In the last decade, guidelines for the presentation of statistical results in medical journals have emphasized confidence intervals (CIs) as an adjunct to, or even a replacement for, statistical tests and p values. Because of the intimate links between the 2 concepts, authors now use statements like