Objective: To explore the usefulness of epidemiological data to guide clinical practice by seeking an answer to the question “What is the risk of cardiovascular disease among users of currently available, low dose, combined oral contraceptives who are aged less than 35 years, do not smoke, and do not have a medical condition known to increase the risk of vascular disease?”
Design: Review of all relevant published studies identified from the library of references held by Royal College of General Practitioners’ Manchester Research Unit, checking of reference lists of identified studies, and Medline search.
Main outcome measures: Identification of methodologically sound studies able to address the specific clinical question.
Results: Our literature search identified 74 papers about the relation between current use of combined oral contraceptives and cardiovascular disease: 23 papers reporting risk of venous thromboembolism, 22 on ischaemic stroke, 13 on haemorrhagic stroke or subarachnoid haemorrhage, 13 on all stroke, and 33 on myocardial infarction. Only five papers provided information that directly addressed our clinical question; all related to the risk of venous thromboembolism. Fourteen of the discarded papers probably had the potential to answer our clinical question.
Conclusions: Much of the epidemiological data about the risk of cardiovascular disease in users of combined oral contraceptives is not useful to clinicians. Some of the discarded data could be made more useful to clinicians by reanalysis. This situation is unlikely to be unique to use of contraceptives.
Key messages Epidemiological studies investigate overall, average effects within populations, but clinicians need information about specific risks and benefits faced by the individual patients consulting them We explored the clinical usefulness of epidemiological data in defining the risk of cardiovascular disease associated with currently available low dose combined oral contraceptives for young, healthy women who do not smoke Our literature search identified 74 papers about the subject, but only five provided information that directly addressed our clinical question Fourteen other studies probably had the potential to answer our question if their data were reanalysed Clinicians need to be cautious when extrapolating results from epidemiological studies to guide their clinical practice