Sexual activity and cardiac risk: overview of epidemiologic research methods

The goal of clinical research is to differentiate among the different possible types of associations between the factors under study: no association, artifactual association, indirect association, and causal association. This article is intended as a brief overview of the research methods used in clinical research to differentiate these types of associations. It begins with an overview of the scientific method and the types of errors that can be made (i.e., random error, confounding, and bias). It then discusses the criteria supporting the causal nature of an association (i.e., how can one tell when an association is causal?). The criteria reviewed are (1) coherence with existing information; (2) time sequence; (3) specificity; (4) consistency; and (5) strength. Included within the last are quantitative strength, dose–response relation, and study design. Finally, the article provides a very brief outline of the available research designs, i.e., experimental studies, cohort studies, case–control studies, analyses of secular trends, case series, and case reports, and the advantages and disadvantages associated with each.