Cluster-randomized trial of routine syphilis screening among men attending HIV outpatient clinics in Ontario, Canada

Title

Cluster-randomized trial of routine syphilis screening among men attending HIV outpatient clinics in Ontario, Canada

Presenter

Ann Burchell

Authors

A. Burchell * (1,2), D. Tan (1,2), R. Grewal (1,2), A. Kroch (3), P. MacPherson (4,5), S. Walmsley (6,2), A. Rachlis (2,7), N. Andany (7,2), S. Gardner (2), J. Raboud (2), S. Mishra (1,2), C. Cooper (4,5), D. Fisman (2), K. Gough (1,2), I. Salit (6,2), T. Mazzulli (2,8), R. Rousseau (2), J. Maxwell (9), V. Allen (2,8)

Institutions

(1) Unity Health Toronto, Toronto, Canada, (2) University of Toronto, Toronto, Canada, (3) Ontario HIV Treatment Network, Toronto, Canada, (4) The Ottawa Hospital, Ottawa, Canada, (5) University of Ottawa, Ottawa, Canada, (6) University Health Network, Toronto, Canada, (7) Sunnybrook Hospital, Toronto, Canada, (8) Sinai Health, Toronto, Canada, (9) ACT, Toronto, Canada

BACKGROUND: The Enhanced Syphilis Screening Among HIV-positive Men (ESSAHM) Trial was a stepped wedge cluster-randomized trial that paired opt-out syphilis tests with routine HIV bloodwork to improve early syphilis case detection (https://doi.org/10.1093/cid/ciab582). Few patient covariates were collected given the pragmatic trial approach. We re-analysed intervention effects using data from a concurrent cohort study that ascertained sociodemographic characteristics and sexual histories.
METHODS: Population: HIV-positive men; intervention: standing orders for syphilis testing with HIV viral loads; control: usual syphilis testing practice; outcomes: case detection, proportion screened ('coverage'), and screening frequency; time: 01/02/2015 to 31/07/2017. Trial clinics were also sites for the Ontario HIV Treatment Network Cohort Study (OCS); annual interviews measure income, education, sexual orientation, and number of sexual partners in the past 3 months. We quantified intervention effects among OCS participants using time-adjusted generalized linear mixed-effect models to estimate odds ratios (OR) and rate-ratios (RR) with 95% confidence intervals (CI) and explored evidence for confounding and effect modification by sociodemographic and sexual history covariates.
RESULTS: 34% (1325/3895) of trial participants were also OCS participants (mean 52.0 years old, 86% men who have sex with men). At baseline, 25% (n=316) reported 2+, 30% (n=365) reported 1, and 45% (n=556) reported 0 sex partners. Comparing intervention to control periods, there were increases in case detection (OR 1.83, CI 0.85, 3.93), the proportion screened (OR 2.81, CI 2.20, 3.59), and number of tests per year (RR 1.73, CI 1.51, 1.97). There was no evidence of confounding by sociodemographic characteristics. However, intervention effects were modified by number of sex partners (Table).

CONCLUSIONS: Routine screening tripled detection of early syphilis among men more likely to have had sexual exposure. Although the magnitude of increases in screening coverage and frequency were less among men with multiple partners compared to men without, there was still improvement compared to usual care.

Download the e-Poster (PDF)