Low rates of pregnancy with three methods of contraception in the ECHO trial

Polly Clayden, HIV i-Base

Both perfect and typical use of all contraceptive methods – DMPA-IM, copper IUD, or LNG implant –  resulted in low pregnancy rates at 18 months in a secondary analysis of the ECHO trial. Typical copper IUD use was associated with statistically significant higher pregnancy risk compared to LNG implants.

The ECHO Trial looked at HIV incidence among 7829 women from sites in Eswatini, Kenya, South Africa and Zambia who were be randomised to one of the three methods.

The pregnancy analyses were performed for perfect use and typical use among 7710 women – these findings were presented at IAS 2019.

Baseline demographics and behavioral data were similar across the threegroupsof women: median age was 23 years, the majoritywomen were single and never married (80%), had some or complete secondary education (74%), a BMI <30kg/m2 (74%), and 1–2 living children (66%). Approximately half had used DMP-IM before (51%) compared with very few who used either LNG implant (6.4%) or copper IUD (0.8%). There was high prevalence of STIs: 18% chlamydia trachomatis and 4.7% gonorrhea.

There were 70 pregnancies during perfect use and and 85 during typical use.

Incidence pregnancy rates for perfect use at 18 months were: 0.61 per 100 woman-years (wy) for DMPA-IM (95% CI: 0.36 to 0.96), 1.06 for copper IUD (95% CI: 0.72 to 1.50) and 0.63 for LNG implants (95% CI: 0.39 to 0.96). But the copper IUD appeared to have a different slope: starting lower (to approximately 6 months) and rising relatively higher when compared to the other two methods.

In pairwise comparisons, there was no statistically significant difference in terms of the risk of pregnancy incidence between method pairs.

Typical use incidence rates were 0.87 per 100 wy for DMPA-IM (95% CI: 0.58 to1.25), 1.11 for copper IUD (95% CI: 0.77 to 1.54), and 0.63 for LNG implants (95% CI: 0.39 to 0.96). Typical use of copper IUD was associated with statistically significant higher risk of pregnancy compared to LNG implants: aHR 1.74  (95% CI: 1.01 to 2.99), p=0.044. The other comparisons did not reach significance.


Presenting author Dr Maricianah Onono noted that DMPA-IM pregnancy incidence with typical use was much lower than usually reported in routine settings. She suggested that this might be attributed to proactive tracing by research staff.

One suggestion during questions after the presentation was, as the copper IUD appeared to be working better up to 6 months, this would be a good time for a routine visit to check that it is in place. It would also be a good opportunity to check for HIV and STIs.


Onono M et al. Comparison of pregnancy incidence among African women in a randomised trial of intramuscular depomedroxyprogesterone acetate (DMPA-IM), the levonorgestrel (LNG) implant, and the copper intrauterine device (IUD).10th IAS Conference on HIV Science. Mexico City, Mexico (IAS 2019), 21–24 July 2019. Oral abstract MOAX0103LB.

Links to other websites are current at date of posting but not maintained.