CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

723 TB PREVENTIVE THERAPY UPTAKE IS HIGH WITH COMMUNITY ART DELIVERY IN SOUTH AFRICA Adrienne E. Shapiro 1 , Alastair van Heerden 2 , Heidi van Rooyen 2 , Torin T. Schaafsma 1 , Olivier Koole 3 , Deenan Pillay 3 , Jared Baeten 1 , Connie L. Celum 1 , Ruanne V. Barnabas 1 , for the DO ART Study Team 1 University of Washington, Seattle, WA, USA, 2 Human Sciences Research Council, Pretoria, South Africa, 3 Africa Health Research Institute, Mtubatuba, South Africa Background: Isoniazid (INH) preventive therapy (IPT) reduces mortality and tuberculosis (TB) in persons with HIV and is recommended for all HIV+ persons in high TB prevalence settings, including South Africa, but uptake is low. Barriers to IPT include lack of provider education and prioritization, screening with non-specific TB symptoms, IPT provided separately from HIV services, and monthly clinic visits for refills. Our objective was to increase uptake of IPT for HIV-infected persons newly initiating ART. Methods: IPT was integrated into community-based ART in the DO ART Study in peri-urban communities in KwaZulu-Natal, South Africa. DO ART is an ongoing randomized clinical trial with intervention arms providing community-based ART delivery, quarterly refills, mobile monitoring, and access to facility-based services only as needed. Between 7/2017-9/2018, 388 HIV-infected adults (149 (38%) men) received care in the community-based arms including ART initiation and at least one follow-up visit. Participants were screened by lay health workers for TB symptoms and contraindications to IPT at every visit. Eligible participants were offered IPT starting 1 month after ART initiation. IPT refills were quarterly and synchronized with ART. We assessed IPT acceptance, refusal, and receipt of a first refill as indicators of feasibility and acceptability. In 7/2018, we began testing urine for INH metabolites at community IPT refill visits to confirm self-reported IPT adherence. Results: 388 participants who received community-based ART were screened for IPT eligibility. 355 (91%) were eligible and initiated IPT. There were 10 refusals, 5 of whom initiated IPT at a subsequent visit. 99% participants reported no side effects or toxicities. 3 persons self-discontinued IPT due to side effects. Self-reported adherence was high. 94% completed a first refill visit and received an indicated IPT refill. Urine testing confirmed presence of INH in 21 of 28 (75%) persons spot-tested for adherence. Among 125 participants who initiated ART at a clinic, 7 (6%) reported receiving IPT. No incident TB cases were reported. Conclusion: High levels of IPT uptake and continuation were achieved in a community-based ART project, and demonstrated feasibility, high safety, adherence, and acceptability in South Africa. Community-based IPT can be effectively provided in a differentiated HIV care model with infrequent clinic contacts, and may have better uptake than clinic-based IPT. Urine testing can complement self-reported adherence.

722 RURAL AND URBAN DIFFERENCES IN THE IMPACT OF THE XPERT MTB/ RIF TEST ON TB CARE Simon Walusimbi 1 , Stella Zawedde-Muyanja 1 , Julius Sendiwala 2 , Abudnoor Nyombi 2 , Stavia Turyahabwe 2 , Andrew Kambugu 1 , Barbara Castelnuovo 1 , Yukari C. Manabe 3 1 Makerere University, Kampala, Uganda, 2 Ministry of Health Uganda, Kampala, Uganda, 3 Johns Hopkins Hospital, Baltimore, MD, USA Background: Since 2014, utilization of the Xpert MTB/Rif test (Xpert) for diagnosis of Tuberculosis (TB) has increased compared with smear microscopy in Uganda. In 2016, more than half of the health facilities with onsite Xpert were rural. However, the impact of the increased uptake of Xpert on the care cascade and health outcomes for TB patients remains unclear. We hypothesized that the care cascade for HIV-associated TB in rural health facilities with onsite Xpert would be similar to that of urban health facilities with onsite Xpert. Methods: We retrieved electronic data on health facility outpatient attendance, number of TB patients diagnosed, treated, and their outcomes from the national HMIS database (June 2016 to July 2017), and rural versus urban placement status of Xpert from the National TB Reference Laboratory reports. We estimated prevalent TB using the total number of outpatients with a diagnosis of any cough or pneumonia. Based on review of local and regional literature, we assumed that 2% of individuals with any cough had TB while 12% of individuals with pneumonia had TB. Of the total prevalent TB, we assumed that 42% had TB/HIV co-infection based on the national TB report of 2016. We computed the absolute counts and percentages for each of the following steps of the TB care cascade: number of prevalent TB patients at the health facility, number of diagnosed TB patients, number of TB patients cured or completed treatment and number of TB deaths. Results: Data was obtained from a total of 758,823 patients from 106 health facilities with onsite Xpert of which 57/106 (54%) were rural. Rural health facilities had 299,643 patients (39%) with any cough and 30,197 patients (4%) with pneumonia, while urban facilities had 386,293 patients (51%) with any cough and 42,690 patients (6%) with pneumonia. Rural facilities diagnosed 1,855/4039 (46%) of the estimated prevalent TB/HIV cases versus 5,101/5397 (95%) at urban facilities. Treatment cure/completion was 60% for rural facilities versus 52% for urban facilities (p<0.001). Mortality was similar in rural and urban health facilities (12% rural versus 12% urban). Conclusion: Despite increased placement of Xpert in rural health facilities, they detected less than half of their prevalent TB cases compared with urban health facilities. Rural facilities had better cure/completion treatment outcomes however, mortality was similar in both settings. Focused interventions are required to address these distinct quality gaps in TB care.

Poster Abstracts

CROI 2019 278