I think you have it backwards. The risk of the individual surgeries is not increased. Only the risk that a surgeon will be confronted with an above average critical surgery is increased.

It is not even clear that there actually is a problem. It's just a weird way to slice the data to produce an effect.