Meta-analysis

Meta-analysis is a statistical method that combines results from multiple studies to estimate an overall effect and assess variation among their findings.

Effect size: A quantitative measure of the magnitude of a difference, association, or treatment effect. Meta-analysis combines comparable effect-size estimates across studies.
Systematic review: A structured review that uses explicit methods to identify, appraise, and synthesize research on a question. It typically establishes the study set that a meta-analysis statistically combines.
Evidence-based medicine: Clinical practice that integrates research evidence with professional expertise and patient values. Meta-analyses often summarize treatment evidence used in clinical decisions.
Narrative review: A literature review that summarizes research without a formal, reproducible systematic synthesis method. It contrasts with meta-analysis by integrating findings qualitatively rather than pooling estimates.
Inverse-variance weighting: A method that gives estimates with lower variance greater weight when combining them. It often determines how strongly each study contributes to the pooled estimate.
PICO framework: A framework specifying a research question through population, intervention, comparison, and outcome. It helps define which studies address a sufficiently shared question for synthesis.
Network meta-analysis: A method comparing multiple interventions by combining direct and indirect evidence across studies. It extends pairwise synthesis when trials connect several treatments.
Vote counting: A synthesis approach that counts studies with statistically significant results in each direction. It ignores effect magnitudes and precision, unlike quantitative pooling.
Fixed-effect model: A meta-analytic model that assumes all included studies estimate one common true effect. It differs from random-effects analysis in how it treats variation between studies.
Cochrane risk-of-bias tool: A framework for assessing risks of bias in randomized trials. Study limitations can distort pooled estimates even when statistical synthesis is correct.

Linked from 30 pages