Epidemiological study designs and principles of data analysis: A conceptually integrated suite of methods and techniques
Epidemiological studies are unified by their common goals and by their basis in defined populations. The case series (or register-based study) includes examination of trends in deaths, cancers, notifiable diseases, and hospitalizations. Case–control studies are analysed by comparing the exposure to risk factors in cases to those in controls. In a population studied at a specific time and place (a cross-sectional study), measurements can be made of disease, the factors which may cause disease, or both simultaneously. Cohort studies produce data on disease incidence and are especially good on associations between risk factors and disease outcomes. Trials compare treated and untreated populations and are used, primarily, for information on effectiveness of health interventions. Natural experiments, including Mendelian randomization studies, may provide causal evidence. The principles for the analysis of all studies are similar. The design and interpretation should be in the context of traditional, systematic, and meta-analytic reviews.