Biotech Values

[iwfal]

The events where BMS’s had a small, non-statsig advantage were deaths and non-fatal MI’s.

The article never actually says that. There is some implication that the 9.3% event rate for DES and 7.9% for BMS is this statistic or something like it (for a total of 65 deaths plus MIs). We'll have to actually read the paper to know. Side note - later they point out that the 65 events is broken down into 16 late-thrombosis events and 49 'other' events, so if the difference is due to late thrombosis the difference in rates must be dramatic, but that breakdown is never given.

Perhaps the referenced paper is interesting ... but the article itself reads like yellow journalism.

BTW - Note that even the formal papers/articles you see on late thrombosis do a pretty lousy job of looking at the risks. E.g. they often cite death rates with restenosis being a censoring event. I.e. if you get restonosis and then die while getting CABG surgery you don't count as a death for the BMS arm. It would be interesting to know whether the referenced paper repeated this flaw. (This strange confusion about what events are important is not new to the cardiac stage - e.g. the AHA recommendation for whether to take aspirin or not is based upon ... ? Drumroll please .... Death rates of bleeding vs heart attack? ... Nope. It is based upon death/hospitalization rates for the two arms. But of course it is generally *much* less consequential to be hospitalized for an aspirin induced bleed than for an MI (and the root paper does make reference to this but doesn't otherwise factor it in).)