Not Pretty: Year 5 Physician Group Practice Demonstration Results (and lessons for Accountable Care Organizations - ACOs)

The Disease Management Care Blog was alerted by a colleague to this hot-off-the-presses CMS release announcing that the five year Physician Group Practice (PGP) Demo results have been compiled.  All ten participating groups increased their quality of care measures and seven achieved scores of 100%.  Four out of the ten groups were awarded $29.4 million in shared savings.  You can look at a results summary here and read the background report here.

Here's a visual display, courtesy of your DMCB, of the payouts for each of the five years to each of the participating clinics:

As you can see, four out of the ten, Billings ($0), Everett ($129,000, barely registering on the graph), Forsyth ($0), Middlesex ($0) didn't do very well for any of the years.  Dartmouth hit a $6 million jackpot only in Year 2, Park Nicolette hit $5 million in Year 5, Geisinger had modestly successful Years 2 and 3.  The only institutions with any consistency were Marshfield, St. Johns and Michigan. 

While the CMS press release generously refers to "positive results" and "significant progress," it also uses the word "lessons" in reference to "Accountable Care Organizations."  In looking at the graph above, the DMCB's initial learnings are that.....

1) assuming these results are generalizable, there is a 40% chance that an organization participating in a shared savings arrangement with CMS will come away practically empty handed, 30% of the participating organizations will have some good and and more bad years, there is 10% chance that an organization can get a huge amount of money and that this will work as intended only 20% of the time;

2) assuming that these Medicare claims results are the product of randomness and had nothing to do with actual planning, anything can happen and it mostly isn't good;

3) even if there is "savings," we don't know if the millions of  those shared savings are greater than the millions in investments that each of the clinics had to make.  That key information on "return on investment" remains conspicuously absent;

4) there is little relationship between achieving "quality" and saving money.  If improving diabetes, heart failure, coronary disease and preventive care was all it took (see page 3), each of the clinics should have made a bundle.

The DMCB will continue to commune with these outcomes data and await the on-line and printed interpretations of academics and bloggers.  In the meantime, if any reader would like a PowerPoint slide with the graph above, just email.