What Constitutes “Proof” in Epidemiologic Studies?
There has been vigorous philosophic and scientific debate over what constitutes “proof” when epidemiologic and other evidence is evaluated with respect to policy. Cohort and case control studies are, in principle, less definitive than are randomized clinical trials because they cannot control completely for unmeasured or incompletely measured risk factors that may be associated with both the exposure and disease of interest. However, the information from observational studies is often considered sufficient for public health action if chance, bias, or confounding can be reasonably excluded as alternative explanations for the findings. Reliance on observational data as a basis for policy is inevitable in situations where randomized trials are unethical or unfeasible. The criteria frequently used to make a judgment of causation were first proposed in the 1964 Surgeon General’s report on smoking and health,56 subsequently expanded by Sir Austin Bradford Hill,57 and later incorporated by or
