casecontrol

Case-control study, also known as case-control design, is an observational study used to identify factors associated with a particular outcome, most commonly a disease. It starts with people who already have the outcome (cases) and compares them with people who do not (controls) to examine prior exposure history. The aim is to determine whether exposure is more common among cases, suggesting a possible association.

Selection: Cases are individuals with the disease or outcome of interest; controls are chosen from the same source population that gave rise to the cases, and should be at risk of developing the outcome. Matching on confounders (e.g., age, sex) is common to reduce confounding. Data on past exposures are collected retrospectively, often via interviews, records, or questionnaires.

Analysis: The primary measure is the odds ratio, estimating how the odds of exposure differ between cases and controls. Conditional logistic regression is used when matching; unmatched designs use standard logistic regression. A case-control study is efficient for rare diseases and can evaluate multiple exposures. It is often used in etiologic research and can be conducted quickly and at lower cost than cohort studies.

Limitations: Susceptible to biases, notably recall bias and selection bias in control selection. Confounding can distort associations; temporality can be difficult to establish, so causality is harder to infer. Not able to provide incidence or risk directly, except with certain sampling methods in nested designs.

Variants: Nested case-control studies within a cohort, population-based or hospital-based designs, matched or unmatched. Nested designs use controls drawn from a defined cohort and can employ incidence-density sampling to estimate rate ratios.