Principles of Cancer Epidemiology

A clinical epidemiological study must address a research question, which can be explanatory or descriptive and which can be categorized as etiological, diagnostic, or prognostic. The main outcome measures of cancer epidemiology studies are prevalence, incidence, mortality, time to event, and quality of life. Surrogate endpoints are sometimes used instead of clinical endpoints but are controversial. There are many measures of association, which include relative risk, hazard ratio, odds ratio, differences in risk, and differences in mean score. The precision of estimates is described by p values and confidence intervals. Inaccuracies can be caused by various forms of bias, such as confounding, selection bias, and information bias. There are several research designs, which include case series, cross-sectional studies, cohort studies, randomized controlled trials, case-control studies, pilot studies, and systematic reviews.