I agree it looks fishy on the surface, but I did some digging a while back (you may remember the post) and found peer-reviewed reports which suggested the wisdom that bare metal stents do not undergo late thrombosis is incorrect. Excluding revascularized vessles from analysis, as appears to have been the old standard, is simply a misleading strategy which was based on the presumption that bare metal stents do not thrombose after 30 days. The poor timing of this redefinition certainly raises eyebrows, but does not in itself make a conspiracy. Heck, this new definition is apparently mandated by the FDA, what do you make of that?
I stand by my working hypothesis that the observed "signal" of increased late thrombosis with Cypher is, in a large part, due to the new ways stents are being used in the DES era. Perhaps we will see BMS resurgent use level off early next year.
Notes:
-Even if DES market share stabilizes, I expect the price of Cypher will continue to decline as competetion increases.
-Late thrombosis is real, which means a solution is still needed. (return to BMS is not a good solution)
Market Data 
Markets 
