Table 2

The predefined non-inferior margins of non-inferiority trials with the assumed and observed event rates and the impact on the relative risk difference

TrialHypothesisNon- inferior margin
(absolute difference)
Assumed event rate on standard treatmentCalculated relative risk difference with assumed event rateObserved event rate on standard treatmentCorresponding relative risk difference with observed event rate (with the predefined absolute difference non-inferior margin)
PARTNER IA27 TAVI non-inferior to surgical AVR in high-risk patients7.5%32%1.23 (32%+7.5%)/32%26.8%1.28
(26.8%+7.5%)/26.8%
TREAT28 Ticagrelor non-inferior to clopidogrel after fibrinolysis for STEMI1%1.2%1.83 (1.2%+1%)/1.2%0.69%2.44
(0.69%+1%)/0.69%
iFR SWEDEHEART29 iFR guided non-inferior to FFR-guided intervention3.2%8%1.4
(8%+3.2%)/8%
6.1%1.53
(6.1%+3.2%)/6.1%
PARTNER III30 TAVI non-inferior to surgical AVR in low-risk patients6%16.6%1.36 (16.6%+6%)/16.6%15.1%1.4
(15.1%+6%)/15.1%
BIONYX12 Resolute Onyx stent non-inferior to Osiro stent2.5%6%1.42 (6%+2.5%)/6%4.7%1.53
(4.7%+2.5%)/4.7%
  • The iFR SWEDEHEART trial showed an absolute difference between instantaneous wave-free ratio (iFR) and fractional flow reserve (FFR)-guided intervention of 0.7% (in favour of FFR-guided, 95% CI −1.5% to 2.8% with a relative risk of 1.12, 95% CI 0.79 to 1.58). This met the criteria for non-inferiority for iFR-guided intervention as the upper bound of the absolute difference did not exceed the predetermined non-inferior margin of 3.2%. However, if the relative risk method had been used, non-inferiority of iFR-guided intervention could not have been declared as the upper bound of the 95% CI of the relative risk was 1.58, exceeding 1.4. Even with the higher relative risk margin based on the observed event rate of the FFR arm (1.53), the actual results of the trial would not have met non-inferiority criteria.

  • AVR, aortic valve replacement; STEMI, ST-elevation myocardial infarction; TAVI, transcatheter aortic valve implantation.