Share
Title
Trends in interruptions in treatment among men, pregnant women and non-pregnant women: retrospective cohort study in Zambézia Province, Mozambique (2013-2021)
Presenter
Erin Graves
Authors
E. Graves1, C. De Schacht2, Z. Yu3, M. Bravo2, C. Cugara4, C. Veríssimo4, J. Tique2, K. Tibana5, G. Amorim3, C. Belo2, A. Couto5, C.W. Wester1,6
Institutions
1Vanderbilt University Medical Center (VUMC), Vanderbilt Institute for Global Health (VIGH), Nashville, United States, 2Friends in Global Health (FGH), Maputo, Mozambique, 3Vanderbilt University Medical Center (VUMC), Department of Biostatistics, Nashville, United States, 4Provincial Health Directorate of Zambézia (DPS-Z), Quelimane, Mozambique, 5Ministry of Health (MOH), National Directorate of Public Health, Maputo, Mozambique, 6Vanderbilt University Medical Center (VUMC), Department of Medicine, Division of Infectious Diseases, Nashville, United States

BACKGROUND: Since Option B+ strategy commenced in Mozambique in July 2013, universal ART for all persons living with HIV (PLHIV) (Test & Start, "T&S") was introduced in 2016. We describe trends over time in interruption in treatment (IIT) outcomes between PLHIV groups (pregnant women [PW], non-PW and men).
METHODS: Retrospective cohort study was conducted involving adult (=15 years old) PLHIV who initiated antiretroviral treatment (ART) between July 2013-June 2021 in 107 health facilities in Zambézia Province. Routine data were used to summarize temporal trends in proportions of PW, non-PW, and men experiencing an IIT (i.e., having no clinical contact for 28 days after last scheduled/expected contact) less than 3 or within 3-5 months post-ART initiation (“IIT<3m”, “IIT3-5m”, respectively). Mixed-effect logistic models were built with district as random effect and splines on time variable.
RESULTS: Data from 228,628 adults were included, 25.5% PW, 40.4% non-PW. Overall, monthly IIT<3m proportions for each group decreased from approximately 55% to 20%. Trends in IIT<3m were more pronounced for PW; while men and non-PW showed continued higher proportions of early IIT, improvements were seen from 2019 on. Similarly, monthly IIT3-5m proportions for all groups decreased from 2013-2021 (~65% to ~18% for PW, ~50-55% to ~20-25% for non-PW and men), with variations among groups: higher proportions of PW experienced an IIT3-5m from 2013 to early 2016, while non-PW and men had biennial increases in IIT3-5m from 2013-2018. After 2018, IIT3-5m trends consistently decreased for all groups (Figure).


CONCLUSIONS: Trend analysis showed prominent decrease in treatment interruptions in Zambézia for all groups. Men and non-PW had overall slightly higher proportions of IIT, with significant improvements among these groups after T&S was introduced, while the established Option B+ strategy showed continued positive effect. Though trends are reassuring on early retention, continued efforts are needed to ensure sustained effect.

Download the e-Poster (PDF)