Compare the registry results with the paper whenever their reported trial findings appear to conflict. Compare the registered results table with the published outcome table field by field to decide whether a discrepancy is a reporting difference or a substantive inconsistency for your review.
ClinicalTrials.gov defines structured registration and results fields covering study identification, status, design, eligibility, locations, outcomes, and reported results. That structure gives you a fixed reference for the primary outcome and the sample size the investigators registered. The published article gives you the outcome table readers will cite.
Align the two tables
- Retrieve the registered results table for the trial record and locate the primary outcome field and the enrolled sample size. Record the exact outcome name, time point, and analysis population as registered.
- Retrieve the published article's outcome table and record the same three items for the outcome the authors label primary. Note the analyzed sample size for that outcome.
- Compare each item in turn: outcome definition, measurement time point, analysis population, and sample size. Mark each as matching, differently worded, or different in substance.
Classify the discrepancy
A reporting difference appears when the outcome and population match but the label, time point wording, or table layout differs. A substantive inconsistency appears when the primary outcome, the analysis population, or the analyzed sample size differs in a way that changes the effect estimate you would extract.
Use that classification to set eligibility. If the discrepancy is substantive and affects the outcome you plan to pool, record it and decide whether to exclude the study, extract both versions, or contact the authors. PRISMA 2020 provides a checklist and flow diagrams for transparent reporting of what studies were found and included, which gives you a place to document the decision.
Locate the records
ClinicalTrials.gov search lets you filter registered studies by condition, intervention, location, status, study type, and dates. PubMed supports field tags, filters, Advanced Search, and citation matching for locating the published article. Match the registration number in the article to the trial record before comparing tables.