Different kinds
of uncertainty.

A well-studied question with no clear benefit is different from an untested claim. Neither should be turned into an absolute statement about every dose, product or person.

When trials do not demonstrate benefit

See the specific questions on routine heart prevention, AMD progression, glucose control and cancer prevention. Population, outcome and follow-up limit the interpretation.

When results are mixed

Dry-eye studies vary by cause, formulation and outcome. Cognitive test findings do not establish prevention of dementia. New studies deserve appraisal, rather than an automatic positive or negative verdict.

When there is an evidence gap

The Atlas has not identified an adequate clinical basis for an MCAS treatment claim. That is not a randomized demonstration of no effect. Mechanistic plausibility cannot fill the gap.

How we score the evidence ↗ · Explore all research questions ↗