Background: This study investigated whether gestodene-containing oral contraceptives (OCs) carry a higher risk of venous thromboembolism (VTE) than OCs containing progestins other than desogestrel and gestodene. The study was conducted based on the hypothesis that the biases and confounding factors that were present initially after the introduction of new so-called "third-generation" OCs (i.e., those containing desogestrel and gestodene) in the 1990s, which likely contributed to the alleged increased risk of VTE, may have vanished after 10 years.
Study design: This was a matched case-control study using data identified for women (aged 15-49 years) with suspected or diagnosed VTE (deep vein thrombosis or pulmonary embolism) that occurred between January 2002 and February 2006 in Austria. All VTE cases were validated by an attending/relevant physician(s), a detailed review of medical records and patient-completed questionnaires. Data were analyzed using an unconditional logistic regression model with adjustment for relevant confounders.
Results: Overall, 451 VTE cases and 1,920 controls without VTE were identified. The adjusted odds ratios for confirmed VTE with OC use versus nonuse were: 3.39 (95% CI 2.36-4.87) for OCs containing gestodene and 3.14 (2.1-4.47) for OCs containing progestins other than desogestrel and gestodene. Adjusted odds ratios for a head-to-head comparison of OCs containing gestodene versus OCs containing progestins other than desogestrel and gestodene were: 0.99 (0.68-1.45) for all cases; 1.01 (0.69-1.47) for confirmed cases and 1.11 (0.73-1.69) for confirmed and idiopathic VTE cases, respectively.
Conclusion: The risk of VTE is not elevated in users of gestodene-containing OCs relative to users of OCs containing progestins other than desogestrel and gestodene. Our study supports the view that (i) the majority of previous results may be explained by differences in the user populations of so-called "third-generation" OCs (containing desogestrel and gestodene) and "second-generation" OCs (containing progestins other than desogestrel and gestodene) that were present shortly after market introduction of gestodene-containing OCs and that (ii) these differences seem to have disappeared over time.