Table 2.
Outcome | Directly assisted HIVST | Unassisted HIVST |
Differencea % (95% CI) |
Non-inferiority of unassisted HIVSTb | Adjusted Risk Ratioc (95% CI) |
p-valuec |
---|---|---|---|---|---|---|
Primary outcomes, n/N (%) | ||||||
- Successful performance of HIV self-test | 245/263 (93.2) | 222/237 (93.7) | 0.5 (−0.1 to 1.1) | Yes | 1.11 [0.68 to 1.81] | 0.817 |
- Correct interpretation of HIV self-test results | 245/263 (93.2) | 206/237 (86.9) | −6.3 (−10.8 to 2.5) | No | 0.60 [0.36 to 0.98] | 0.019 |
Secondary outcome, n/N (%) | ||||||
- Assistance requested | 110/263 (41.8) | 80/237 (33.8)d | −8.0 (−13.9 to 2.7) | No | 0.87 [0.67 to 1.14] | 0.063 |
- Retention ratee | 263/265 (99.2) | 237/265 (89.4)f | −9.8 (−15.9 to 4.9) | No | 0.13 [0.03 to 0.51] | 0.004 |
- Positivity rate with confirmation HIV testing | 9/263 (3.4) | 4/137 (1.7) | −1.7 (−6.1 to 2.7) | Yes | 0.65 [0.32 to 2.02] | 0.854 |
- Linkage to care | 7/9 (77.8) | 3/4 (75.0) | −2.8 (−9.5 to 5.1) | Yes | 0.89 [0.42 to 1.89] | 0.763 |
- Willing to buy HIV self-test kit if locally availableg | 195/263 (74.1) | 217/237 (91.6) | 17.5 (14.1 to 21.1) | Yes | 4.20 [2.42 to 7.32] | < 0.001 |
aDifference assessed with Wald asymptotic test;
bNon-inferiority was defined as a lower limit > −10 of the 95% CI around the difference in outcomes;
cEstimates and confidence intervals are marginal effect from regression of Poisson;
dMajority (72/80; 90%) requested assistance via telephone; and 8 (10%) participants declared to have been assisted by a trusted person;
eReturn rate was defined as the number of included participant who completed the evaluation throughout the follow-up period;
gThe mean of self-test purchase price was estimated at 2.80 USD per test (limit: 0.33–5.41) in the unassisted HIV self-testing group while it was 2.96 USB per test (limit: 0.36–10.25) in the directly assisted group
CI Confidence internal, HIVST HIV self-testing