Knowledge Center

These results are impressive … but I have a question for those involved… if low revenue = physician led then can we eliminate adverse selection (bias) within the attribution results from being a contributor to “performance”? 

I have been beating this horse over and over but based on our analysis of larger delivery systems, and larger less curated panels of physicians (including often lots of specialists) I suspect that it may be more an issue of attribution rather than actual performance … that is to say it is a measurement artifact, not a fact. I could be very wrong … but smaller physician led ACOs have many features, other than physician led, that might correlate with the findings. PCP to specialists ratios, smaller geographic service areas.

We all are comfortable with “correlation not being causation” … but does apply here?

At Nascate we have seen loose and temporary relationships for a specific specialty episode – result in terrible apparent performance (because episode intensity ties in with temporary assignment – with all the costs attached) — but it is the result of a broken assignment methodology combined with a large and diverse Tax ID being used to define the ACO’s providers.