Frequency of Condomless Sex Among Women Using Reversible Methods of Contraception

A version of this article was originally published on Contemporary OB/GYN. This version has been lightly edited.

A substudy of the Evidence for Contraceptive Options and HIV Outcomes (ECHO) trial, which compared 3 highly effective, reversible methods of contraception, concluded that women assigned either a copper intrauterine device (Cu-IUD) or the levonorgestrel (LNG) implant may have had condomless sex more frequently than women assigned to intramuscular depo-medroxyprogesterone acetate (DMPA-IM).

“The ECHO trial informed participants during the informed consent process that the World Health Organization (WHO) had communicated there was evidence of a possible increased risk of HIV acquisition associated with DMPA, but that the evidence was uncertain,” said substudy principal investigator Jennifer Deese, PhD, who at the time of the research was a scientist at the nonprofit human development organization FHI 360 and a senior epidemiologist at RTI International.

The substudy, which was published in the journal AIDS and Behavior, evaluated postrandomization sexual behavior using an objective marker of condomless vaginal sex in a subset of 458 participants at 3 of the 12 sites: Cape Town, South Africa; Johannesburg, South Africa; and Kisumu, Kenya.

Participants reported a median of 10 vaginal sex acts per month in the 3 months prior to enrollment, 96.9% reported having had the same primary sex partner during that period, and 73.3% reported at least 1 instance of unprotected sex.

The investigators assessed the frequency of condomless vaginal sex from December 2015 to September 2017, measured by prostate specific antigen (PSA) detection in vaginal swabs, collected at 6 months and the participants’ final visit, along with the concordance of self-reported condomless vaginal sex with PSA detection.

PSA was detected less frequently in the DMPA-IM group than in the Cu-IUD or LNG implant group: 16% vs 21% vs 24%, respectively.

Results were not statistically significant in the unadjusted model, when accounting for prespecified multiple-testing criteria.

However, in the adjusted model, there were significant differences in PSA detection between the DMPA-IM and LNG-implant groups (odds ratio, 0.61; 95% CI, 0.40-0.94).

There was also similar moderate discordance between self-reported condomless vaginal sex and detection of PSA among the 3 groups.

Deese, now a director and clinical scientist of vaccines at Pfizer Inc, told Contemporary OB/GYN that the substudy was not designed to have direct clinical implications, “but rather to identify if there was any biological evidence to suggest that women randomized to different contraceptive methods had different sexual behavior, and thus potential differential exposure to HIV.”

Deese said it is not surprising to discover biological evidence that indicates women randomized to DMPA may have modified their sexual behavior, “due to concerns about HIV risk.”

But because the substudy represented only 6% of the total ECHO trial population, “the women may not be generalizable to the full ECHO trial population or the larger population of women represented by trial participants,” said Deese, an adjunct assistant professor of epidemiology at the University of North Carolina at Chapel Hill.

The substudy findings are consistent with the outcomes of numerous prior studies, according to Deese, indicating that studies relying on self-reported sexual behavior “must be interpreted with caution, given continuing evidence of poor validity of these measures across studies in different settings.”

Instead, when feasible, objective markers of sexual exposure should be incorporated to support results interpretation, Deese noted.

Reference

Deese J, Chen PL, Gao X, et al. Post-randomization differences in condomless vaginal sex among women randomized to intramuscular depot medroxyprogesterone acetate injections, a copper intrauterine device or a levonorgestrel implant in the ECHO trial. AIDS Behav. Published online November 11, 2022. doi:10.1007/s10461-022-03834-y