LBPEB16

Title

Risk of sexually transmitted infections among women randomized to DMPA-IM, the copper IUD, and levonorgestrel implant in the Evidence for Contraceptive Options and HIV Outcomes (ECHO) trial

Presenter

Jen Deese

Authors

J. Deese¹, N. Philip², M. Lind³, K. Ahmed⁴, J. Baeten³, J. Batting⁵, M. Beksinska⁶, D. Donnell^3,7, V. Edward⁸, C. Louw⁹, T. Mastro¹, N. Mugo¹⁰, K. Nanda¹, M. Onono¹⁰, H. Rees¹¹, R. Sihlongonyane², J. Smit⁶, C. Morrison¹, ECHO Trial team

Institutions

¹FHI 360, Durham, United States, ²ICAP at Columbia University, New York, United States, ³University of Washington, Seattle, United States, ⁴Setshaba Research Centre, Pretoria, South Africa, ⁵Effective Care Research Unit, East London, South Africa, ⁶MatCH Research Unit, University of the Witwatersrand, Durban, South Africa, ⁷Fred Hutchinson Cancer Research Center, Seattle, United States, ⁸The Aurum Institute, Johannesburg, South Africa, ⁹Madibeng Centre for Research, Brits, South Africa, ¹⁰Kenya Medical Research Institute, Nairobi, Kenya, ¹¹Wits Reproductive Health and HIV Institute, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa

Background: Women and girls in need of contraception are at risk of both pregnancy and sexually transmitted infections (STI), yet minimal data are available regarding STI risk with modern contraceptive methods, and no data are available from randomized controlled trials.
Methods: Women aged 16-35 years were randomized to injectable depot medroxyprogesterone acetate (DMPA-IM), copper intrauterine device (IUD), or levonorgestrel (LNG) implant and were followed for up to 18 months. Nucleic acid amplification testing for Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) was conducted at baseline, final visit and at interim visits per clinical indication; treatment was provided based on syndromes and laboratory results. We estimated pairwise NG and CT prevalence ratios (PR) at the final visit using intention-to-treat (ITT) and best achievable use (BAU, i.e., excluding women who did not initiate their randomized method at enrollment and censoring at time of randomized method discontinuation) analytic approaches. The ITT model included adjustment for site; confounders resulting in a ≥10% change in PR estimates were retained in BAC models.
Results: NG and CT prevalence at baseline were 4.7% and 18.2% respectively, with comparable prevalence by randomized arm and higher prevalence among women ≤24 years of age vs. >24 years. Follow-up time (mean 15.7 months) and method continuation were high (81.3% used their assigned method throughout follow-up and completed the final visit) and comparable across randomized arms. STI prevalence at the final visit was high despite syndromic treatment during follow-up: NG 4.8% (95% CI 4.4, 5.4) and CT 15.4% (14.6, 16.2). Results of the ITT and BAU analyses were comparable:

	ITTa	ITTa	BAU	BAU
	NG PR (95% CI)	CT PR (95% CI)	NGb PR (95% CI)	CTc (PR (95% CI)
DMPA-IM vs. copper IUD	0.7 (0.5, 0.9)*	0.9 (0.8, 1.0)	0.7 (0.5, 0.9)*	0.9 (0.7, 1.0)
DMPA-IM vs. LNG implant	0.8 (0.6, 1.0)	0.8 (0.7, 0.9)*	0.8 (0.6, 1.0)	0.8 (0.7, 0.9)*
Copper IUD vs. LNG implant	1.2 (0.9, 1.5)	0.9 (0.8, 1.0)	1.1 (0.9, 1.5)	0.9 (0.8, 1.0)
Prevalence ratio (PR); intention to treat (ITT); best achievable use (BAC) *significant at p≤0.05 level; aadjusted for study site; badjusted for study site, HIV status at final visit and pelvic exam number (total); cadjusted for study site, age group (≤24 vs. >24) and CT status at screening

[Relative risks of sexually transmitted infections by contraceptive methods]

Conclusions: The consistency of ITT and BAU analysis results suggest that DMPA-IM users may have a lower risk of NG compared to copper IUD users and of CT compared to LNG implant users. However, any true decreased risk must be evaluated along with all potential risks and benefits of the contraceptive method.

Download the e-Poster