Drospirenone-containing combined oral contraceptives and the risk of arterial thrombosis: a population-based nested case-control study

Objective To compare the rate of arterial thromboembolism (ATE) of drospirenone-containing COCs to that of levonorgestrel-containing COCs. Design Population-based cohort study. Setting United Kingdom's Clinical Practice Research Datalink (CPRD), which contains clinical records for >11 million patients. Population Women aged 16-45 years prescribed a drospirenone- or levonorgestrel-containing COC between May 2002 and June 2012. Methods We conducted nested case-control analyses using risk set sampling to randomly select up to 10 controls for each ATE case, matched on age, cohort entry year, CPRD registration year, COC user type (first-time ever, new, switcher, or prevalent users), duration of COC use, duration of progestin-only or implantable contraceptive use, pre-cohort entry duration of drospirenone and levonorgestrel use, and duration of follow up. Main outcome measures We used conditional logistic regression to estimate hazard ratios and 95% confidence intervals (CIs), adjusted for high-dimensional propensity scores. Results Our cohort included 339 743 women followed over a mean 4.4 years, during which 228 ATE cases occurred: 37 myocardial infarctions, 170 strokes, and 21 other ATEs; overall rate: 1.5 events per 10 000 person-years (PYs). After adjusting for potential confounders, the hazard ratio for ATE with current use of drospirenone-containing COCs versus current use of levonorgestrel-containing COCs was 0.89 (95% CI 0.35, 2.28), corresponding to a rate difference of -0.16 events per 10 000 PYs. Conclusions The overall rate of ATE in this population is low regardless of which COC was taken. We found little evidence of a difference in the rate of ATE with drospirenone-versus levonorgestrel-containing COCs.