The impact of guideline changes on timing of HIV diagnosis, ART uptake, and IPT provision among people living with HIV in South Africa
SHEARER-DISSERTATION-2019.pdf (1.020Mb) (embargoed until: 2022-12-01)
MetadataShow full item record
Background South Africa’s National Department of Health adopted universal test and treat (UTT) for people living with HIV (PLHIV) in September 2016. We sought to describe the impact of this policy change on timing of HIV diagnosis, antiretroviral therapy (ART) uptake, and provision of isoniazid preventive therapy (IPT) among newly diagnosed PLHIV. Methods We included adult PLHIV without tuberculosis who were diagnosed at 14 clinics participating in a cluster randomized trial between January 2015-May 2017. We evaluated the impact of UTT on 30-day ART uptake using difference-in-differences to compare newly eligible PLHIV (CD4 ≥500 cells/mm3) to always eligible PLHIV (CD4 ≤500 cells/mm3). We used quantile (median) regression to estimate the change in median CD4 count at diagnosis and ART initiation before and after implementation of UTT and used Poisson regression to evaluate the association between CD4 count strata and ART initiation. We present IPT provision as simple proportions and evaluated predictors of IPT provision to participants who initiated ART using multi-level Poisson regression. All regression analyses incorporated cluster-robust standard errors. Results Removal of the CD4-based eligibility threshold resulted in a 47% (95% CI: 35%, 59%) increase in 30-day ART uptake among newly eligible participants. The median (IQR) baseline CD4 count was 315.5 (168.0-496.0) cells/mm3, with no change observed before and after UTT. After adjustment, a clinically meaningful increase in CD4 count was observed at ART initiation (92.9 cells/mm3; 95% CI: 54.0, 131.8). Compared to patients with CD4 counts between 51-200 cells/mm3, patients diagnosed with CD4 counts between 351-500 cells/mm3 were less likely to initiate ART under pre-UTT (aRR: 0.79; 95% CI: 0.64, 0.97) while no association was observed under UTT. Of 1,184 PLHIV enrolled at standard of care sites, 24% started IPT. Among participants who started ART ≤30 days, pregnant women (PR: 0.67; 95% CI: 0.49, 0.94) were less likely to be prescribed IPT than non-pregnant women. Conclusions Removal of the ART-eligibility threshold resulted in a large increase in ART uptake, with similar initiation rates across all CD4 strata. However, ART and IPT initiation are below target levels and interventions to improve uptake will likely be required.