CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

decline in female sex workers (FSW) prevalence during the SIDA1/2/3 FSW HIV prevention intervention. We aimed to estimate the population attributable fraction (PAF) of young women to overall HIV incidence during the HIV epidemic compared to FSW. Methods: We used a deterministic age-stratified model of HIV transmission informed and calibrated to Cotonou demographic, sexual behaviour and HIV prevalence data by subgroups and time. The model was used to derive the fraction of all HIV infections acquired (acquisition PAF, A-PAF) or transmitted (transmission PAF, T-PAF) by young women aged 15-24 (including/excluding FSW) over different periods (epidemic onset-1993,1993-2012, 2005-2015). We compared these to PAFs for FSW of all ages, defined as the relative difference in number of incident HIV infections between the baseline scenario and a counterfactual with HIV acquisition risk or transmission risk set to 0 in the relevant subgroup. Results: Young women contributed more to HIV acquisition than transmission, but FSW contributed more to transmission over each time period. Young women’s T-PAF and A-PAF increased over time, and their acquisition of infection was estimated to have contributed to 22% (12-36%) of all HIV infections between the epidemic onset and 1993, increasing to 47%(28-67%) between 1993-2012, and 29%(9-50%) between 2005-2015, and slightly more when young FSWwere included(Figure, A-PAF). Young women alone may have contributed to less than 25% (12-43%) of all HIV transmissions over the different periods (Figure,T-PAF). The FSW T-PAF declined over time (T-PAF= 92% onset-2013, but 7% between 2005-15), to the same order as young women’s T-PAF. Conclusions: Although the contribution of FSW to HIV transmission has declined over time, their role in driving onward HIV transmission remains as important as the role of young women, suggesting that HIV prevention, especially HIV treatment for FSW, must be maintained and enhanced. Young women remain highly vulnerable to HIV acquisition, highlighting the need for prevention interventions focused on them, but their risk alone cannot sustain the HIV epidemic in Benin, nor can interventions focused only on young women end it.

1048 Effect of Eliminating CD4 Thresholds on Number of New ART Initiators in South Africa Jacob Bor 1 ; Shahira Ahmed 2 ; Matthew P. Fox 2 ; Sydney Rosen 2 ; FrankTanser 3 ; Deenan Pillay 3 ;Till Bärnighausen 4

1 Boston Univ Sch of PH, Boston, MA, USA; 2 Boston Univ, Boston, MA, USA; 3 Africa Cntr for Hlth and Pop Studies, Mtubatuba, South Africa; 4 Harvard Sch of PH, Boston, MA, USA Background: WHO now recommends eliminating CD4 count eligibility criteria for ART, with the goal of expanding the numbers of HIV-infected persons on therapy. Using a novel quasi-experimental method to obtain empirical estimates of the effect of CD4 eligibility on ART uptake, we predict the total number of new ART initiators in South Africa that would result from eliminating CD4 thresholds. Methods: We analyzed clinical records from all patients (n=11,307) in the Hlabisa sub-district public sector ART programwith a first CD4 count between August 2011 and December 2012. Using a regression discontinuity design and the 350-cell threshold, we estimated the proportions of patients: initiating ART within 6 months due to a Stage 3 or 4 condition, initiating ART due solely to CD4 count, and not initiating ART despite being eligible. We also estimated the distribution of first CD4 counts: <350, 350-500, and >500. Using national (NDoH) data on the number of ART initiators in 2013 (n=614,000), we estimated the number of new initiators per year if CD4 criteria were eliminated. Results: In our study, 18% of patients initiated ART due to condition and would initiate under any threshold (Fig, bottom). An additional 25%would initiate if the threshold was increased (Fig, middle) and 57%would not initiate despite having an eligible CD4 count (Fig, top). Under a policy extending eligibility to all patients regardless of CD4 count, just 30% {25/(25+57)} of patients newly eligible would be expected to initiate ART within 6 months. Of all patients seeking care, 55% (6256) had a first CD4<350 cells; 20% (2223) CD4 350-500 cells; and 25% (2858) CD4>500 cells. Relative to the proportion of all patients initiating within 6 months (32%, 3658), raising the threshold to 500-cells is expected to have increased the number of initiators by 16% {20*25/32}. Eliminating CD4 criteria will increase new initiators by an additional 17% {25*25/(32*1.16)}. If these numbers hold nationally, then South Africa can expect 98,000 additional initiators per year from raising the threshold to 500 and a further 121,000 initiators per year from eliminating CD4 criteria. Conclusions: Eliminating CD4 thresholds would lead to timely ART initiation by an additional 121,000 South Africans per year (a 4% increase in the total number on ART). Twice that number will enter care and not initiate. Removing CD4 criteria alone, without improving HIV testing, linkage to care, and ART initiation procedures, will not achieve the country’s 90-90-90 targets.

Poster Abstracts

Pa#ents who did not ini#ate despite being eligible with CD4 < 350. These pa#ents would not be expected to ini#ate under an increase in CD4 threshold even though they would be eligible. Pa#ents who did not ini#ate because they were not CD4 eligible (CD4 > 350) but would have if eligible. These pa#ents would be expected to newly ini#ate under an increase in CD4 threshold.

57%

25%

Started ART Within 6 Months Pa#ents who ini#ate despite not being eligible by CD4 count (i.e. WHO IV). These pa#ents would be expected to con#nue to ini#ate under an increase in CD4 threshold.

18%

0 .1 .2 .3 .4 .5 .6 0 100 200 300 400 500 600 700 800 Earliest CD4 Count, cells/ µ L

449

CROI 2016