Trust & Transparency
How We Evaluate Research
No single badge or score can replace study design. We evaluate each claim through a consistent set of questions.
Evidence level
We identify whether a claim comes from randomized trials, observational studies, case reports, animal work, cell studies, or mechanistic reasoning.
Design and comparison
We examine randomization, blinding, control groups, pre-specified endpoints, and the ability of the design to answer the stated question.
Population and duration
Sample size, eligibility, attrition, treatment duration, and follow-up shape both confidence and applicability.
Outcomes
We distinguish primary from exploratory endpoints, clinical outcomes from surrogates, relative from absolute effects, and statistical from practical importance.
Bias, conflicts, and replication
Funding, investigator relationships, selective reporting, missing data, and independent replication are part of the evidence—not administrative footnotes.
The stopping point
Our conclusion should be no broader than the studied population, intervention, comparator, outcome, and time period allow.
Last reviewed August 16, 2026.