Appraise a systematic review by checking its question, protocol, search, eligibility, study quality, synthesis, heterogeneity, certainty, and applicability.
Match the review question to your decision
Write the population, exposure or intervention, comparator, outcomes, setting, and study designs that matter to you. Then compare that frame with the review’s eligibility criteria. A rigorous review can still be irrelevant if its question differs from yours.
Inspect protocol and search coverage
Look for registration or a dated protocol, databases searched, complete search dates, language restrictions, grey-literature methods, and duplicate screening. A PRISMA flow diagram explains movement through the review, but reporting completeness alone does not guarantee that the search found every relevant study.
Examine included studies and bias
Read the characteristics table and risk-of-bias assessment. Note sample sizes, settings, interventions, outcome definitions, follow-up, missing data, and funding. A pooled estimate does not erase weaknesses shared by the underlying studies.
Interpret synthesis and heterogeneity
Check whether combination was appropriate, which effect measure was used, how statistical and clinical heterogeneity were handled, and whether sensitivity analyses change the result. Avoid reducing uncertainty to whether a p-value crossed a threshold.
Judge certainty and transfer
Review certainty assessments, publication-bias concerns, absolute effects, harms, and applicability to your population and setting. Separate the authors’ data-supported conclusion from recommendations that require values, resources, feasibility, or professional judgment.
Frequently Asked Questions
Is PRISMA a quality score?
No. It is primarily a reporting guideline; appraisal still requires judgment about methods and evidence.
Why does protocol registration matter?
It helps reveal changes made after results became known.
Does meta-analysis guarantee a strong conclusion?
No. The result depends on study quality, compatibility, bias, and uncertainty.
What is heterogeneity?
Variation among studies in populations, methods, interventions, outcomes, or effects.
Should I use the abstract alone?
No. Important limitations and applicability details are usually in the full review and supplements.
Can this replace professional guidance?
No. Consequential decisions require relevant expertise and context.
Put the Reading Into Practice
Open a systematic review in Readever and create a one-page appraisal covering question match, search, bias, synthesis, certainty, and transfer. Link every judgment to a passage or table.
Related Readever Pages
- evidence appraisal: How to read scientific papers
- evidence appraisal: AI reading for researchers
- evidence appraisal: Organize research papers and notes
- evidence appraisal: Build an evidence table
