo, Uganda, 3 University of Michigan, Ann Arbor, MI, USA, 4NIAID, Bethesda, MD, USA Background: Effective strategies are needed to increase engagement in HIV services in HIV hyperendemic settings. We conducted a cluster-randomized trial in a fishing community on Lake Victoria (HIV prevalence ~41%) in Rakai, Uganda to assess the impact of a community health worker intervention called “Health Scouts” which used motivational interviewing strategies, a situated Information, Motivation, and Behavioral Skills framework, and mobile health (mHealth) counseling support tools to promote engagement in HIV treatment and prevention services. Methods: From September 2015 to December 2018, the Health Scout intervention was deployed in the community which had been divided into 40 contiguous, similarly populated clusters (20 intervention; 20 control). Community-wide surveys of consenting 15-49 year-old residents with HIV viral load testing of HIV-positive participants were conducted at mid-study (~15 months) and end-of-study (~39 months) to assess self-reported antiretroviral therapy (ART) and male circumcision coverage and HIV viral load suppression (defined as <400 copies/mL). The primary analytic method was an as-treated analysis using generalized estimating equations models including participants from both surveys in a pragmatic analysis due to high participant mobility and contamination by study arm. Results: 2522 and 1891 community residents completed the mid-study and end-of-study surveys respectively. By end-of-study, 95.7% (1789/1891) of residents reported awareness of the Health Scouts; 31% (580/1891) of residents reported having been visited and counseled by a Health Scout (i.e. exposed); 2.2% (41/1311) reported being approached but refusing to be seen. Health Scout exposure was higher in intervention (38%) compared to control clusters (23%), among those living with HIV (39%) compared to those who were not (23%), and among women (32%) compared to men (26%). As shown in Table 1, residents who reported having received the intervention (exposed) were more likely to report being on ART and to be virologically suppressed compared to residents who reported not having received the intervention (unexposed); however, there were no differences in male circumcision coverage. Conclusion: A novel community health worker intervention using motivational interviewing techniques and mHealth tools was associated with improved ART coverage and HIV virologic suppression. However, intervention uptake varied by subgroups and cluster contamination was substantial. This intervention may be a useful community-based component of a comprehensive HIV response.

45 IMPROVED TIME IN CARE AND VIRAL SUPPRESSION WITH STREAMLINED CARE IN THE SEARCH STUDY Matthew D. Hickey 1 , James Ayieko 2 , Dalsone Kwarisiima 3 , Fredrick J. Opel 2 , Asiphas Owaraganise 3 , Laura B. Balzer 4 , Gabriel Chamie 1 , Vivek Jain 1 , James Peng 1 , Edwin D. Charlebois 1 , Craig R. Cohen 1 , Elizabeth A. Bukusi 2 , Moses R. Kamya 5 , Maya L. Petersen 6 , Diane V. Havlir 1 1 University of California San Francisco, San Francisco, CA, USA, 2 KEMRI-UCSF, Kisumu, Kenya, 3 Infectious Diseases Research Collaboration, Kampala, Uganda, 4 University of Massachusetts Amherst, Amherst, MA, USA, 5 Makerere University, Kampala, Uganda, 6 University of California Berkeley, Berkeley, CA, USA Background: HIV differentiated service delivery (DSD) models are being scaled up in resource-limited settings for stable patients; less is known about DSD outcomes for patients newly linked or re-linked to care. We evaluated the effect of the SEARCH streamlined care intervention by comparing care engagement and viral suppression (VS) between intervention and control arms among HIV+ persons ART eligible by country guidelines at study start who were already enrolled or who linked to care after universal HIV testing in the SEARCH trial (NCT:01864603). Methods: Our analysis included HIV+ adults (age ≥15 yrs) at baseline (2013) who were country guideline ART eligible (prior ART experience or CD4≤350) and had ≥1 clinic visit for HIV care between 2013-2017 in SEARCH communities randomized to intervention (N=16) or control (N=16). We assessed the effect of streamlined care (patient-centered care, increased appointment spacing, improved clinic access, reminders, and tracking) on time in care (TIC) and viral suppression (VS) at 3 years. TIC was defined as the proportion of total follow up time that patients adhered to visit schedules. Analysis was stratified by baseline care status, namely: 1) ART-experienced with baseline VS, 2) ART-experienced with baseline viremia, or 3) ART-naïve with baseline CD4≤350. Comparisons between study arms used cluster-level TMLE. Results: Among 4,391 HIV+ persons (35%men, 8% youth 15-24 yrs) in care and eligible for ART by country guidelines, 2,958 (67%) were ART-experienced with baseline VS, 568 (13%) were ART-experienced with baseline viremia, and 865 (20%) were ART-naïve with CD4≤350. Among ART-experienced patients with baseline viremia, streamlined care was associated with both higher TIC (RR 1.11, 95% CI 1.01-1.21) and VS (67% vs 47%, RR 1.41, 95% CI 1.04-1.92). Among ART- naïve persons, streamlined care was associated with higher TIC (RR 1.10, 95% CI 1.05-1.21) but VS was not significantly higher (83% vs 78%, RR 1.06, 95% CI 0.95 -1.19). Among ART-experienced persons with baseline VS, effects of streamlined care were observed on TIC (RR 1.07, 95% CI 1.01-1.13), although nearly all were virally suppressed after 3 years regardless of the care delivery model (97% intervention vs 95% control, RR 1.02, 95% CI 1.00-1.03). Conclusion: Streamlined care was associated with better engagement in care for all groups and viral suppression for ART-experienced patients with viremia in this randomized comparison of patients ART eligible at study start who linked to care after universal HIV testing.

Oral Abstracts

46 COLLABORATIVE DATA-TO-CARE MODEL IMPROVES HIV CARE OUTCOMES IN PLWH IN PHILADELPHIA Sindhu Shamasunder 1 , Crystal Lucas 1 , Shedane Shaw 1 , Briana Gibson 1 , Olivia Kirby 1 , Melissa Miller 1 , Kathleen A. Brady 1 1 Philadelphia Department of Public Health, Philadelphia, PA, USA

16

CROI 2020