Systematic Review and Evidence Synthesis
Systematic review and evidence synthesis is the family of methods that locate, appraise, and combine the findings of multiple studies on a question using an explicit, reproducible protocol. Rather than relying on a single trial or an unstructured narrative, these methods aim to minimise bias and give a transparent, defensible summary of what the available research collectively shows, making them a cornerstone of evidence-based practice and health technology assessment.
Definition
A systematic review is a review that uses explicit, pre-specified and reproducible methods to identify, select, appraise, and synthesise all the studies relevant to a defined question; evidence synthesis is the wider set of approaches that combine study results quantitatively or qualitatively to inform conclusions.
Scope
This area orients the reader to the synthesis methods used to inform decisions in health policy, systems, and practice: the prototypical systematic review with meta-analysis, the broader landscape of evidence-synthesis and review typologies, exploratory scoping reviews, comparison of many treatments through network meta-analysis, and the publication-bias problem that threatens any synthesis. It is a reference overview; the detailed methodology lives in the child topics.
Sub-topics
Core questions
- What review or synthesis design fits the question being asked?
- How are studies searched for, screened, and selected without bias?
- How is the risk of bias in included studies and across the body of evidence assessed?
- When is it appropriate to pool results statistically versus synthesise narratively?
- How confident can decision-makers be in the synthesised estimate?
Key concepts
- Explicit, reproducible protocol
- Comprehensive and reproducible searching
- Risk-of-bias assessment
- Quantitative pooling (meta-analysis) versus narrative synthesis
- Certainty of evidence (GRADE)
- Review typology and design choice
- Reporting standards (PRISMA)
Mechanisms
Evidence synthesis works by replacing the implicit judgements of a traditional review with documented, repeatable steps: a protocol defines the question and eligibility criteria in advance; a structured search retrieves candidate studies; independent screening and data extraction reduce reviewer error; risk-of-bias tools appraise each study; and, where studies are sufficiently similar, results are pooled statistically. The certainty of the resulting body of evidence is then rated so that the synthesis communicates not only an effect estimate but how much trust it warrants. The Cochrane Handbook codifies these steps for intervention reviews, PRISMA standardises their reporting, and GRADE structures the certainty rating (higgins-handbook-2019; page-2021-prisma; guyatt-2008-grade).
Clinical relevance
Syntheses of this kind sit at the top of conventional evidence hierarchies and underpin clinical guidelines, formularies, and health technology assessments. Understanding how a synthesis was conducted is part of appraising the evidence behind a recommendation; the methods describe how collective evidence is generated and graded, and are not themselves instructions for treating an individual patient.
Evidence & guidelines
Reporting and conduct in this area are governed by widely adopted standards. PRISMA 2020 sets out what a systematic review report should contain, the Cochrane Handbook describes accepted methods for intervention reviews, and GRADE provides a structured approach to rating certainty of evidence and strength of recommendations. Review typologies clarify which design answers which kind of question (page-2021-prisma; higgins-handbook-2019; guyatt-2008-grade; grant-booth-2009; gough-2012).
History
Quantitative combination of studies has roots in early statistics, but organised evidence synthesis grew rapidly from the late twentieth century with the founding of the Cochrane Collaboration in 1993 and the spread of the systematic review as a distinct research product. Reporting standards followed: QUOROM and then PRISMA for systematic reviews, and the GRADE framework for certainty, with PRISMA updated in 2021 to reflect methodological advances (page-2021-prisma; higgins-handbook-2019).
Debates
- Are systematic reviews being produced too readily and redundantly?
- The growth of synthesis methods has raised concern about overlapping, low-quality, or unnecessary reviews; choosing the right review type for the question and registering protocols are proposed safeguards.
Key figures
- Iain Chalmers
- Cynthia Mulrow
- David Moher
- Matthew Page
- Gordon Guyatt
- Julian Higgins
Related topics
Seminal works
- page-2021-prisma
- higgins-handbook-2019
- guyatt-2008-grade
Frequently asked questions
- What is the difference between a systematic review and a meta-analysis?
- A systematic review is the whole structured process of finding, appraising, and synthesising studies; a meta-analysis is the optional statistical step that pools results numerically. A systematic review may or may not include a meta-analysis.
- Why are systematic reviews considered high-level evidence?
- Because they bring together all eligible studies using explicit, reproducible methods that limit bias, they give a more complete and less arbitrary picture than any single study or unstructured review.