CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

cost per patient achieving the primary outcome and the incremental cost-effectiveness ratio (ICER) for the intervention and conducted sensitivity analysis for 3 key parameters: omitting the POC TB test; shifting some initiation tasks to junior nurses; and setting patient volume equal to the clinic’s full patient load. Results: In the standard arm, 46/108 (43%) patients were suppressed by 10 months, the average cost per patient suppressed was $483, and the total cost per patient suppressed (taking into account costs for patients not suppressed) was $836. In the rapid arm, 67/105 (63%) patients were suppressed by 10 months, the average cost per patient suppressed was $524, and the total cost per patient suppressed was $899. Standard and rapid arm patients averaged 9.9 and 6.9 clinic visits during the 10 month-period, respectively. Laboratory costs for baseline CD4 counts, TB tests, and other blood tests were much higher in the POC-based rapid arm, accounting for most of the difference; attributed fixed costs for infrastructure and management were lower in the rapid arm due to the smaller number of clinic visits. The ICER for the rapid strategy was $780 per additional patient suppressed by 10 months. Sensitivity analysis (see figure) indicates that omitting POC TB tests and increasing patient volume may reduce costs substantially. Conclusion: Same-day treatment initiation using POC tests is more effective and more expensive than standard initiation. Variations on the strategy, in particular omitting POC tests at initiation and/or focusing on high volume clinics, have the potential to reduce costs substantially and should be evaluated in routine settings.

Poster and Themed Discussion Abstracts

1051 COST OF STREAMLINED HIV CARE IN RURAL KENYAN AND UGANDAN CLINICS IN THE SEARCH STUDY Starley B. Shade 1 , Thomas Osmand 1 , Florence Mwangwa 2 , Asiphas Owaraganise 2 , Tamara D. Clark 1 , Edwin D. Charlebois 1 , Moses R. Kamya 3 , Maya L. Petersen 4 , Diane V. Havlir 1 , Vivek Jain 1 1 Univ of California San Francisco, San Francisco, CA, USA, 2 Infectious Diseases Rsr Collab, Kampala, Uganda, 3 Makerere Univ Coll of Hlth Scis, Kampala, Uganda, 4 Univ of California Berkeley, Berkeley, CA, USA Background: The SEARCH Study-an HIV test-and-treat community-randomized trial-has achieved >90% HIV testing and antiretroviral (ART) treatment for all HIV+ stable adult residents. Our “streamlined” HIV care model reduces wait times and facilitates viral suppression (84%). Standard HIV care delivery is estimated to cost $224-$1089 per-person-per- year (ppy). We sought to estimate the cost of streamlined HIV care delivery in SEARCH, and to model the cost of streamlined HIV care under optimized scale-up conditions. Methods: We estimated the cost ppy of streamlined HIV care delivery in 17 health facilities in intervention communities in Kenya and Uganda within the SEARCH Study (NCT:01864603). Streamlined HIV care utilizes a patient-centered, multi-disease approach in a supportive environment; nurse-driven appointments (q 3 months) with co-located clinical/phlebotomy/laboratory services; appointment reminders; telephone access to clinicians; and viral load (VL) testing and counseling (q 6 months). We calculated costs using standard micro-costing techniques, time-and-motion studies, interviews of supervisory staff, and administrative records review. Cost categories included clinical and supervisor staff salaries, ART medications, VL testing, and fixed and recurring costs. We modeled HIV care costs under optimal scale-up conditions: (1) transition to governmental personnel; (2) lowest available ART costs (using UNDP negotiated rates); and (3) actual costs of centralized VL testing. Results: Streamlined HIV care delivery averaged $275 ppy. ART medications ($118/ppy for TDF/3TC/EFV) and VL testing ($110 ppy for 2 tests per year) dominated costs relative to staff salaries [$38 ppy for clinical ($37 ppy) and supervisory ($1 ppy) staff], fixed costs for infrastructure and equipment ($5 ppy) and other recurring goods and services ($4 ppy). In an optimized scale-up model featuring government salaries ($27 ppy), lowest available ART costs ($100 ppy) and annual VL testing ($12 ppy), the overall cost of streamlined HIV care dropped to $148 ppy. Conclusion: Costs of streamlined HIV care within the SEARCH test-and-treat trial were similar to or lower than previous standard HIV care cost estimates, even after including costs for VL testing and counseling. Optimized models of care delivery would substantially reduce these costs below prevailing estimates. These data can inform global cost and policy formulations focused on financing the expansion of ART to achieve UNAIDS 90-90-90 targets. 1052 DIFFERENTIATED HIV RNA VIRAL-LOAD MONITORING IN UGANDA: A COST-EFFECTIVENESS ANALYSIS Diana M. Negoescu 1 , Zhenhuan Zhang 1 , Heiner C. Bucher 2 , Eran Bendavid 3 , for the Swiss HIV Cohort Study 1 Univ of Minnesota, Minneapolis, MN, USA, 2 Basel Inst for Clinical Epi and Biostatistics, Basel, Switzerland, 3 Stanford Univ, Stanford, CA Background: Viral load (VL) monitoring for patients receiving antiretroviral therapy (ART) is recommended worldwide. However, the costs of frequent monitoring are a barrier to implementation in resource-limited settings. The extent to which personalized monitoring frequencies may be cost-effective is unknown. Methods: We created a simulation model parameterized using person-level longitudinal data from the Swiss HIV Cohort Study to assess the benefits of flexible monitoring frequencies. Our data-driven model estimated CD4 evolution during and prior to virologic failure, and the relationship between virologic failure and self-reported adherence, duration on regimen, age, and gender. Our model tracked a population of HIV+ individuals for 10 years following ART initiation. Adherence was modeled as a time-varying process that depends on the patient’s previous adherence status, age, gender, and education level. We used a Markov version of the model to optimize the interval between viral load tests

CROI 2017 454