Posters · Compared
Systematic review vs meta-analysis
A systematic review uses a planned, reproducible method to find, select, appraise and summarise all the studies that answer a specific question. A meta-analysis is a statistical technique that combines the results of two or more of those studies into one pooled estimate. Many reviews include one, but only when the studies are similar enough to combine.
The two terms are often used as if they were one, which hides an important decision inside every review: whether pooling the numbers is justified at all. Knowing the difference helps readers judge a single pooled figure and helps authors defend choosing not to produce one.
Indunil Asanka · Co-founder
5 min read · Published
| Narrative review | Systematic review | Meta-analysis | |
|---|---|---|---|
| Question | Broad, often a topic overview | Specific and prespecified | A specific quantitative effect |
| Search | Selective, not usually reported in full | Comprehensive, reported so it can be repeated | Uses the studies a systematic review has found |
| Selection and appraisal | Author judgement | Eligibility criteria set in advance, risk of bias assessed | Depends on the review's appraisal |
| Synthesis | Descriptive discussion | Structured, narrative or statistical | Statistical pooling with a weight for each study |
| Output | An expert summary | A summary of all the evidence and its certainty | A pooled estimate with its interval and heterogeneity |
| Reporting guideline | None specific | PRISMA | PRISMA, with meta-analysis items |
What makes a review systematic
The Cochrane Handbook describes a systematic review as an attempt to collate all the empirical evidence that fits pre-specified eligibility criteria in order to answer a specific research question, using explicit, systematic methods chosen to minimise bias. In practice that means a protocol written before searching, a search of several databases and other sources documented well enough to repeat, screening against fixed criteria by more than one reviewer, data extraction, assessment of each study's risk of bias, and a synthesis that states how certain the evidence is. The method, not the topic or length, is what makes it systematic.
What a meta-analysis adds
A meta-analysis takes the effect estimates from included studies and combines them, giving more weight to more precise studies, usually those with more participants or events. The Cochrane Handbook's chapter on analysing data and undertaking meta-analyses describes it as the statistical combination of results from two or more separate studies. The result is a pooled estimate with a narrower confidence interval than any single study, plus statistics describing how much the studies disagree. It can reveal an effect too small for individual studies to detect, but it cannot fix problems in the studies it combines.
When pooling is not appropriate
Pooling assumes the studies are estimating comparable effects. When populations, interventions, comparators or outcome measures differ widely, a single number can mislead, averaging over things that should not be averaged. Heterogeneity statistics help, but the first judgement is clinical and methodological. Reviews in that situation use structured narrative synthesis instead, grouping studies and describing patterns, sometimes with a chart of individual study results without a pooled summary. Choosing not to pool is a legitimate and often correct decision, and the review should explain it rather than apologise for it. Subgroup analyses and meta-regression can explore why studies differ, but with few studies they are weak, and the poster should present them as exploration rather than explanation.
Reporting both on a poster
PRISMA, updated in 2020, is the reporting guideline for systematic reviews and includes items for meta-analyses. A review poster needs the question, the eligibility criteria, the sources and search dates, the screening flow from records identified to studies included, a summary of study characteristics, the risk of bias results and the synthesis. If a meta-analysis was done, report the pooled estimate, its interval, the model used and the heterogeneity. State the certainty of the evidence in words. Registration of the protocol, for example in a prospective review register, should be named because reviewers look for it.
Common mistakes
Calling a literature review systematic because it searched a database, without a protocol, criteria or appraisal, is the most common. Pooling studies that measured very different outcomes into one estimate is next. Others are reporting a pooled estimate without heterogeneity, ignoring risk of bias when interpreting the result, and counting the same participants twice when several papers report one trial. A review poster that shows a pooled number but not how many studies and participants it rests on leaves readers unable to judge it. Another is presenting a review of reviews as though it were a review of primary studies, which can count the same trials several times over.
Where it shows up in the poster builder
The participant flow block, on ten of the forty layouts, holds six nodes and one aside, which fits a PRISMA style path from records identified to studies included, with exclusion reasons folded into the aside. There is no forest plot among the ten chart kinds, so a pooled analysis is usually shown as a horizontal bar chart of study effects, without intervals, with the pooled estimate and its interval in a key number block. The systematic review example uses its flow for screening and a chart of how studies measured output, and explains in its introduction why a pooled number would mislead.
Questions people ask
Can a meta-analysis be done without a systematic review?
Technically yes, for instance pooling a set of trials from one research program, but it is prone to bias because the studies were not selected by a comprehensive, prespecified process. Readers cannot tell whether other relevant studies with different results were left out. For published evidence summaries, a meta-analysis should sit inside a systematic review.
How long does a systematic review take?
Often many months and sometimes more than a year for a full review, depending on the number of records to screen and studies to extract. Rapid reviews shorten the process by limiting searches or using a single reviewer for some steps, and should say which shortcuts were taken, because each one increases the chance of missing or misjudging evidence.
What is a scoping review?
A scoping review maps the extent and nature of research on a broad topic, identifying concepts, evidence types and gaps, without usually appraising quality or pooling results. It suits emerging fields or questions not yet specific enough for a systematic review. It has its own PRISMA extension and should not be presented as a test of effectiveness.
What does a random effects meta-analysis assume?
It assumes the studies estimate different but related true effects, which vary around an average, rather than one identical effect. The pooled result is that average, and its interval is usually wider than in a fixed effect analysis when studies disagree. Many reviews choose random effects because real studies rarely share exactly the same conditions.
Do systematic reviews include unpublished studies?
Good ones try to. Searching trial registries, conference abstracts, theses and other grey literature reduces publication bias, the tendency for positive results to be published more often. The review should report which sources were searched and how many unpublished studies were found, and consider whether missing studies could change its conclusions.
How should certainty of evidence be described?
Many reviews use the GRADE approach, rating the certainty of each important outcome as high, moderate, low or very low, based on risk of bias, inconsistency, indirectness, imprecision and publication bias. On a poster, one line per key outcome giving the effect and its certainty rating helps readers avoid over reading a pooled number built on weak studies.
Make one with posters
The button opens the generator with this use case already described. Change the wording to match your own.
Create a poster with OneCraftRelated questions
- What is a PRISMA flow diagram?A PRISMA flow diagram tracks records through a systematic review: identified, screened, assessed, included, with exclusions and reasons at each stage.
- What is a forest plot?What is a forest plot: a chart of each study's effect and interval with a diamond for the pooled result. How to read one, and how to show one on a poster.
- What is an RCT (randomised controlled trial)?What is an RCT: a study that randomly allocates people to an intervention or a control so groups differ only by chance. How it limits bias, and what it costs.
Step by step in the builder: Choose the template, theme and size.
Written and checked by the OneCraft team. Last checked .