The risk of unintended pregnancies in users of the contraceptive patch compared to users of oral contraceptives in the UK General Practice Research Database

BackgroundThe transdermal contraceptive patch EVRA has been marketed since 2002 as an alternative to oral contraceptives (OCs), but large follow-up studies of effectiveness are lacking.Study DesignWe used a cohort plus a nested case-control study to estimate rates and hazard ratios of unintended pregnancy among users of different hormonal contraceptives (HC) and to evaluate the effects of various potential predictors on unintended pregnancy.ResultsRates of unintended pregnancy were all low (overall rate 0.17 per 100 women-years), though they were slightly higher in users of progestogen-only OCs and EVRA compared to users of second-generation OCs. The hazard ratios for unintended pregnancy were also higher for users of progestogen-only OCs and the patch compared to users of second-generation OCs, although these estimates may be confounded by compliance. Rates of unintended pregnancy were inversely associated with age and duration of HC use as well as with other predictors of pregnancy but not body mass index or use of antibiotics.ConclusionThe rates of unintended pregnancy were low for all study HCs, and while the data on unintended pregnancy in EVRA users compared to OC users are sufficient to rule out major differences in the rate of unintended pregnancy, they are not sufficient to rule out modest ones.