Evaluating the relevance of epidemiological studies for the single patient: How can evidence of mechanisms help?

As a general conclusion, we might say that evidence of mechanisms is most reliable for discarding inadequate interventions. It can provide a key contribution to predicting that an intervention that was significantly effective and safe in the study population might not be so for the patient. However, establishing that an intervention will be as effective and safe in the patient as it was in the study population, based on relevant similarities between general and local mechanisms, requires caution. An open question, which exceeds the scope of this commentary, is which (and how) particularities of evidence of mechanisms contribute to this divergence. Finally, we have shown how insights from a growing body of conceptual and methodological research that addresses inferences from population‐level evidence to particular cases can help guide clinical thinking. We suggest that the philosophy of medicine should have a larger place in medical education, medical journals and medical research.