Start with the record
Identify the registry entry, protocol, publication, sponsor, update date, recruitment status, and whether results are posted.
DIRECT
Study reader
Read the protocol before the promise.
A six-pass field guide for extracting the actual question, methods, result, and limit from a research record.
Reconstruct the question
Write the population, intervention, comparator, outcomes, and time horizon in one sentence.
Check the design
Look for randomization, masking, expectancy assessment, allocation, supervision, and prespecified analysis.
Follow every participant
Compare enrolled, assigned, completed, analyzed, withdrawn, and excluded counts. Read why data are missing.
Read effect and uncertainty
Capture absolute results, between-group difference, confidence interval, and adverse events—not only the p-value.
Write the boundary
State what the study did not test and what would be needed before applying the result elsewhere.
Registry status
A study listing is a map, not proof.
A ClinicalTrials.gov record can show that a study was planned or is underway. Recruitment status, posted results, peer review, protocol changes, and final publication must be checked separately.
View one current microdose-safety study record Learn how trial records workOffline toolkit
Keep the extraction auditable.
Download a study-appraisal checklist and a structured evidence log. Both are blank research aids, not medical questionnaires.