The terms are used interchangeably in a great many titles, including published ones. They describe different things, and reviewers notice when an author does not appear to know which they have done.
What a systematic review is
A systematic review answers a defined question by searching for all relevant studies according to a pre-specified protocol, screening them against explicit inclusion criteria, appraising their quality, and synthesising what they collectively show.
The defining feature is that the method is decided in advance and reported in full, so another researcher could repeat the search and reach the same set of studies.
What a meta-analysis is
A meta-analysis is a statistical procedure that combines the numerical results of multiple studies into a single pooled estimate, weighted by the precision of each.
It is a technique, not a study design. It is normally performed within a systematic review, on the subset of included studies that report comparable outcomes.
The relationship between them
- Systematic review without meta-analysis. Common and entirely legitimate. If studies are too different in design, population or outcome measure, pooling them produces a number that means nothing. The synthesis is then narrative.
- Systematic review with meta-analysis. The usual arrangement when outcomes are comparable.
- Meta-analysis without a systematic review. Rare and generally indefensible, because the studies included were not identified by a reproducible process.
If the studies are too heterogeneous to pool, saying so is a finding. Pooling them anyway is an error.
What reviewers check
For a systematic review: a registered protocol, a reproducible search strategy, explicit inclusion criteria, a risk-of-bias assessment, and PRISMA 2020 reporting.
For a meta-analysis, additionally: a justified statistical model, a heterogeneity assessment, and some examination of publication bias.
Which should you do?
The question decides it. If the studies in your field measure the same outcome in comparable ways, a meta-analysis adds precision. If they do not, attempting one will be the first thing a reviewer objects to.
Our guides on writing a systematic review and conducting a meta-analysis cover each in detail, and this comparison of review types covers the wider family.