CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Background: A prior cohort study from Kenya demonstrated reduced effectiveness of contraceptive implants when used in combination with efavirenz-containing antiretroviral therapy (ART). To further validate this finding, we conducted two-phase random sampling with an expanded cohort of women living with HIV (WLHIV) using data from the East Africa International Epidemiology Databases to Evaluate AIDS. Methods: We conducted a random sampling study of WLHIV, from 15 to 45 years of age enrolled in HIV care in western Kenya between January 2011 and December 2015, to validate the exposure of a combination of contraceptive method and ART regimen and primary outcome of incident pregnancy. We generated a cohort of WLHIV utilizing electronic medical records, and then conducted detailed file reviews for a stratified random subset of the women. We used multivariate Poisson models to compare pregnancy rates among women using different contraceptive and ART combinations, accounting for the second phase sampling with generalized raking inverse probability weighted methods. Lastly, we conducted phone interviews with a further subset of women sampled for the file reviews to compare self-reports against medical records. Results: 85,819 women contributed 172,378 person-years (p-y) to this analysis. We conducted file reviews for 4,987 women (contributing 16,991 p-y) and phone interviews for 1,275 women (contributing 5,775 p-y). Based on data from the file review, among women using implants in the overall cohort, the pregnancy incidence was 1.1 and 3.3 per 100 p-y for nevirapine- and efavirenz- containing ART users, respectively (incidence rate ratio [IRR] 3.1, 95% CI 2.1-4.5; Table). Among the subset of women using implants with whomwe conducted phone interviews, the pregnancy incidence was 2.4 and 9.0 per 100 p-y for nevirapine- and efavirenz-containing ART users, respectively (IRR 3.8, 95% CI 2.0-7.2). Conclusion: Using probabilistic subsampling, we confirm the prior finding that contraceptive implant effectiveness is reduced with concomitant efavirenz use. Dolutegravir-containing ART, which is not anticipated to reduce implant effectiveness, should be considered for WLHIV already using or interested in contraceptive implants. Self-reports largely corroborated medical records, though the higher rates may be due to recall bias. Our robust and novel validation methodology also highlights a way forward for other studies conducted with electronic medical records.

aged ≥16 years were interviewed pre and post-intervention about their fertility desires and FP use, and clinical data was abstracted from their medical charts. Differences between pre and post-intervention participants were tested using Pearson Chi-square tests. Unadjusted and adjusted logistic regression models were used to examine differences in self-reported FP uptake between the two time periods. Results: A total of 629 WLHIV were interviewed pre-intervention and 684 post-intervention. During the pre-intervention period, only 38% of women not desiring a pregnancy reported currently using an effective FP method compared to 49% post-intervention (p=.003, Table 1). Uptake by method at the two time points was: pills (10% vs. 8%, p>.05), injectables (15% vs. 25%, p<.0001), implants (5% vs. 8%, p>.05), and intrauterine devices (IUDs, 1% vs. 1%, p>.05). The percent of women reporting dual method use increased from 9% to 18% (P=.0003); while, unmet need for FP decreased from 59% to 46% (P=.0003). Among women wanting to get pregnant, receipt of safer pregnancy counseling increased from 27% to 39%. The total intervention cost was estimated at $83,293 (2018 USD) over the 12-month period including labor (40%), supplies (26%), training (14%) and administration (20%). model of FP/HIV integration was associated with a significant increase in the number of WLHIV reporting use of an effective FP method and a met need for FP. These results support continued efforts to integrate FP and HIV services to improve women’s access to these services.

Poster Abstracts

1059 A COMBINED ESTROGEN/PROGESTIN VAGINAL RING IMPROVES VAGINAL MICROBIAL COMMUNITIES Nicole H. Tobin 1 , Sarah L. Brooker 1 , Fan Li 1 , Yoninah S. Cramer 2 , Susan L. Rosenkranz 2 , Grace M. Aldrovandi 1 , Robert Coombs 3 , Susan E. Cohn 4 , Carmen D. Zorrilla 5 , Laura E. Moran 6 , Baiba Berzins 4 , Kimberly K. Scarsi 7 , Catherine Godfrey 8 , for the ACTG 5316 Team 1 University of California Los Angeles, Los Angeles, CA, USA, 2 Harvard T.H. Chan School of Public Health, Boston, MA, USA, 3 University of Washington, Seattle, WA, USA, 4 Northwestern University, Chicago, IL, USA, 5 University of Puerto Rico, San Juan, Puerto Rico, 6 Social & Scientific Systems, Silver Spring, MD, USA, 7 University of Nebraska Medical Center, Omaha, NE, USA, 8 DAIDS, NIAID, Bethesda, MD, USA Background: ACTG study A5316 found that during contraceptive intravaginal ring (IVR) use over 3 weeks, efavirenz-based ART (EFV) significantly decreased both ethinyl estradiol (EE) and etonogestrel (ENG) plasma exposure, while atazanavir/ritonavir-based ART (ATV) decreased EE, yet increased ENG. We explored the role of the IVR on vaginal microbial communities and vaginal small chain fatty acids (SCFA) as well as the role of the vaginal microbes/SCFA on hormone concentrations. Methods: Of the 74 participants (25 ART Naïve; 25 EFV, 24 ATV), 71 had 16S rRNA sequencing of the V4 region on vaginal swabs at weeks 0 (pre-IVR insertion), 1, 2, 3, and 4 (1 week post-IVR removal); and 73 had vaginal aspirate SCFAs measured by Metabolon® at weeks 0, 1 or 2, and 4. Sequences were filtered and taxa assigned using DADA2, species using SPINGO with SILVA database, and Lactobacillus using BLAST. Negative binomial and linear regression models identified differentially abundant microbiome and SCFA features, respectively. Spearman correlation assessed relationships between microbiome relative abundance and weekly EE/ENG concentrations. Results: At baseline, microbial communities of participants could be robustly classified as L. crispatus¬-dominant (Community State Type (CST) I, n=8), L. gasseri-dominant (CST II, n=2), L. iners-dominant (CST III, n=20), or mixed anaerobic communities (CST IV, n=41). Start of IVR therapy was associated with an increased probability of transition into Lactobacillus-dominant community types (OR=3.39, CI=0.36-32.15), Fisher's exact test, p<0.001). ENG levels were negatively correlated with abundance of Prevotella timonensis.

1058 AN EVALUATION OF AN ENHANCED MODEL OF FP/HIV SERVICE INTEGRATION IN LUSAKA, ZAMBIA

Amy M. Medley 1 , Fatima Tsiouris 2 , Sherri Pals 1 , Brenda Senyana 2 , Susan Hanene 3 , Shadrick Kayeye 3 , Rocio R. Casquete 2 , Arielle Lasry 1 , Mollie Braaten 2 , Meagan Cain 1 , Tiffiany Michelle Aholou 1 , Prisca Kasonde 3 , Tina Chisenga 4 , Keith Mweebo 5 , Tiffany Harris 2 1 CDC, Atlanta, GA, USA, 2 ICAP at Columbia University, New York, NY, USA, 3 ICAP at Columbia University, Lusaka, Zambia, 4 Government of Zambia Ministry of Health, Lusaka, Zambia, 5 CDC Zambia, Lusaka, Zambia Background: Women living with HIV (WLHIV) in sub-Saharan Africa continue to experience high rates of unplanned pregnancies. Ready access to family planning (FP) within HIV treatment programs allows women to make informed fertility choices. We implemented an enhanced model of integrating FP and HIV services at 6 health facilities in Lusaka, Zambia aimed at increasing contraceptive uptake among WLHIV wanting to avoid pregnancy and to improve safer conception counseling for those desiring a pregnancy. Methods: The model included: training HIV clinic staff in FP service delivery; offering a full range of FP methods within the HIV clinic; improving FP documentation within HIV monitoring systems; and introducing facilitated referral to community-based distributors to support FP use between HIV treatment visits. For the evaluation, systematic, independent samples of WLHIV

CROI 2020 398