How to read a clinical trial record

01

What the evidence says

Find the registry identifier and initial registration date. Check phase, sponsor, recruitment status, locations, eligibility, assignment, masking, intervention, comparator, primary and secondary outcomes, measurement times, enrollment target, and data-monitoring structure. Registration before enrollment reduces but does not eliminate selective reporting; a late registration should be clearly acknowledged.

Sources for this section: [1]

02

How to use the finding

For results, follow every randomized participant through allocation, treatment, follow-up, and analysis. Determine whether the primary analysis follows intention to treat, how missing outcomes were handled, and whether stopping rules or interim analyses were planned. Compare event counts and absolute risks, not only relative measures. Read adverse events and withdrawals with the same attention as efficacy.

Compare absolute and relative changes in risk

Sources for this section: [1]

03

Limits and open questions

A recruiting study has no efficacy result, a completed study may have no posted result yet, and a registry’s status depends on submitted updates. One trial rarely settles comparative care. Place it beside prior evidence, biological plausibility, other trials, systematic reviews, external validity, and regulatory information. ClinicalTrials.gov explicitly states that listing does not mean the U.S. government evaluated safety or science.

Sources for this section: [1]