JL isn't confused about anything - Will is - this statement is the problem:
The numbers used in the math are for the primary events which I thought are non-fatal events. The secondary events include cv death and others and the risk % is even smaller, e.g., 3.5%
1. The numbers used were from the primary endpoint of R-IT. which was 5 pt MACE.
2. Primary events are defined as the first CVE a patient suffers - it's not JL's definition, it's a widely known and used definition in CVOTs. Same goes for the term secondary event - that's when a patient with established CVD that has already had at least one event has a second one, or 3rd, or 4th, etc., and it need not be an MI or stroke, it could be hospitalization from unstable angina. R-IT looked at secondary [event] prevention and primary [event] prevention - in practical use, you don't write the words contained in brackets.
3. The primary endpoint of R-IT included CV death, it's part of 5 pt MACE. The most important secondary endpoint was 3 pt MACE, which includes CV death - so Will' statement "The numbers used in the math are for the primary events which I thought are non-fatal events." contains two errors, the improper term used (events instead of endpoints) and "non-fatal", because obviously CV death is, erm.....fatal.