Pooled incidence of continuation and pregnancy rates of four contraceptive methods in young women: a meta-analysis.

BACKGROUND: Continuation rates of contraceptive methods in young women vary among studies, and there is scarce data regarding the pregnancy rate in this population. METHODS: Four independently systematic searches were performed in PUBMED, EMBASE, LILACS, and Cochrane databases from inception until January 2021 for oral contraceptive pill (OCP), copper IUD, levonorgestrel intrauterine system (LNG-IUS), and subdermal implant. Inclusion criteria were observational or RCT studies that reported continuation for at least 12 months and/or pregnancy rate of these contraceptives methods in girls aged 22 years old or younger. Two authors extracted data from the study design and the outcomes. Pooled proportions of each method were applied using the inverse variance in all calculations with LOGIT transformation, using the random-effects model. Cochrane collaboration tool and New Castle-Ottawa were used to assess the quality and bias of all included studies. GRADE criteria evaluated the quality of evidence. RESULTS: Continuation rate for OCP was 51% (95%CI 34%-68%), while for cooper IUD was 77% (95%CI 74%-80%), LNG-IUS 84% (95%CI 80%-87%), and implant 85% (95%CI 81%-88%). The pooled estimated pregnancy rate for OCP was 11% (95%CI 6%-20%), while for cooper IUD was 5% (95%CI 3%-7%), LNG-IUS 1.6% (95%CI 1.2%-2.3%), and implant 1.8% (95%CI 0.4%-8.4%). CONCLUSION: Long-acting contraceptive methods presented higher continuation rates and lower pregnancy rates when compared to OCPs.