Accuracy of unsupervised versus provider-supervised self-administered HIV testing in Uganda: a randomized implementation trial
Asiimwe et al.
Type of approach
Type of assistance
Vulnerable population: Fisherfolks
UNAIDS HIV prevalence (2017)
7.1 [6.6 - 7.7]
The study surveyed 246 high-risk fisherfolk in western Uganda on the most important attributes of HIV tests, including mode of HIV test and specimen collection method, location of HIV test service, price, availability of counseling services, timeliness, and accuracy. Both arms received a brief demonstration on how to use the self-test.
Summary of findings
The HIVST sensitivity was 90% in the unsupervised arm (non-observed) and 100% among the provider-supervised (observed), yielding a difference of -10% (90% CI -21, 1%); non-inferiority was not shown. In a per protocol analysis, the difference in sensitivity was -5.6% (90% CI -14.4, 3.3%) and did show non-inferiority.
Given no costs of service, a home-based HIV self-test had the largest share of preference (24.5%), twice that of the rapid testing currently done at public clinics.
Willingness to pay
Willingness to pay details
In the unsupervised oral HIVST arm (non-observed), sensitivity and specificity were 90% and 95.2%, respectively. In the provider-administered oral HIV testing arm, sensitivity and specificity were 100% and 99.1%, respectively. HIV prevalence did not differ by arm and the absolute difference in sensitivity between study arms was 10% (overall) and 6% when excluding 6 participants who did not report their HIVST results. Overall, the lower limit of the 90% confidence interval was 14%, indicating non-observed oral HIVST was not inferior to the provider-administered oral HIV testing. 75% of participants reported oral HIVST was very easy and 67% would strongly recommend it to a friend/family. 24% of non-observed self-testers reported additional help was necessary and 42% of provider-supervised self-testers requested help. In both arms users primarily reported challenges using a timer. Differences in request for help were not statistically significant across arms. While no self-testers reported errors, errors were observed in 1/5 of participants in provider-supervised arm. Most common errors were: (1) incorrect swabbing of the upper and lower gums (10%), (2) touching the flat pad (5%), and (3) spills of developer fluid (4%).
Linkage to prevention, care and treatment
Within 12-72 hours all participants found HIV-positive received confirmatory rapid testing were linked to care/referral.