Epidemiologic assessment of the risks of oral contraception.

Relative risk is the epidemiologic measure of risk presented most often to indicate the benefits and hazards of oral contraception. This measure of risk is prone to distortion and, unlike the attributable risk measure, does not indicate the absolute probability of a hazardous or beneficial effect of oral contraception. Estimates of the risk of cardiovascular disease in oral contraceptive users quoted most often are derived from studies that are at least a decade old. Changes in the formulation of oral contraceptives and selection of healthy women for use of such contraceptives have probably reduced these risks, although precise estimates for currently marketed oral contraceptive formulations in women free of cardiovascular risk factors are not available. Even considering the probably inflated estimates of the risk of cardiovascular disease in oral contraceptive users derived from the older British studies, oral contraception compares favorably with such activities as automobile use in women less than age 35. The hazards of everyday living are the most appropriate framework for interpreting the risks of oral contraception.