Inclusion and exclusion criteria are the explicit rules that decide which studies a systematic review keeps and which it discards. Inclusion criteria describe the features a study must have to be eligible; exclusion criteria describe the features that rule a study out even when it otherwise qualifies. Together they make study selection transparent and reproducible, so that a second reviewer applying the same rules to the same records would reach the same set of included studies.
Because these criteria are pre-specified before screening begins, they are one of the strongest defences a review has against selection bias. They stop reviewers from quietly redrawing the boundaries once they have seen which studies would or would not support a hoped-for conclusion.
Why these criteria are written before you search
The criteria are fixed in the protocol, not invented during screening, for the same reason a referee agrees the rules before a match rather than during it. If you decide what counts as eligible after seeing the results, the selection becomes a way of engineering the answer. Locking the criteria into the systematic review protocol removes that temptation and lets peer reviewers check that your final included set genuinely follows your stated rules.
This pre-specification is also what makes a systematic review reproducible in a way an ordinary literature review is not. The rules are public, and anyone can audit them against the studies you eventually included.
Deriving the criteria from your PICO question
You do not invent eligibility criteria from scratch. They fall out of the PICO question: each element of the question becomes one or more eligibility criteria. A study is eligible only if its population, intervention, comparison, and outcome all match what the question specifies.
| PICO element | Inclusion criterion | Exclusion criterion |
|---|---|---|
| Population | Adults aged 18 and over with type 2 diabetes | Children, or type 1 diabetes |
| Intervention | Structured aerobic exercise programme | Diet-only or medication-only programmes |
| Comparison | Usual care or no intervention | Studies with no comparison group |
| Outcome | Reports HbA1c as an outcome | No glycaemic outcome reported |
| Study design | Randomised controlled trials | Case reports, editorials, protocols |
Notice that some criteria, such as study design, language, and publication date, sit outside PICO but still belong in the eligibility set. Each one should have a defensible methodological reason rather than being added for convenience.
Worked example: from topic to criteria
Suppose the question is whether mindfulness reduces anxiety in university students. The inclusion criteria would specify the population (currently enrolled university students), the intervention (a structured mindfulness programme), a validated anxiety outcome measure, and an eligible design such as a controlled trial. The exclusion criteria would then rule out non-students, mindfulness blended inseparably with other therapies, studies without a validated anxiety measure, and conference abstracts lacking full data.
Written this way, the criteria are specific enough that two reviewers screening the same abstract will almost always agree, which is the whole point. Vague criteria such as "relevant studies on student wellbeing" guarantee disagreement and force endless adjudication.
Applying the criteria during screening
Criteria are applied in two passes: a broad title and abstract screen, then a stricter full-text screen. At the full-text stage you must record the single specific reason each excluded study failed, because those reasons populate the exclusions box of your flow diagram. Good practice in this phase, including duplicate screening and pilot testing, is covered in our guide to title and abstract screening best practices, and the counts you record feed straight into the numbers shown in the PRISMA flow diagram.
Common mistakes to avoid
The frequent failures are vague criteria too imprecise to apply consistently, criteria invented after seeing results, language or date limits with no methodological justification, and a failure to record exclusion reasons at full text. Each of these weakens reproducibility, and each is easy for a careful reviewer to spot. When the criteria are clean, the rest of the selection, and the PRISMA flow diagram you build from it, follows almost mechanically.
Frequently Asked Questions
What do inclusion and exclusion criteria mean in a systematic literature review?
Inclusion criteria are the characteristics a study must have to be eligible for the review, such as the right population, intervention, outcome, and study design. Exclusion criteria are characteristics that disqualify a study even if it meets the inclusion rules. Together they define the exact boundary of the review and make study selection reproducible.
How do you write inclusion and exclusion criteria in research?
Start from your research question, usually framed with PICO, and turn each element into a rule a study must satisfy. Add criteria for study design, language, publication type, and date where there is a methodological reason. Write each criterion specifically enough that two reviewers would apply it the same way, then pilot the criteria on a sample of records before full screening.
What is an example of inclusion and exclusion criteria in a literature review?
For a review of mindfulness and student anxiety, an inclusion criterion might be "currently enrolled university students receiving a structured mindfulness programme with a validated anxiety outcome", while an exclusion criterion might be "studies combining mindfulness inseparably with other therapies, or lacking a validated anxiety measure". Each rule maps to an element of the research question.
Do systematic reviews have inclusion and exclusion criteria?
Yes. Explicit, pre-specified inclusion and exclusion criteria are a defining feature of a systematic review and one of the main things that distinguish it from a narrative literature review. They are stated in the protocol before searching begins and are reported in the final review so that the selection process is transparent and reproducible.