conferencePaper Cape Town communication orale Nishimwe Marie Protopopescu Camelia Iwuji Collins Okesola Nonhlanhla Spire Bruno Orne-Gliemann Joanna McGrath Nuala Pillay Deenan Dabis François Larmarange Joseph Boyer Sylvie The impact of early ART initiation on HIV disclosure and social support among people living with HIV and followed within a universal test and treat programme in rural South Africa (ANRS 12249 TasP trial) Aim HIV status disclosure and social support have been associated with increased antiretroviral treatment (ART) adherence and better clinical outcomes. We aimed to investigate the impact of early ART initiation on HIV status disclosure and social support among patients in care within the universal test and treat (UTT) cluster-randomized TasP trial conducted in rural South Africa between 2012 and 2016. Method/Issue Following home-based HIV testing, HIV-positive individuals were referred to trial clinics and offered ART regardless of CD4 in intervention arm or according to national guidelines (CD4≤350 cells/mm3 until December 2014, then ≤500 cells/mm3) in control arm. This analysis included patients not ART-treated at baseline (i.e. first clinic visit) and with CD4>500 cells/mm3 who attended at least two clinic visits. HIV disclosure and social support indexes (0-5) were estimated every 6 months: one point was attributed when HIV status was disclosed to - or when the patient reported social support from - at least one member of the following groups: regular partner, family, friends, neighbours, others. We used Poisson mixed effects models, adjusted on individual factors, exploring (i) the impact of time since baseline clinic visit, trial arm and interaction between arm and time; (ii) the impact of ART initiation (time-dependent); and (iii) interaction between arm and ART initiation. Results/Comments Of 3019 patients entering in care in trial clinics, 1597 were not on ART at baseline, 477 had CD4>500 cells/mm3 and 182 had at least two visits (93 in intervention, 89 in control arm). The 182 participants had a median [interquartile range (IQR)] follow-up duration of 13.2 [7.0-18.5] months. After 6 months, 22% had initiated ART in control arm versus 97% in intervention arm, while after 12 months they were 29% and 98% respectively. At baseline, median [IQR] HIV disclosure index was 2 [1;2] versus 1 [1;2] in intervention and control arms, respectively (p=0.361) and median [IQR] social support index was 1 [0;2] versus 2 [1;3] (p=0.053). Both outcomes significantly improved over time. The social support increased significantly faster in intervention arm (Incidence Rate Ratio (IRR) [95% Confidence Interval (CI)]=1.25 [1.04;1.49] per year for the interaction). HIV disclosure was higher in intervention arm (IRR (95% CI)=1.23 [1.10;1.38]), with a similar increase over time between arms. When introducing ART initiation, treatment appeared positively associated with both outcomes (IRR [95% CI]=1.39 [1.17;1.65] for HIV disclosure and 1.36 [1.13;1.63] for social support). For HIV disclosure, differences in ART initiation between arms explained the higher disclosure observed in intervention arm, the effect of arm not being significant anymore. Initiating ART was also associated to increased social support, and this effect was stronger in intervention arm (IRR [95% CI] =1.51 [1.13;2.01] for the interaction). Discussion Our findings suggest that, besides clinical benefits, early ART initiation at high CD4 cell count may also positively influence HIV disclosure and social support. For this second outcome, a longer time of follow-up may however be required to observe this benefit. These findings are encouraging for countries that made the choice of implementing UTT strategies. 2017-11-13 AIDS Impact 13th International Conference http://www.aidsimpact.com/abstracts/-KohAV6cKhSNCJEku24i All rights reserved