Do your notes say the 0.4% refers to the most negative bound of the 95% CI for the difference between the two treatments (as is typical in NI studies)?
“The efficient-market hypothesis may be the foremost piece of B.S. ever promulgated in any area of human knowledge!”