Systematic Review and Meta-Analysis
Systematic review and meta-analysis is the area of epidemiology and evidence-based health care concerned with locating, appraising, and combining the results of multiple primary studies to answer a defined research question. A systematic review applies an explicit, reproducible protocol to find and evaluate all relevant evidence; a meta-analysis is the optional statistical step that pools the quantitative results into a single summary estimate.
Definition
Systematic review and meta-analysis comprises the set of structured, protocol-driven methods for identifying, critically appraising, and synthesising the findings of primary studies, where the systematic review is the qualitative synthesis and the meta-analysis is the quantitative pooling of effect estimates.
Scope
This area orients the reader to the family of methods used to synthesise research evidence: the formulation of a focused question, comprehensive and reproducible searching, study selection and risk-of-bias appraisal, statistical pooling, the assessment of heterogeneity and publication bias, and the rating of the overall certainty of evidence. It treats these as methodological and reporting topics, not as clinical instructions.
Sub-topics
Core questions
- What does all the available evidence say about a defined question, once it has been gathered systematically and appraised?
- Can the results of separate studies be combined into a single summary estimate, and how consistent are they?
- How much can the synthesised body of evidence be trusted?
Key concepts
- Protocol-driven, reproducible methodology
- Comprehensive literature searching
- Risk-of-bias appraisal
- Effect-size pooling (meta-analysis)
- Statistical heterogeneity
- Publication and reporting bias
- Certainty-of-evidence rating (GRADE)
- Transparent reporting (PRISMA)
Mechanisms
A systematic review reduces the bias and chance that affect any single narrative summary by specifying, in advance, how studies will be sought, selected, appraised, and combined. Comprehensive searching aims to capture the whole relevant evidence base; explicit eligibility criteria and duplicate study selection limit selective inclusion; risk-of-bias appraisal weighs the trustworthiness of each included study. Where studies are similar enough, meta-analysis combines their effect estimates, weighting each by its precision, to produce a more precise summary than any single study and to characterise how much the true effect varies across studies. Tools such as PRISMA standardise the reporting of this process so it can be scrutinised and reproduced, and frameworks such as GRADE rate how much confidence the synthesised result warrants.
Clinical relevance
Well-conducted systematic reviews and meta-analyses sit near the top of conventional evidence hierarchies and inform clinical guidelines, health-technology assessment, and policy. Understanding how they are built and where they can mislead is central to evidence appraisal in the health sciences. This entry describes how aggregate evidence is generated and judged; it is reference material for evaluating evidence, not guidance for individual diagnosis or treatment.
Epidemiology
Synthesis methods are used across clinical medicine, public health, and the social sciences. The Cochrane Collaboration, founded in the 1990s, organised the large-scale production of systematic reviews of health-care interventions, and reporting standards such as PRISMA are now expected by most biomedical journals. The number of published systematic reviews and meta-analyses has grown rapidly, which has itself raised concerns about redundancy and quality.
Evidence & guidelines
Reporting of systematic reviews and meta-analyses is governed by the PRISMA 2020 statement (Page et al., 2021), with PRISMA-P covering protocols. The certainty of the synthesised evidence is commonly rated with the GRADE approach (Guyatt et al., 2008). These are reporting and appraisal frameworks rather than treatment recommendations.
History
The idea of statistically combining studies dates to early twentieth-century statistics, and Gene Glass coined the term meta-analysis in 1976 in the context of educational research. In medicine, the random-effects model of DerSimonian and Laird (1986) became the standard pooling method, and Archie Cochrane's call for systematic summaries of trials inspired the founding of the Cochrane Collaboration in 1993. Reporting standards followed: QUOROM, then PRISMA (2009), updated as PRISMA 2020. Frameworks for rating heterogeneity, publication bias, and certainty of evidence matured in parallel.
Debates
- Have systematic reviews and meta-analyses been overproduced?
- The rapid growth in published reviews has prompted concern that many are redundant, methodologically weak, or conflicted, raising questions about how much new synthesis adds and how to prioritise it.
Key figures
- Archie Cochrane
- Iain Chalmers
- Rebecca DerSimonian
- Nan Laird
- Julian Higgins
- David Moher
- Gordon Guyatt
Related topics
Seminal works
- dersimonian-laird-1986
- page-2021-prisma
- guyatt-2008-grade
Frequently asked questions
- What is the difference between a systematic review and a meta-analysis?
- A systematic review is the whole process of finding, appraising, and summarising studies under a pre-specified protocol; a meta-analysis is the optional statistical step that pools the quantitative results into a single summary estimate. Every meta-analysis should rest on a systematic review, but a systematic review need not include a meta-analysis.
- Why are systematic reviews considered high-level evidence?
- Because, when done well, they gather and appraise the whole relevant evidence base in a reproducible way rather than relying on selectively cited studies, and they can yield more precise and less biased summaries than any single study. Their reliability still depends on the quality of the included studies and the rigour of the methods.