CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

antiretroviral treatment (ART) to prevent mother-to-child transmission could be indicated but inappropriate ART may have negative psychosocial, interpersonal, and health systems impacts. Methods: In a cluster randomized trial in Kenyan public health programs (NCT03070600), PrEP is offered to HIV seronegative women at ANC. Repeat HIV testing is done at each follow-up visit (monthly in pregnancy, tri-monthly in postpartum). The Kenyan national HIV testing algorithm indicates that if one rapid (Determine) is reactive, a second (First Response) is performed; if discrepant, both tests are repeated by a separate provider and a DNA PCR is performed using standard of care national referral systems. Results: Among 2,231 women enrolled during pregnancy and followed for postpartum care, 3,135 repeat HIV tests have been performed, 7 of which had discrepant rapid results (0.22%, 95% CI: 0.09-0.46%) among 5 individuals. DNA PCR samples were collected on the same day as discrepant results; median time to receipt of PCR results was 22 days (range 16-37). In all 5 initial cases, DNA PCR was negative and none of the women were initiated on ART. Two of 5 women subsequently had repeat discrepant rapid results with repeat negative PCRs, one of whom had subsequent concordant positive rapid results (PCR pending) at delivery and declined ART due to disbelief in rapid test results. Conclusion: False positive results are expected to occur at a low frequency with repeated rapid testing. For individuals who are pregnant or using PrEP, positive results demand urgent ART, but false results could trigger inappropriate ART. As repeat HIV testing during pregnancy and PrEP monitoring expands, the volume of discrepant rapid test results will increase. Our data provide evidence that discrepant results are more likely false positive than true positive. Management of discrepant results needs to balance benefits of rapid ART for PMTCT among true positives, with specific counseling about temporary ART and “disclosure” among women with false positive results. Expedited point-of-care HIV PCR could prevent unclear diagnosis, messaging, and treatment. 779 PILOTING POINT-OF-CARE HIV TESTING AT BIRTH AND 6 WKS POSTNATAL IN 4 KENYAN HOSPITALS Catherine Wexler 1 , Melinda Brown 1 , Matthew Sandbulte 1 , May Maloba 2 , Brad Gautney 2 , Kathy Goggin 3 , Thomas A. Odeny 4 , Shadrack Babu 4 , Elizabeth Muchoki 4 , Maosa Nick 4 , Sarah Finocchario Kessler 1 1 University of Kansas Medical Center, Kansas City, KS, USA, 2 Global Health Innovations, Dallas, TX, USA, 3 Children’s Mercy Hospital, Kansas City, MO, USA, 4 Kenya Medical Research Institute, Nairobi, Kenya Background: Point-of-care (POC) testing can expedite HIV diagnosis and treatment among HIV-exposed infants, particularly if performed at birth. Repeat testing at 6 wks is needed to detect intrapartum infections. Methods: A non-blinded pilot study was conducted at 4 Kenyan hospitals with randomly assigned POC technologies (n=2 GeneXpert, n=2 Alere q). Implementation was tailored to a hospital’s layout, departmental collaborations, and patient flow. Exposed infants enrolled in the study were targeted for POC testing at birth (0 to <2 wks) and the 6-week period (4-8 wks). We report preliminary results for median age at birth and 6-week POC testing for 434 infants born November 3, 2017 to July 5, 2018, uptake of the POC test at both stages, and median age at ART initiation and drug resistance (DR) status for HIV+ infants. Results: Of 434 infants, 358 (82.5%) received POC testing at any timepoint; 219 (61.2%) of these received a POC test within the birth window. Infants tested with POC at birth included 100 (45.7%) tested in Maternity (≤2 d of age) and 119 (54.3%) tested on a return visit (3-14 d of age). The median age at birth POC was 0.4 wks (IQR, 0.1-1.3). An additional 52 (14.5%) infants received a first POC test in the near-birth (2 to <4 wks) period, at median age 2.5 wks (IQR, 2.1-3.0). In the 6-wk window 257 (71.8%) received a POC test, at median age 6.1 wks (IQR, 6.0- 6.3). Among infants receiving an initial POC test at or near-birth, 170 (62.7%) returned for a repeat test in the 6-wk period. The optimum test sequence (first at 0-2 wks, then at 4-8 wks) was achieved for 144 (40.2%) infants. A total of 91 missed opportunities for POC were due to machine breakdown (12: all Alere q),

777 PREFERENCES FOR HOME VS CLINIC AND BLOOD VS SALIVA HIV RETESTING IN PREGNANCY Anjuli D. Wagner 1 , Jill Neary 1 , David A. Katz 1 , Grace John-Stewart 1 , Daniel Matemo 2 , John Kinuthia 2 , Alison L. Drake 1 1 University of Washington, Seattle, WA, USA, 2 Kenyatta National Hospital, Nairobi, Kenya Background: HIV retesting during pregnancy and postpartum is critical to reduce mother-to-child HIV transmission (MTCT) due to incident maternal infections. However, widespread scale-up of this policy may confer additional strain on health systems. HIV self-testing may be an innovative solution for maternal retesting by addressing client access barriers and staffing shortages. Methods: HIV-negative pregnant women were enrolled between November 2017 and August 2018 in Nyanza and Nairobi regions in Kenya. At enrollment, retesting preferences were assessed for location (clinic or home), test type (saliva- or blood-based rapid), and test performer (self or provider). Reasons for preferences were assessed and women were asked to select a test strategy for retesting during the current pregnancy: blood-based testing by a provider in clinic (clinic-based testing [CBT]) or self-testing at home using a saliva-based test (home-based testing [HBT]). Chi-squared and t-tests were used to compare reasons for choice. Generalized linear models (log link, binomial family) were used to assess cofactors for testing strategy. Results: Overall, 1,000 pregnant women were enrolled, with a median gestational age of 28 weeks (Interquartile range [IQR]: 22-32) and median age 24 years (IQR: 21-27). More women elected CBT (665 [67%]) than HBT (335 [34%]) for retesting (p<0.001). Later gestational age was associated with lower likelihood of electing HBT (PR per week: 0.99, 95%CI: 0.98-1.0, p=0.04). Maternal age, parity, income, education, same day HIV testing, marital status, relationship duration, and partner testing history were not associated with choice (p>0.05 for all). Preferences for test location (33% home vs 67% clinic), test operator (31% self vs 69% provider), and test type (32% saliva vs 68% blood) mirrored choice of HBT or CBT. Women who elected HBT were more likely to report being unavailable during clinic hours than women who elected CBT (18% vs 10%, p<0.001) and report longer clinic wait times (73 vs 53 minutes, p<0.001). Conclusion: While pregnant women generally preferred CBT for HIV retesting, HIV self-testing at home was preferred by one-third of women, particularly those with challenges accessing health centers. As HIV retesting scales up in pregnancy and postpartum, HBT may reduce burden on health systems, increase retesting rates, and facilitate efforts to eliminate MTCT. 778 CHALLENGES OF POTENTIALLY FALSE-POSITIVE HIV TESTS IN PREGNANT WOMEN IN THE PrEP ERA Anjuli D. Wagner 1 , John Kinuthia 2 , Julia C. Dettinger 1 , Nancy M. Ngumbau 2 , Laurén Gómez 1 , Salphine A. Wattoyi 2 , Alison L. Drake 1 , Felix Abuna 2 , Jillian Pintye 1 , Ben O. Odhiambo 2 , Grace John-Stewart 1 , Jared Baeten 1 1 University of Washington, Seattle, WA, USA, 2 Kenyatta National Hospital, Nairobi, Kenya Background: HIV testing, done repeatedly over time, is a cornerstone of both antenatal care (ANC) and PrEP care. In many settings, HIV rapid tests are done in sequence to confirm infection, but discrepant results (i.e. one positive, one negative) can occur. Guidelines are lacking for how to make treatment decisions after discrepant rapid results in the context of pregnancy and PrEP where urgent

Poster Abstracts

CROI 2019 301