Objective: To estimate whether doxycycline, a matrix metalloproteinase inhibitor, would decrease unscheduled bleeding associated with initiation of a continuous oral contraceptive pill.
Methods: Participants initiating a continuous oral contraceptive pill (20 micrograms of ethinyl estradiol/90 micrograms of levonorgestrel) were randomly assigned to receive either doxycycline (100 mg orally twice daily) or placebo taken for 5 days at the onset of each bleeding or spotting episode during the first 84 days of the study period. For the final 28 days of the study, participants were observed on the oral contraceptive pill alone. The primary outcome was the number of bleeding and spotting days. A sample size of 66 (33 in each arm) was calculated to detect a 50% reduction in bleeding (beta=0.80, alpha=0.05) and accounted for a 30% dropout rate.
Results: Sixty-six women were randomly assinged (33 in each study group). There were no significant differences during the 84-day treatment in bleeding or spotting days (doxycycline [mean {standard error}, placebo, P=.32) or the length of the longest bleeding or spotting episode (doxycycline, placebo, P=.70) between study groups. Similarly, no significant differences in bleeding patterns existed between groups during the final 28 days.
Conclusion: Doxycycline, administered once bleeding has started, does not decrease unscheduled bleeding or shorten episodes of unscheduled bleeding in continuous oral contraceptive pill users.
Level of evidence: I.
Trial registration: ClinicalTrials.gov NCT00480532.