ArticleBMJ (Clinical research ed.)2000

Confidence intervals should be used in reporting trials.

M BlandPubMed ↗Full text ↗Publisher ↗

No numbers read from the abstract.