Vaginal bleeding decreases over time with a levonorgestrel intrauterine system

Référence

Maldonado LY, Sergison JE, Gao X, Hubacher D. Menstrual bleeding and spotting with the levonorgestrel intrauterine system (52 mg): a systematic review and meta-analysis. Am J Obstet Gynecol 2020;222(5):451-468.e9.

Sommaire

The authors of this systematic review and meta-analysis conducted it for the stated purpose of providing information to help clinicians counsel patients about what to expect regarding vaginal bleeding after the placement of a LNG-IUS. The authors identified 7 studies—4 randomized controlled trials (RCTs), 2 cohort studies, and 1 re-analysis of RCT data—that met their inclusion criteria with a total of 5098 participants from 10 countries. Included studies reported outcomes of interest for women with a history of normal menstrual periods for at least 90 days after the placement of LNG-IUS (52 mg; Mirena or Liletta) according to bleeding diaries. Bleeding outcomes were the number of bleeding and/or spotting days based on World Health Organization definitions. The authors excluded studies if the study population was women with heavy or prolonged menstrual periods, who were actively breastfeeding, who used long-acting contraception methods within 6 months prior to placement, or who had anatomic pathologies (such as fibroids or cancer). The authors extracted mean or median days of bleeding only, spotting only, and bleeding and/or spotting in 90-day intervals for the year after placement. They found that bleeding decreased markedly after the first 90-day interval. For bleeding and/or spotting, the results over succeeding periods were 35.6 days (95% CI 32.2 - 39.1), 19.1 days (16.1 - 21.5), 14.2 days (11.7 - 16.8), and 11.7 days (9.7 - 13.7). The results for bleeding-only days were 13.3 days (10.8 - 15.7), 7.8 days (6.5 - 9.1), 6.2 days (4.9 - 7.5), and 5.0 days (3.9 - 6.1), respectively, for the four 90-day periods. The authors conducted several sensitivity analyses to account for heterogeneity between studies, which did not materially change the findings.