Estimating the effect of COVID-19 pandemic restrictions on self-reported antiretroviral therapy use and late refill visits among people living with HIV in rural South Africa
Ashley S. Tseng, Kenneth K. Mugwanya, Adam A. Szpiro, Alastair van Heerden, Xolani Ntinga, Torin T. Schaafsma, Ruanne V. Barnabas- Pharmacology (medical)
- Infectious Diseases
Background:
People living with HIV require reliable access to and adequate supply of antiretroviral therapy (ART) for viral suppression. The Deliver Health Study, a randomized trial conducted during the COVID-19 pandemic, found that home-delivered ART significantly increased viral suppression compared to clinic-based care. The effect of changing COVID-19 alert levels on self-reported ART use has not been quantified.
Setting:
KwaZulu-Natal, South Africa.
Methods:
Adults living with HIV were followed in the Deliver Health Study during October 2019-December 2020. We used difference-in-differences (DiD) to estimate the effect of changing COVID-19 alert levels during three distinct periods on self-reported missed ART doses (missed 0 vs. 1 doses in past week) for participants receiving home-delivered vs. clinic-based refills. We additionally estimated the effect of changing COVID-19 alert levels on late clinic ART refill visits (late vs. on-time). We used relative risk regression for both binary outcomes.
Results:
Of 155 participants, 46% were women and median age was 36 years. The mean number of missed weekly doses was 0.11, 0, and 0.12 in the home-delivery group and 0.09, 0.08, and 0.18 in the clinic group during periods 1, 2, and 3, respectively. There were no differences in relative risk (RR) of self-reported daily ART use between refill groups when comparing across periods (DiDperiod 2 vs. 1=1.05; 95% confidence interval [CI]: 0.97, 1.13 and DiDperiod 3 vs. 2=0.99; 95% CI: 0.91, 1.08). In the clinic group, the risk of late refill visits was significantly higher during COVID-19 restrictions (vs. before alert level 5 implementation) and even after the COVID-19 alert level was downgraded to level 1 (RRperiod 2 vs. 1=1.83, 95% CI: 1.34, 2.51 and RRperiod 3 vs. 2=1.71; 95% CI: 1.43, 2.04).
Conclusion:
The COVID-19 pandemic did not differentially impact self-reported ART adherence by method of ART refills, but the risk of late clinic refill visits was significantly higher during COVID-19 restrictions and sustained after restrictions were loosened.