Menstrual bleeding and spotting with the Levonorgestrel Intrauterine System (52 mg) during the first-year after insertion: a systematic review and meta-analysis

Background

Changes in menstrual bleeding concern many users of the 52 mg Levonorgestrel Intrauterine System. Prescribing information for Levonorgestrel Intrauterine System devices describe an overall decrease in bleeding and spotting days over time; however, estimates derived from a variety of existing clinical data are currently unavailable.

Objective

The objective of the study was to systematically calculate the mean days of bleeding-only, spotting-only, and bleeding and/or spotting experienced by a population of reproductive-aged Levonorgestrel Intrauterine System users with normal regular menses prior to insertion during the first year of use.

Data Sources

We identified clinical trials, including randomized controlled trials and randomized comparative trials, as well as cohort studies published in English between January 1970 and November 2018 through searching 12 biomedical and scientific literature databases including MEDLINE and ClinicalTrials.gov.

Study Eligibility Criteria

We considered studies that reported data on Levonorgestrel Intrauterine System devices releasing 20 μg of levonorgestrel per day, collected daily menstrual bleeding data for at least 90 consecutive days, defined bleeding and spotting per World Health Organization standards and evaluated participants with normal regular menses prior to insertion.

Study Appraisal and Synthesis Methods

We assessed study quality using established guidelines. Two reviewers independently conducted all review stages and rated the quality of evidence for each article; any disagreements were resolved by a third. Where possible, we pooled data using a random-effects model.

Results

Among 3403 potentially relevant studies, we included 7 in our meta-analysis. We calculated the mean days of bleeding-only, spotting-only, and bleeding and/or spotting for the first four 90 day intervals after Levonorgestrel Intrauterine System insertion. Combined menstrual bleeding and/or spotting days gradually decreased throughout the first year, from 35.6 days (95% confidence interval, 32.2–39.1) during the first 90 day interval to 19.1 (95% confidence interval, 16.6–21.5), 14.2 (95% confidence interval, 11.7–16.8), and 11.7 days (95% confidence interval, 9.7–13.7) in the second, third, and fourth intervals. Measures for bleeding-only and spotting-only days similarly decreased throughout the first year, with the greatest decreases occurring between the first and second intervals.

Conclusion

Our study provides 90 day reference period measures that characterize menstrual patterns for Levonorgestrel Intrauterine System users with normal regular menses prior to insertion during the first year of use. Our findings provide broader generalizability and more detail than patterns described in the prescribing information. These findings quantify an overall decrease in menstrual bleeding days with longer duration of use, with the greatest decrease occurring between months 3 and 6. Accurately establishing expectations with the Levonorgestrel Intrauterine System may improve informed selection and decrease discontinuation.