Increased demand for antiretroviral therapy (ART) services combined with plateaued levels of development assistance for HIV/AIDS requires that national ART programs monitor program effectiveness. In this pilot study, we compared commonly utilized performance metrics of 12- and 24-month retention with rates of viral load (VL) suppression at 15 health facilities in Uganda.


Retrospective chart review from which 12- and 24-month retention rates were estimated, and parallel HIV RNA VL testing on consecutive adult patients who presented to clinics and had been on ART for a minimum of six months. Rates of VL suppression were then calculated at each facility, and compared to retention rates to assess correlation of performance metrics. Multilevel logistic regression models predicting VL suppression and 12- and 24-month retention were constructed to estimate facility effects.


We collected VL samples from 2,961 patients, and found that 88% had a VL <1,000 copies/mL. Facility rates of VL suppression varied between 77% and 96%. When controlling for patient mix, significant variation in facility performance persisted. Retention rates at 12 and 24 months were 91% and 79%, respectively, with comparable facility-level variation. However, neither 12- (ρ=0.16) nor 24-month (ρ=-0.19) retention rates were correlated with facility rates of VL suppression.


Retaining patients in care and suppressing VL are both critical outcomes. Given the lack of correlation noted in this study, the utilization of VL monitoring may be necessary to truly assess the effectiveness of health facilities delivering ART services.


Duber HC, Roberts DA, Ikilezi G, Fullman N, Gasasira A, Gakidou G, Haakenstad A, Levine AJ, Achan J. Evaluating facility-based antiretroviral therapy program effectiveness: a pilot study comparing viral load suppression and retention rates. Tropical Medicine & Hygiene. 2016 March 21. doi: 10.1111/tmi.12694.