Background: There is concern that earlier initiation of antiretroviral treatment (ART) in lower resource settings may compromise access to care for patients with lower CD4 counts, and that patients with higher CD4 counts may have lower retention in care (RIC). In July 2014, we extended ART initiation criteria from CD4 cell counts of ≤350 to ≤500 copies/ µl in 9 primary health clinics in KwaZulu-Natal, South Africa. Here we assess whether any compensatory reduction in initiation of sicker patients was seen and whether retention among those newly eligible was satisfactory in this public sector setting.
Methods: In this retrospective cohort analysis we compare proportions initiated on ART and RIC at 6 months among patients with baseline CD4 taken between July 1 and December 31, 2014 (CD4≤500 eligibility cohort) and between July1 and December 31, 2013 (CD4≤350 eligibility cohort). Pregnancy, TB, age < 15 years and WHO stage 3 or 4 were exclusion criteria. Outcomes were determined from baseline CD4 and analysed using survival analysis.
Results: There were 768 patients in the CD4≤350 eligibility cohort, with 31% having baseline CD4 ≤200; 51% 201-350, and; 12% 351-500. Of the 856 in the CD4 ≤500 eligibility cohort 23% had a baseline CD4 ≤200, 37% 201-350 and 33% 350-500. In both cohorts, median age was 31 years and 67% were female. Among participants with CD4 351-500, percentage initiated on ART within 3 months increased 10 fold between the periods from 7% (95%CI:3.4-13.0) to 70% (95%CI: 61-78); among those with CD4≤200 this increased from 70% (95%CI:55-80) to 86% (95%CI:79-93). The proportion initiated within 3 months among those with baseline CD4 201-350 remained unchanged at approximately 75%. RIC at 6 months was 82% (95%CI:79%-85%) in the CD4≤500 cohort and 80% (95%CI:76%-84%) in the CD4≤350 cohort.
Conclusions: Expanding eligibility for ART to CD4≤500 resulted in rapid change in time to ART initiation among those with baseline CD4 351-500 without compromising initiation or RIC among those with a CD4 ≤ 350. Extended initiation criteria can be successfully implemented in high HIV prevalence, low resources-settings without compromising access to care for more vulnerable patients.