Report / research literacy
Reading a new clinical study
Begin with the study question and the design, then inspect the findings and their limits. A title, registered plan or press release is a lead to the evidence, not a substitute for it.
First identify what you are reading
Distinguish a trial registration, study protocol, results report, preprint and review. A registration describes intended work and can help compare the plan with later reporting. It is not a positive result. WHO emphasizes complete public reporting because missing or selectively reported results can leave a misleading evidence base. Absence of posted results means the record alone cannot answer whether a treatment worked.
Evidence: WHO: Reporting on trial findings
Write the question before reading the conclusion
Identify who was studied, what intervention or exposure was examined, what it was compared with, and which outcomes were measured over what period. FDA describes a trial protocol as the plan for eligibility, study size, duration, controls, assessments and analysis. These choices determine what question the trial can answer. They do not make its findings applicable to every patient or care setting.
Evidence: FDA: Clinical research and trial design
Read the design, not just the phase
For drug development, early clinical work commonly examines safety and dose, while later studies address treatment benefit in larger groups and gather more safety information. A phase label is a broad description, not a quality rating. Inspect the actual control group, allocation, follow-up, missing data and analysis. These details matter more than assuming every phase 3 study establishes benefit.
Evidence: FDA: Clinical research and trial design / ICMJE: Reporting methods and results
Look for the size of benefit and the harms
Read the primary outcome first, then the other planned outcomes and harms. Separate the number of participants from the number of measurements or events. ICMJE recommends effect estimates and uncertainty rather than relying only on P values. Statistical significance does not by itself describe the importance of a change. A short study also cannot settle risks that need longer observation.
Evidence: ICMJE: Reporting methods and results
End with a bounded conclusion
State what was observed in this population under these conditions, then name the limits. ICMJE asks authors to relate conclusions to the study goals, avoid unsupported claims and distinguish new hypotheses from established findings. Compare the result with the original objectives and methods, including which analyses were planned or exploratory. WHO warns that missing and selectively reported trial results can distort the evidence base. A paper should leave readers with a bounded account of what was studied, not a universal treatment recommendation. Individual decisions still require clinical judgment.
Evidence: WHO: Reporting on trial findings / ICMJE: Reporting methods and results
Source note
This is general health-science education. No clinical review has been performed.
No automatic quality score, individual treatment recommendation or unverified live-update claim.