Risk of endometrial cancer in relation to use of combined oral contraceptives. A practitioner's guide to meta-analysis

A review of the English-language literature related to the association between use of combined oral contraceptives (OCs) and endometrial cancer is used to illustrate the principles of meta-analysis. Guidelines are presented on defining the problem, finding the relevant studies, abstracting the pertinent information, imputing values to duration of use and recency, and analyzing and interpreting the resulting data. The meta-analysis of the epidemiologic literature indicates that combined OC use is associated with a reduced risk of endometrial cancer. There also appears to be a residual protective effect that persists long after discontinuation of OC use. The absolute reduction in risk is inconsequential at the individual level, however. A US woman's probability of remaining free of endometrial cancer through age 74 years increases from 97.6% in the absence of OC use to 98.6% after 12 years of OC use. Although meta-analyses can provide a quantitative perspective on findings in the literature, they cannot distinguish between correlation and causation.