Like other countries, Rwanda started implementation of decentralization in 2003 however there has not
been evaluation of successes of this policy. Therefore the aim of this study is to evaluate impact of decentralization
of ART services in two rural districts of Rwanda. We used routinely corrected data of adult HIV patients that
visited 21 health facilities. We matched patients who were decentralized (exposed group) to patients who were not
decentralized using coarsened exact matching method. Chi-square test was used to assess association between each
variable and outcome. We then performed conditional logistic regression to assess impact of decentralization on
patient’s outcomes. Stata V12 (College Station TX: StataCorp LP) and ArcGIS version 10.3 software were used in
the analysis. We found that there was no significant difference in baseline characteristics among these groups.
Before decentralization patients in exposed group were more likely to have gap in treatment compared to patients
in unexposed group (OR: 1.31; 95% CI: 1.88, 2.06). However, after decentralization, patients who were
decentralized were less likely to have gap in treatment compared to patients who were not decentralized (OR: 0.14,
95% CI: 0.074, 0.26) and there was no significant difference in number of deaths or lost to follow between both
groups.
This study informs Ministry of Health and Government of Rwanda in general that decentralization policy was
successful. It also informs other countries with similar settings as Rwanda that decentralization of ART services is
a feasible policy that can be implemented to increase adherence of ART patients.