Table 2.
Estimated item response probabilities of self-reported use of HIV prevention strategies among the HIV-negative/unknown participants of the Engage-Montreal study, 2017–2018 (n = 968): 4 class modela
| Variable | Class 1: Biomedical prevention use n = 113 (7%, 95% CI 4–10%) |
Class 2: Condom use n = 341 (40%, 95% CI 34–45%) |
Class 3: Seroadaptive behaviour use n = 241 (21%, 95% CI 17–26%) |
Class 4: Low use of prevention n = 273 (32%, 95% CI 27–37%) |
|---|---|---|---|---|
| HIV testing | 100% | 53% | 67% | 18% |
| PrEP | 84% | 0% | 7% | 0% |
| PEP | 53% | 13% | 18% | 2% |
| Consistent condom use | 34% | 83% | 45% | 34% |
| Strategic positioning | 35% | 38% | 46% | 6% |
| Serosorting or viral load sorting | 65% | 0% | 100% | 33% |
CI confidence interval, PrEP pre-exposure prophylaxis, PEP post-exposure prophylaxis
aRDS-II adjusted class sizes (%) and 95% confidence intervals are presented. RDS-II weights are inverse probability of sampling weights that are proportional to participant network size