Table 2. Before-After Trend in Positivity for Participants With No Record of Prior PrEP Use and 2 or More Tests in Year Before PrEPa.
| Disease and site of infection | Before PrEP, change/quarter (95% CI), % | P valueb | After PrEP, change/quarter (95% CI), % | P valueb | RR (95% CI) | P valueb |
|---|---|---|---|---|---|---|
| Chlamydia | ||||||
| Anorectal | 1.09 (1.03-1.15) | .002 | 1.00 (0.98-1.03) | .86 | 0.92 (0.86-0.97) | .005 |
| Pharyngeal | 1.06 (0.93-1.22) | .38 | 1.04 (0.97-1.12) | .25 | 0.98 (0.84-1.14) | .80 |
| Urethral | 1.13 (1.04-1.23) | .004 | 1.01 (0.97-1.05) | .57 | 0.89 (0.81-0.98) | .02 |
| Any site | 1.11 (1.06-1.16) | <.001 | 1.01 (0.99-1.03) | .16 | 0.91 (0.87-0.96) | <.001 |
| Gonorrhea | ||||||
| Anorectal | 1.07 (0.99-1.15) | .11 | 0.99 (0.96-1.02) | .53 | 0.93 (0.85-1.01) | .09 |
| Pharyngeal | 1.02 (0.96-1.09) | .47 | 0.96 (0.93-0.99) | .01 | 0.94 (0.87-1.004) | .07 |
| Urethral | 1.02 (0.91-1.13) | .78 | 0.99 (0.95-1.04) | .70 | 0.98 (0.87-1.10) | .68 |
| Any site | 1.07 (1.02-1.12) | .002 | 1.00 (0.98-1.02) | .81 | 0.94 (0.89-0.98) | .007 |
| Chlamydia or gonorrhea | ||||||
| Anorectal | 1.08 (1.03-1.13) | .001 | 1.00 (0.98-1.02) | .82 | 0.93 (0.88-0.97) | .002 |
| Pharyngeal | 1.03 (0.97-1.09) | .36 | 0.97 (0.94-1.00) | .04 | 0.94 (0.89-1.01) | .08 |
| Urethral | 1.09 (1.02-1.17) | .01 | 1.00 (0.97-1.03) | .97 | 0.92 (0.85-0.99) | .02 |
| Any site | 1.08 (1.05-1.12) | <.001 | 1.01 (0.99-1.02) | .35 | 0.93 (0.90-0.96) | <.001 |
| Syphilis | 1.10 (0.94-1.29) | .24 | 1.12 (1.02-1.23) | .01 | 1.02 (0.85-1.22) | .85 |
| Any disease, any site | 1.08 (1.05-1.11) | <.001 | 1.01 (0.99-1.02) | .29 | 0.93 (0.90-0.96) | <.001 |
Abbreviations: PrEP, Preexposure prophylaxis; RR, rate ratio.
a
A total of 2404 EPIC-NSW participants with testing after enrollment and no record of prior PrEP and with 2 or more test in the year prior to enrollment were included; 587 of 30 549 tests (1.9%) did not have specified anatomical site, including 263 of 12 733 (2.1%) before PrEP and 324 of 17 492 (1.8%) during PrEP (χ2 test for difference, P = .12). Multivariable binomial-log regression was adjusted for PrEP status and quarter of follow-up, which are patient-level random effects. Intervention effect RRs are available in eTable 3 in the Supplement.
b
Wald test.