Bias versus causality : interpreting recent evidence of association between endometriosis and ovarian cancer

Objective: To establish whether causality versus biases may explain the association of endometriosis and ovarian cancer. Design: We employ the nine criteria proposed by Austin Bradford Hill, which still stand as foundation milestones for causal inference. Setting: University-based clinic. Patient(s): None. Intervention(s): Search strategies included online searching of the MEDLINE database and hand searching of relevant publications and reviews. Additional reports were collected by systematically reviewing all references from retrieved papers. Main Outcome Measure(s): Examination of the causal relationship between endometriosis and ovarian cancer. Result(s): Based on the present analysis, the criterion of strength has not been fulfilled. There were insufficient or mixed data for four criteria (biological gradient, biological plausibility, analogy, coherence), and four criteria were fulfilled (consistency, temporality, specificity, experimental evidence in animal model). Conclusion(s): A causal relationship between endometriosis and specific histotypes of ovarian cancer should be recognized, but the low magnitude of the risk observed is consistent with the view that ectopic endometrium undergoes malignant transformation with a frequency similar to its eutopic counterpart.