Koo,

The 14% relative risk reduction (RRR) is the percentage reduction in the incidence of 3-point MACE achieved by those patients treated with empagliflozin compared to the incidence of 3-point MACE in those patients treated with placebo. 282 out of 2333 patients treated with placebo had a 3-point MACE event (12.1%) and 490 out of 4687 patients treated with empagliflozin had a 3-point MACE event (10.5%). If you divide 10.5% by 12.1%, you get a hazard ratio of 0.86. My understanding is that to get the RRR from the hazard ratio, one subtracts the hazard ratio from 1. So 1 - 0.86 = 0.14, giving a RRR of 14%.

I know there are others on this board that can likely explain statistics better than I can. In the EMPA-REG Outcome NEJM paper, they state "Since interim data from the trial were included in a new-drug application submitted to the FDA, under the Haybittle–Peto rule, a two-sided P value of 0.0498 or less was considered to indicate statistical significance in the final analyses." A similar statement was made in the presentation I linked to in my earlier post. The p-value is a measure of statistical significance. There are many different statistical tests, but the investigators had to define what statistical tests they would use in advance of the study I believe. Statistical significance is the likelihood that a relationship between two or more variables is caused by something other than random chance. In most cases, a p-value of 0.05 is the threshhold for statistical significance. Above 0.05 is interpreted as non-significant and below 0.05 is interpreted as significant.

My point was that in the EMPA-REG Outcome study, they barely met their statistical goal of p<0.0498 with their achieved value of p=0.04. If EMPA-REG Outcome had acheived a 14% RRR but with a p-value greater than 0.0498, then it would have been concluded to be a non-significant change and they would have failed to achieve the primary outcome of their study. 

BearDownAZ