Deep vein thrombosis and the oestrogen content in oral contraceptives. An epidemiological analysis

Epidemiological studies have pointed to a correlation between the oestrogen content of oral contraceptives and the risk of deep vein thrombosis (DVT). The correlation has been strongest in studies which partially consisted of adverse drug reaction reports to the Swedish Adverse Drug Reaction Advisory Committee (SADRAC). The present study analyzes the epidemiological basis of the adverse drug reaction reports on DVT in women on oral contraceptives to SADRAC. It verifies the reported correlation between the oestrogen content of the pills and the risk of DVT but it also demonstrates that this correlation probably was secondary to differences in the diagnostic standard of DVT, to differences in reporting policies to SADRAC and to an age difference between women on low-oestrogen-pills and those on high-oestrogen pills and is thus due to bias. It is concluded that adverse drug reaction reporting on oral contraceptives has been very unreliable, for which reason it cannot support any epidemiological conclusion concerning the relative thrombogenicity of high-oestrogen pills compared with that of low-oestrogen pills.