CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

62.9% (56.0-69.3) of the discordant couples in Ndhiwa and 62.1% (50.0-73.0) in Gutu, but about half of HIV-positive partners were men in Chiradzulu (48.8%; 41.3-56.4) and Nsanje (52.3%; 40.2-64.1). HIV status awareness among HIV- positive partners in discordant couples ranged from 42.4% in Ndhiwa, to 72.7% in Gutu. VL suppression ranged from 34.2% in Ndhiwa to 69.5% in Nsanje. VL suppression was similar between men and women in three settings, Ndhiwa (38.4% vs 27.8%, p=0.14), Nsanje (65.6% vs 74.1%, p=0.49) and Gutu (54.1% vs 66.7%, p=0.96) but lower in men than women in Chiradzulu (44.4% vs 62.7%, p=0.02). Conclusion: High rates of discordant couples with low status awareness among positive partners are one of the major gaps in this high risk group. The low rates of status awareness among HIV-positive partners must be addressed in order to promote timely initiation of ART and/or PREP to reduce transmission within this high-risk group. Nancy Puttkammer 1 , Tricia Rodriguez 1 , Ermane Robin 2 , Gracia Desforges 2 , Canada Parrish 1 , Jean Gabriel Balan 3 , Jean Guy Honore 3 , Joelle Deas Van Onacker 2 , Scott Barnhart 1 , Jane M. Simoni 1 , Kesner Francois 2 1 University of Washington, Seattle, WA, USA, 2 Ministry of Public Health and Population, Port-au-Prince, Haiti, 3 I-TECH, Petion-Ville Background: Haiti’s Ministry of Health recently endorsed a national initiative to lengthen prescribing intervals for HIV antiretroviral therapy (ART), known as multi-month scripting (MMS). With MMS, virally-suppressed patients on ART for >6 months are moved frommonthly prescribing intervals to intervals of 2-6 months. This decreases patient travel and clinic waiting time, and reduces congestion in ART clinics. Differentiated models of HIV care seek to optimize quality and efficiency of HIV services; however, few studies have described results of MMS in resource-limited settings. Methods: To describe the evolution of ART prescribing patterns in Haiti, we analyzed 867,449 ART prescription records from 65,460 patients in 82 health facilities from January 2012 to December 2016, drawn from the iSanté electronic medical record (EMR) system. We assessed the relationship between prescribing interval and being retained in care, defined as returning within 90 days of the next expected ART pick up date. The outcomes analysis used a subset of 45,604 ART patient records during 2015 – 2016. A multilevel logistic model was used to estimate the association between MMS and retention on ART, after adjustment for clinic site and for patient age, sex, baseline WHO stage, time on ART, and starting ART regimen. Results: By March 2016, MMS intervals of 36 – 70 days were most common in Haiti (see Figure). Among patients on ART for at least 6 months, MMS accounted for at least half of prescriptions in 81% of sites by December 2016 (increasing from 66% of sites 1 year earlier). Patients receiving MMS tended to be older, have been on ART longer, and have more advanced WHO stage at baseline. Retention was highest (80.8%) among patients with MMS intervals of 71-100 days, and lower (63.4%) among patients with intervals of 0-35 days. After adjustment, longer MMS intervals were positively associated with retention. Odds of retention were 2.3 times higher for intervals of 36-70 days (p<.001), and 2.6 times higher for intervals of 71+ days (p<.001), compared to intervals of 0 -35 days. Conclusion: Haiti has aggressively moved toward MMS across a majority of ART sites. The association between longer MMS intervals and improved retention on ART is promising, although these favorable results may reflect the preferential selection of stable patients for MMS, rather than a direct causal effect of the strategy. Nevertheless, the fact that no unintended negative relationship was observed between MMS and retention is important.

1086 MULTI-MONTH SCRIPTING (MMS) AND RETENTION ON HIV ANTIRETROVIRAL THERAPY IN HAITI

1087 ROUTINE RETENTION IN CARE AT HOWARD BROWN, 2012-2017: ARE QUARTERLY VISITS TOO MUCH? Laura Rusie 1 , Carlos Orengo 1 , Diane Burrell 1 , Magda Houlberg 1 , Arthi Ramachandran 2 , Kristin Keglovitz-Baker 1 , David Munar 1 , John A. Schneider 2 1 Howard Brown Health Center, Chicago, IL, USA, 2 University of Chicago, Chicago, IL, USA Background: PrEP retention in care is a critical yet understudied component of the PrEP care continuum. Most PrEP retention in care analyses come from placebo-controlled and open-label studies, but data are limited on real-world PrEP use. Methods: We conducted a retrospective cohort study from July 2012 through August 2017 of Howard Brown Health patients initiating PrEP in Chicago. We abstracted unique PrEP starts as well as other demographic, clinic visit and clinic history data and examined drivers of visit constancy. Multinomial regression models were built beginning with all associated predictors removing one variable at a time based on p-value. Results: Overall the cohort included 4787 participants who initiated PrEP from January 2012 through August of 2017 accumulating 4058 person-years of PrEP use (see Figure 1). Cohort members are young with over half being under 30 years of age, 36% Latinx or Black and 24% uninsured. There were approximately 178 monthly PrEP starts in 2017, 858 (62%) of clients have been on PrEP for 24 months or more, 1687 (35%) had an STI during follow-up, and 30% had two or more other clinical diagnoses at PrEP initiation. With respect to PrEP retention in care, among those with the opportunity for at least 12 months of follow-up, 42.9% remained in care at 12 months, yet only 15% had high visit constancy of 4 out of 4 quarters with a PrEP visit during the first 12 months of PrEP care. In final multinomial regression models, factors associated with at least 1/4 quarters with a PrEP visit included number of other comorbidities. In addition with increasing visit constancy of 3/4 quarters and 4/4 quarters with a PrEP visit, uninsured clients were less likely to be retained (aOR, 0.53; 95% CI, 0.34-0.84) and (aOR, 0.36; 95% CI, 0.21-0.62) respectively. Conclusion: While overall client engagement in PrEP was modest, adherence to CDC recommended guidelines for quarterly visits was low. Insurance status and prevalent co-morbidities were the main drivers of PrEP retention in care. It is unclear whether low rates of retention are due to need for longer gaps between scheduled visits for adherent patients, whether risk and perception of risk is dynamic or whether social and structural factors are impeding clinic visits. PrEP retention in care interventions are needed to realize the full potential of PrEP in the context of HIV elimination in the United States.

Poster Abstracts

CROI 2018 417