Explainer
How systematic reviews and meta-analyses work (Cochrane and GRADE)
A systematic review tries to gather all the studies that meet pre-set criteria on one question, using explicit methods designed to reduce bias.[1] A meta-analysis is the statistical step that combines their results, which can improve precision but can also mislead if biases and differences between studies are ignored.[2][3] Cochrane reviews rate the certainty of each finding with GRADE as high, moderate, low or very low.[4]
This page explains how research is pooled. It is general information, not medical advice. A review describes average effects in the people studied, so talk to a clinician about how a finding applies to you.
Systematic reviews look at all the relevant studies on a question together. Cochrane says they were developed so decisions can be informed by an up-to-date and complete understanding of the evidence, because with the literature growing ever faster, individual decision makers cannot assess all the primary research themselves.[5]
What makes a review “systematic”
A systematic review tries to collect all the empirical evidence that meets pre-specified eligibility criteria to answer a specific question, using explicit methods chosen to minimise bias.[1] The key steps are:
- A protocol written in advance. Decisions made after seeing the results, such as dropping certain studies or changing the analysis, are highly susceptible to bias. Publishing the protocol and registering it on PROSPERO reduces duplication and promotes accountability.[6]
- A wide search. Cochrane reviews look for studies in several sources, including unpublished ones, select and appraise them against predefined criteria, then collect and synthesise the data systematically.[7]
- A qualified team without conflicts. Cochrane says review teams should combine subject and methods expertise and be free of potential conflicts of interest.[8]
- Transparent reporting. The PRISMA guideline, currently PRISMA 2020, sets out what a published review should report.[9]
A Cochrane review is one published in the Cochrane Database of Systematic Reviews. Faster “rapid reviews” streamline or skip some of these methods.[10]
Meta-analysis: combining the numbers
A meta-analysis statistically combines results from two or more separate studies.[2] Cochrane lists its advantages as greater precision, the ability to answer questions single studies did not ask, and a way to settle conflicting claims. It also warns that a meta-analysis can seriously mislead if study design, biases within studies, variation between studies and reporting biases are not carefully considered.[3]
Results are usually drawn as a forest plot. Each study is a block on a horizontal line. A bigger block means the study carries more weight, and the line shows its confidence interval.[11] To read the effect sizes, it helps to know how relative and absolute risk differ.[12]
When study results differ more than chance alone would explain, that is statistical heterogeneity.[13] A fixed-effect model assumes every study estimates exactly the same quantity, while a random-effects model assumes the true effects vary across studies.[14] Cochrane also asks authors to report confidence intervals with exact P values rather than lean on “statistically significant” thresholds.[15]
Rating the certainty: GRADE
Cochrane defines certainty as how confident one can be that an effect estimate is close to the quantity of interest.[16] Using the GRADE approach, reviewers rate each outcome as high, moderate, low or very low certainty. They judge five domains: risk of bias, inconsistency, indirectness, imprecision and publication bias.[4] Evidence from non-randomised studies is generally rated down two levels for confounding and selection bias.[17] It can be rated up for a large effect or a dose-response gradient.[18] The epidemiology explainer covers why.
Pooling many studies does not by itself make a finding high certainty: each outcome is still judged on those five domains.[4]
From review to recommendation
Cochrane tells review authors not to make recommendations about healthcare decisions.[19] The handbook says judgements about what to recommend are typically the domain of clinical practice guideline developers, though review authors can highlight actions that might fit particular values and preferences.[19] Guideline groups such as the US Preventive Services Task Force grade their recommendations by the certainty and size of the net benefit. See how medical guidelines are made.[20]
Read next
US Preventive Services Task Force (USPSTF)
The USPSTF sets the US cancer screening recommendations that shape free coverage. How it grades evidence, what it recommends, and its 2026 overhaul.
Questions readers ask
What is the difference between a systematic review and a meta-analysis?
A systematic review is the whole process of finding, appraising and synthesising all the evidence that meets pre-set criteria on a question. A meta-analysis is the statistical combination of results from two or more studies, and is often one part of a review.[1][2]
What is a Cochrane review?
It is a systematic review of health care or health policy research published in the Cochrane Database of Systematic Reviews. Types include intervention, diagnostic test accuracy and prognosis reviews, and rapid reviews.[10]
What does GRADE certainty mean?
GRADE rates how confident one can be that an effect estimate is close to the true value, on four levels from high to very low. Reviewers judge it on risk of bias, inconsistency, indirectness, imprecision and publication bias.[4][16]
How do I read a forest plot?
Each study appears as a block and a horizontal line. The block's area shows how much weight the study gets in the meta-analysis, and the line shows its confidence interval.[11]
Do systematic reviews tell you what to do?
Not directly. Cochrane tells review authors not to make recommendations about healthcare decisions, and says such judgements are typically the domain of clinical practice guideline developers.[19]
Sources
· 20 claims from 6 publishers
Each numbered claim is a statement we checked against the sources listed with it. Tap a number in the text to see its sources in place.
- [1]
A systematic review tries to collect all the empirical evidence that meets pre-specified eligibility criteria to answer a specific research question, using explicit methods chosen to minimise bias. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 1: Starting a review · Cochrane · Primary source · Section 1.1 (retrieved 2026-11-07)
- About Cochrane Reviews · Cochrane Library · Primary source · What is a systematic review? (retrieved 2026-11-07)
- [2]
A meta-analysis statistically combines the results of two or more separate studies on the same question. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- NCI Dictionary of Cancer Terms: meta analysis · National Cancer Institute · Primary source · definition (retrieved 2026-11-07)
- [3]
Cochrane lists greater precision, answering questions single studies did not ask, and settling conflicting claims as advantages of meta-analysis, but warns it can seriously mislead if study design, within-study biases, variation between studies and reporting biases are ignored. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- [4]
GRADE rates the certainty of a body of evidence for each outcome as high, moderate, low or very low, judged on five domains (risk of bias, inconsistency, indirectness, imprecision and publication bias). confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- [5]
Cochrane says systematic reviews were developed so that decisions can be informed by an up-to-date and complete understanding of the research, because with the literature growing ever faster, individual decision makers cannot assess all the primary research themselves. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 1: Starting a review · Cochrane · Primary source · Section 1.1 (retrieved 2026-11-07)
- [6]
Reviews pre-specify their questions and methods in a protocol, because decisions made after seeing the results, such as dropping studies or changing the analysis, are highly susceptible to bias; registering protocols on PROSPERO reduces duplication and promotes accountability. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 1: Starting a review · Cochrane · Primary source · Section 1.5 (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 1: Starting a review · Cochrane · Primary source · Section 1.5 (retrieved 2026-11-07)
- [7]
Cochrane reviews search several sources including unpublished ones, select and appraise studies against predefined criteria, collect data systematically and synthesise it. confirmedas of 2026-11-07
- About Cochrane Reviews · Cochrane Library · Primary source · Cochrane review methods (retrieved 2026-11-07)
- [8]
Cochrane says reviews should be done by a team with both subject and methods expertise whose members are free of potential conflicts of interest. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 1: Starting a review · Cochrane · Primary source · Section 1.3 (retrieved 2026-11-07)
- [9]
PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) is a reporting guideline for systematic reviews; its main version is PRISMA 2020. confirmedas of 2026-11-07
- PRISMA statement · PRISMA · Primary source · Home page (retrieved 2026-11-07)
- PRISMA statement · PRISMA · Primary source · Home page (retrieved 2026-11-07)
- [10]
A Cochrane review is a systematic review of health care or health policy research published in the Cochrane Database of Systematic Reviews; types include intervention, diagnostic test accuracy, prognosis and rapid reviews. confirmedas of 2026-11-07
- About Cochrane Reviews · Cochrane Library · Primary source · What is a Cochrane review? (retrieved 2026-11-07)
- About Cochrane Reviews · Cochrane Library · Primary source · Types of Cochrane review (retrieved 2026-11-07)
- About Cochrane Reviews · Cochrane Library · Primary source · Types of Cochrane review (retrieved 2026-11-07)
- [11]
Meta-analyses are usually shown as forest plots, in which each study is a block whose area shows its weight and a horizontal line shows its confidence interval. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Section 10.2 (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Section 10.2 (retrieved 2026-11-07)
- [12]
Relative risk (also called the risk ratio) compares the risk of an event in one group with the risk in another group; a relative risk of one means no difference between the groups. confirmedas of 2026-11-07
- NCI Dictionary of Cancer Terms: relative risk · National Cancer Institute (NIH) · Primary source · definition (retrieved 2026-11-07)
- NCI Dictionary of Cancer Terms: relative risk · National Cancer Institute (NIH) · Primary source · definition (retrieved 2026-11-07)
- [13]
Statistical heterogeneity means the results of the included studies differ from each other more than chance alone would explain. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Section 10.10.1 (retrieved 2026-11-07)
- [14]
A fixed-effect meta-analysis assumes every study estimates exactly the same quantity, while a random-effects meta-analysis assumes the true effects vary across studies. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Section 10.3 (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 10: Analysing data and undertaking meta-analyses · Cochrane · Primary source · Section 10.3 (retrieved 2026-11-07)
- [15]
Cochrane guidance tells review authors not to lean on P-value thresholds or the label "statistically significant", and to report the confidence interval with the exact P value instead. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 15: Interpreting results and drawing conclusions · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- [16]
Cochrane defines certainty of evidence as the extent to which one can be confident that an estimate of effect is close to the quantity of interest. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Section 14.2 (retrieved 2026-11-07)
- [17]
Under the GRADE approach used by Cochrane, evidence from non-randomized studies should generally be rated down two levels because of the risk of confounding and selection bias. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Section 14.2.1 (retrieved 2026-11-07)
- [18]
GRADE allows certainty in non-randomized evidence to be rated up for a large effect, a dose-response gradient, or when plausible confounding would only weaken the observed effect. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence · Cochrane · Primary source · Table 14.2.a (retrieved 2026-11-07)
- [19]
Cochrane review authors are told not to make recommendations about healthcare decisions, though they can highlight actions that might fit particular values and preferences; the handbook says such judgements are typically the domain of clinical practice guideline developers. confirmedas of 2026-11-07
- Cochrane Handbook, Chapter 15: Interpreting results and drawing conclusions · Cochrane · Primary source · Key Points (retrieved 2026-11-07)
- Cochrane Handbook, Chapter 15: Interpreting results and drawing conclusions · Cochrane · Primary source · Section 15.6.2 (retrieved 2026-11-07)
- [20]
A USPSTF grade A means high certainty of substantial net benefit; grade B means high certainty of moderate net benefit or moderate certainty of moderate to substantial net benefit. confirmedas of 2026-11-07
- Grade Definitions · US Preventive Services Task Force · Primary source · Grade Definitions table (retrieved 2026-11-07)
- Grade Definitions · US Preventive Services Task Force · Primary source · Grade Definitions table (retrieved 2026-11-07)
Revision history (1)
- Page created.
Created Nov 7, 2026. Last reviewed by an editor on Nov 8, 2026. Next scheduled review: Nov 7, 2027.
Cite this page
"How systematic reviews and meta-analyses work (Cochrane and GRADE)." ContentLora, updated Nov 7, 2026. https://contentlora.com/explain/how-systematic-reviews-work
Spotted an error? Suggest a correction or emailcorrections@contentlora.com.