Oral contraceptives and venous thromboembolism

In October, 1995, the United Kingdom Committee on Safety of Medicines recommended that women should use oral contraceptives containing desogestrel or gestodene only if prepared to accept an increased risk of thromboembolism. This action was based on observational studies that indicated a 2-3-fold increase in the risk of thromboembolism when desogestrel and gestodene-containing contraceptives were compared to products with levonorgestrel. The fact that these studies point in the same direction is concerning, but it is possible that these observational studies, somewhat similar in design, are influenced by the same unrecognized biases. Furthermore, the case numbers are relatively small (20-40 cases), and, for example, preferential prescribing to women at greater risk could affect the results. It is difficult to reconcile the conclusions with the strong belief supported by good evidence that thrombosis is an estrogen dose-related complication, and that progestational agents have no impact on clotting parameters. In my view, these equivocal reports are not of sufficient strength to escape biases and to change our prescribing of oral contraceptives.