Systematic reviews and meta-analysis
Synthesising a whole literature rather than trusting a single study.
A systematic review answers a defined question by searching the literature to a pre-specified protocol, applying explicit inclusion criteria, and appraising study quality — the aim being to remove the selection bias that makes narrative reviews unreliable. A meta-analysis goes further and pools results statistically for a combined estimate.
Their value is that single studies mislead. Effects are exaggerated in small trials, and publication bias means positive results are likelier to appear at all. Pooling counteracts both — which is why the systematic reviews of growth-factor-loaded scaffolds and of GLP-1 agonists and body composition carry more weight than any individual paper within them.
They are not immune to their inputs. Pooling heterogeneous studies produces a precise-looking number that may mean little — the "garbage in, garbage out" objection — and heterogeneity statistics deserve as much attention as the headline estimate. They sit at the top of the hierarchy in Peer review and the evidence hierarchy, but only when the underlying studies support it.
Related articles
- MethodsPeer review and the evidence hierarchyWhat peer review does, what it does not do, and how evidence is ranked.
- MethodsReproducibility in biomedical researchWhy a substantial share of published findings do not replicate — and what reduces the risk.
- MethodsClinical trial phasesWhat each stage of clinical testing is actually designed to answer.
- MethodsHow to read a research paperA practical order of operations for judging a study without taking its abstract at face value.
- MethodsPreclinical modelsWhat animal models can and cannot tell you about human regeneration.
- MethodsBench to bedsideThe pathway from laboratory observation to clinical use, and where it usually breaks.