CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

1006 INCIDENCE AND CORRELATES OF UNINTENDED PREGNANCY IN HIV- POSITIVE KENYAN SEX WORKERS Erica Lokken 1 , Barbra A. Richardson 1 , George Wanje 2 , Katherine Wilson 1 , Walter Jaoko 2 , John Kinuthia 3 , R. Scott McClelland 1 1 University of Washington, Seattle, WA, USA, 2 University of Nairobi, Nairobi, Kenya, 3 Kenyatta National Hospital, Nairobi, Kenya Background: HIV-positive female sex workers (FSWs) often have high rates of unmet contraceptive need, but data on the incidence of planned, mistimed, and unwanted pregnancies are sparse. We examined incidence and correlates of pregnancy in HIV-positive Kenyan FSWs. Methods: Non-pregnant FSWs enrolled in a cohort study in Mombasa, Kenya were eligible. Participants returned for monthly visits to ascertain sexual risk behavior, and were pregnancy tested quarterly. Pregnancies were considered planned, mistimed, or unwanted according to quarterly fertility desire and pregnancy intention questions. Cox proportional hazards models were used to estimate hazard ratios (HR) for the association between demographic, sexual, clinical, and behavioral characteristics and pregnancy. Correlates associated with pregnancy in univariate analysis at p<0.10 were included in the adjusted model. Results: A total of 316 FSWs contributed 785.7 person-years of follow-up. Most women had a current/regular partner in the last 3 months (50.8%, 160/315), were not using modern non-barrier contraception (69.3%, 219/316), did not desire a child (69.9%, 221/316), and had CD4 counts >350 cells/mm3 (69.0%, 218/316). There were 46 first incident pregnancies, for a rate of 5.9/100 person-years (p-y; 95%CI 4.4-7.8). The incidences of planned (6.4/100 p-y, n=4), mistimed (10.1/100 p-y, n=12), and unwanted pregnancies (5.0/100 p-y, n=30) were similar (p=0.11), but 90% (n=42) of pregnancies were mistimed or unwanted. In univariate analysis, oral contraceptive pill (OCP) use (vs no contraception or condoms only), condomless sex, vaginal washing, transactional sex in the last year, having a regular partner in the last 3 months, and experiencing intimate partner violence in the last year were significantly associated with a higher pregnancy risk. Being ≥35 years old (vs <25) was associated with a lower pregnancy rate. In multivariable analysis, OCP use (aHR 2.92, 95%CI 1.09-7.80), reporting condomless sex (aHR 2.19, 95%CI 1.08, 4.46), and having a current/regular partner in the last 3 months (aHR 3.64, 95%CI 1.00- 13.34) were associated with increased risk of incident pregnancy. Conclusion: In this cohort of HIV-positive FSWs, 90% of pregnancies were unintended. As part of comprehensive HIV care for FSWs, identifying women’s fertility desire and pregnancy intention could facilitate efforts to increase effective contraceptive use in women not trying to conceive and to implement safer conception strategies for women trying to have a child. 1007 DRIVERS OF UNPLANNED PREGNANCY AND UNMET NEED FOR CONTRACEPTION IN SOUTH AFRICA Witness Chirinda 1 , Debra Jackson 2 , Vundli Ramokolo 1 , Yages Singh 3 , Trisha Ramraj 3 , Duduzile Nsibande 3 , Kondwani Ng’oma 4 , Sanjana Bhardwaj 2 , Mireille Cheyip 5 , Mary Mogashoa 5 , Carl Lombard 1 , Ameena Goga 1 1 South African Medical Research Council, Cape Town, South Africa, 2 UNICEF, New York, NY, USA, 3 South African Medical Research Council, Durban, South Africa, 4 UNICEF Rwanda, Kigali, Rwanda, 5 CDC, Atlanta, GA, USA Background: Preventing unplanned pregnancies amongst HIV positive women is a pillar of the WHO prevention of mother to child transmission of HIV (PMTCT) strategy, yet 60% pregnancies in South Africa are unplanned. We sought to identify predictors of unplanned pregnancies and unmet contraceptive among postpartumwomen in South Africa. Methods: This analysis involves data from a nationally representative, cross- sectional survey measuring PMTCT effectiveness, in 2012/13. A total of 9277 women with known HIV status were included. All data regarding pregnancy planning and contraceptive use were self-reported during interviews. Unmet need for contraception was defined as unintended pregnancy among women not using any contraceptive method. All analyses were weighted and accounted for the survey design. Multivariable logistic regression models were used to estimate factors associated with unplanned pregnancy and unmet need for contraception. Results: The mean age of participants was 26.3 years (SD 6.35), with 31.7% (95%CI: 30.6-32.7) self-reported HIV prevalence. More than a third (35.5%) were unaware of their HIV-positive status before pregnancy. A total of 5524 (61.0%) reported that their pregnancies were unplanned; of these 3868 (70%) reported non-use of contraceptives before pregnancy. Women who were unaware of

1005 PARTNER NOTIFICATION: INCREASING EFFECTIVENESS WITH MODERN COMMUNICATION TECHNOLOGY Amelia S. Oliveira 1 , Julia Bilinska 1 , Hamish Mohammed 2 , Jay Jarman 1 , John Were 2 , Anatole Menon-Johansson 1 , Lisa Hamzah 1 1 Guy’s and St Thomas’ NHS Foundation Trust, London, UK, 2 Public Health England, London, UK Background: Prevalence of sexually transmitted infections (STIs) in STI contacts are high. Partner notification (PN) aims to inform and treat partners and reduce onward transmission. UK standards recommend 0.6 partners tested per index case (0.4 in large conurbations), however PN is time and labor intensive. Online platforms may reduce costs, expand coverage and increase efficiency although data remains limited. SXT is an electronic PN tool using interactive digital contact slips; this study aimed to assess effectiveness with number of contacts tested per index case compared to national data and examine factors associated with successful PN. Methods: A retrospective analysis of PN initiated via SXT in the UK between 01/12/17-31/07/18 was performed using anonymized data on index case demographics, STIs and PN. Number of contacts screened per index case were compared to national data for chlamydia (CT), gonorrhea (GC) and syphilis (STS). Factors associated with testing at least one partner were examined using multivariable logistic regression. Analyses were performed using STATA 12. Results: 6414 index cases initiated PN via SXT across 13 sexual health providers, median age 25 (IQR 21-32) years, 66%white ethnicity, 58%male and 26% men who have sex with men (MSM), with 6779 STIs; the majority CT (65%), GC (21%) and STS (5%). The number of verified tested partners per diagnosis via SXT vs. national data were higher for CT, GC and STS (Table 1). Based on known STI prevalence in partners, a predicted 133 GC, 77 CT and 12 STS additional diagnoses were made using SXT during the 7 month period. 23-34% of PN was self-verified online by partners. Compared to testing ≥1 partner, black vs. white ethnicity (adjusted OR [95% CI] black African 0.75 [0.58-0.96], black Caribbean 0.70 [0.56-0.89] and black other 0.77 [0.61-0.97]), MSM vs. heterosexual (0.74 [0.61-0.90]), living outside large conurbations (0.47 [0.37-0.59]) or testing online (0.30 [0.26-0.35]) vs. urban clinics and a diagnosis of trichomonas vaginalis (TV) vs. CT (0.57 [0.40-0.81]) were associated with fewer partners tested. Conclusion: An electronic PN tool demonstrated increased PN compared to national data, exceeded national targets for CT, GC and STS, reduced workload, and was successful in large conurbations. Being MSM, of black ethnicity and a having a diagnosis of TV was associated with fewer partners tested, highlighting areas to target for future improvements.’

Poster Abstracts

CROI 2019 394