Title

Favorable effect of differentiated models of care on retention in care and viral suppression among adults on antiretroviral treatment: retrospective cohort study in Zambézia Province, Mozambique (2016-2021)

Presenter

José Tique

Authors

C. De Schacht¹, J. Tique¹, C. Veríssimo², C. Cabsela¹, C. Belo¹, J. Matsimbe³, E. Graves⁴, G. Amorim⁵, H. Macul⁶, C.W. Wester⁷

Institutions

¹Friends in Global Health, Maputo, Mozambique, ²Provincial Health Directorate Zambézia, Quelimane, Mozambique, ³Friends in Global Health, Quelimane, Mozambique, ⁴Vanderbilt University Medical Center (VUMC), Vanderbilt Institute of Global Health (VIGH), Nashville, United States, ⁵Vanderbilt University Medical Center (VUMC), Department of Biostatistics, Nashville, United States, ⁶Ministry of Health, National Directorate of Public Health, Maputo, Mozambique, ⁷Vanderbilt University Medical Center (VUMC), Department of Medicine, Division of Infectious Diseases, Nashville, United States

BACKGROUND: Differentiated service delivery (DSD) models have been implemented in Mozambique to de-congest health facilities, and promote retention of persons living with HIV (PLHIV) in care. This study evaluated the effects of two frequently used DSD models (Community Adherence Groups [CAG] and Three Multi-Month Dispensing [3MMD]) on retention in care and viral suppression.
METHODS: A cohort study using routine patient-level data was implemented among adults (=15 years) enrolled in ART services between October 2016-September 2020 and eligible for the DSD models, in 147 health facilities (HF) in Zambézia Province. Propensity score matching was used to match PLHIV in CAG to those in 3MMD. Conditional logistic regression models measured associations between the DSD model and 12-month retention (pick-up within 59 days after last scheduled visit) and viral suppression (viral load <1,000 copies/mL, measured within one year after DSD model enrollment), adjusting by HF location (rural vs urban), DSD model and their interaction.
RESULTS: Data from 46766 PLHIV were collected; 31340 (67%) female, 30512 (65%) registered at rural HF, median age at DSD eligibility 30 years (IQR 24-38). From this cohort, 38118 (82%) PLHIV enrolled in 3MMD, 3129 (7%) in CAG; 5527 (12%) were not included in any DSD. A matched population of 4936 PLHIV were included in retention analysis, and 1610 in viral suppression analysis. The overall 12-month retention was 93% and 94% in the 3MMD and CAG groups, respectively. Viral suppression was 86% overall, 83% for 3MMD and 89% for CAG. In rural areas, the odds of being retained at 12 months was 1.5 times higher for PLHIV in CAG compared to 3MMD (OR 1.50 [95%CI:1.14-1.97], p=0.003). PLHIV in CAG in rural areas also had higher odds of being virally suppressed (OR 2.03 [95%CI:1.43-2.88], p<0.001). There were no differences among PLHIV in urban areas.
CONCLUSIONS: In this cohort, most PLHIV were enrolled in the quarterly dispensation model. Retention in care and viral suppression was high for both models, but advantages were seen for CAG among PLHIV attended in rural areas. Targeted models considering area of service delivery can contribute to maintaining PLHIV in the continuum of care.

Download the e-Poster (PDF)