So 4.8% primary endpoint risk reduction over median 4.9 years.
And we're talking about 3 months (.25 years)
So if the delay kept a million people off V for 3 months, you'd expect the following event reduction
1,000,000 x .048 x .25/4.9
That's 2,448 events
Of which about ~20% would be CV deaths (maybe a little less) -- call it 15% for conservatism -- so ~367 events
But of course you have to look at ultimate usage levels, not upfront
If ultimately 20 million people globally are on V, and this delay pushed that out 3 months, then you're talking 20x that death count