I was surprised to see this study both written up so seriously in the NYTimes and then taken so seriously on HN because when it came out it was kind of used as a textbook example of how poor study design leads to bad conclusions.

Here's the asterisk:

> although adjusted percutaneous coronary intervention (PCI) rates were lower during meetings (20.8% vs 28.2%; P = .02)

IE: there are nearly 30% fewer angioplasties during annual cardiologist meetings.

Unfortunately the study looked at admissions dates rather than treatment dates (which also lines up with there not being angioplasties during those dates). Someone coming in with heart failure is going through the same procedures regardless of who is there to stabilize them, the actual repair is frequently done later, ie: when the senior doctor gets back from the conference. Another, totally reasonable, possibility emerges then: doctors coming back from conferences perform better for a small period of time after the conference.

Let's see this study repeated with treatment dates instead of admission dates and see what happens.