How to Critically Appraise a Medical Research Paper

Critical appraisal is the systematic assessment of a study's validity, results and relevance so you can decide how far to trust and use it. A reliable approach asks three questions: Is the study valid? What are the results? Do they apply to my patient or setting?

A three-part framework

1. Validity (could the results be biased?) Check design fit, selection of participants, comparability of groups, measurement of exposure and outcome, blinding, follow-up and analysis.
2. Results (what was found, and how precise?) Look at effect sizes, confidence intervals and absolute as well as relative differences.
3. Applicability (can I use this?) Compare participants, setting, intervention and outcomes with your situation, and weigh benefits against harms and costs.

Use a checklist that matches the design

Design-specific tools reduce missed items. Examples: CASP checklists and JBI critical appraisal tools for many designs; RoB 2 for randomised trials and ROBINS-I for non-randomised studies of interventions. Choose by design, apply it consistently, and, for reviews, have two people appraise independently. MRN's free critical appraisal worksheet offers a design-neutral starting point.

Illustrative example

Invented for teaching. A cohort study reports that coffee drinkers have less liver disease. Appraisal: groups differ by alcohol intake and socio-economic status (confounding); coffee intake was self-reported once (measurement error); the adjusted hazard ratio is 0.80 (95% CI 0.65 to 0.98), a modest and imprecise effect. Applicability: participants were mostly men aged 50–70. Conclusion: suggestive, not proof of causation.

Common pitfalls

• Using a checklist as a score without thinking.
• Judging quality from the journal name.
• Ignoring funding and conflicts of interest.

Key takeaways

• Appraise validity, results and applicability in that order.
• Match the tool to the study design.
• Express conclusions with appropriate uncertainty.

This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.