HIV suppression was maintained during the COVID-19 pandemic in Malawi: a program-level cohort study

Abstract

Background and ObjectivesMeasures introduced to reduce the spread of SARS-CoV-2 by the Malawi government and the national HIV care program might have compromised treatment outcomes of patients living with HIV on antiretroviral therapy (ART). We studied viral load (VL) outcomes before and during the COVID-19 epidemic in Malawi. MethodsIn this population-based cohort study, we included all routine VL measurements collected from July 2019 to December 2020 in about 650 ART clinics in Malawi. We examined differences between pandemic periods (before/during COVID-19) for i) VL monitoring, and ii) VL suppression (VLS: <1,000 copies/ml). For i) we studied the number of VL measurements over time and assessed predictors of missed measurements before and during COVID-19 in logistic regression models. For ii) we estimated the odds of VLS before and during the COVID-19 epidemic stratified by treatment regimen using generalized estimation equations adjusted for age, sex, time on ART, and type of biological sample. We imputed missing treatment regimens by population-calibrated multiple imputation. ResultsWe included 607,894 routine VL samples from 556,281 patients. VL testing declined during COVID-19 (243,729; 40%) compared to before COVID-19 (365,265; 60%), but predictors of missing tests were similar in the two periods. VLS rates increased slightly from 93% before to 94% during COVID-19. Compared to before COVID-19, the odds of VLS increased during COVID-19 for patients on protease inhibitor-based (PI) regimens (adjusted odds ratio [aOR] 1.22, 95% CI: 0.99-1.49) and for patients on integrase strand transfer inhibitor-based (INSTI) regimens (aOR 1.10, 95% CI: 1.03-1.17). There was no difference in VLS between the two periods among patients on nonnucleoside reverse transcriptase inhibitor-based (NNRTI) regimens. VLS varied by age, sex, regimen, and duration on ART, ranging from 45.1% (95% CI 40.3-50.0%) to 97.2% (95% CI 96.9-97.4%). ConclusionThere was a significant decline in VL monitoring during COVID-19, but we did not find clear evidence that the pandemic reduced VL suppression rates. Routine scheduled VL monitoring, targeted adherence support, and timely regimen switches for patients with treatment failure remain critical to improving VLS.