On the primary safety endpoint of major bleeding, Apixaban was not statsig worse than aspirin; the hazard ratio was 1.14 and the 95% CI was [0.74,1.75].
That's true, but it is not noninferior either. The sample size (5600 patients) may not big enough in this case. If it mains same percentage 1.4 vs 1.2 in ARISTOTLE trial (18,000 patients), it most likely will be stat sig.