Begin with the question
Before reading the conclusion, write down what the study is trying to establish. Is it describing morphology, measuring diagnostic accuracy, testing a treatment or exploring an association? Different questions need different designs and appraisal methods.
CASP provides study-specific critical appraisal tools, including a diagnostic study checklist. [1] A structured approach helps readers inspect methods rather than decide from the title alone.
Examine who and what was studied
For a diagnostic study, identify how participants or samples were selected, the test being assessed and the reference standard. Ask whether the material resembles the cases in which the test would actually be used.
STARD reporting guidance addresses these elements, along with thresholds, missing or indeterminate results, participant flow and estimates of accuracy with uncertainty. [2]
Read beyond a headline percentage
Ask which errors matter clinically. Could a test miss a malignant lesion, or could a false positive lead to unnecessary investigation? Look for the underlying numbers and the uncertainty around estimates.
In an image-based study, a useful additional question is whether multiple images from the same patient were kept together when datasets were divided. Also ask whether performance was assessed on an independent population. These questions help examine how closely the evaluation matches the proposed use.
Separate reporting from validity
A well-reported study is easier to appraise, but complete reporting does not prove that the design is unbiased. Conversely, an omitted detail may prevent confident judgment. Record what is known and what cannot be assessed.
Read the limitations, funding and conflicts disclosures with the methods and results. Check whether the authors’ conclusion stays within what the study actually tested.
Finish with a practical note
My suggested reading record has four entries: the question, the main finding, the principal limitation and the decision it might inform. Add whether the paper should change practice now, prompt further reading or simply raise a hypothesis.
Critical reading makes research more useful by connecting the strength of the evidence to the decision being considered.
References
[1] Critical Appraisal Skills Programme. Diagnostic Study Checklist. Read source[2] EQUATOR Network. STARD 2015, diagnostic accuracy study reporting guidance. Read source