Thursday, July 22, 2010

Pay for Performance (P4P), Large Physician Groups and Accountable Care Organizations. Insights from Behavioral Economics


To: Sr. VP and Chief Financial Officer, Health Insurance Operations

From: VP, Network Provider Relations, Health Insurance Operations

Re: Provider Pay for Performance "P4P"

Thank you for so repeatedly sharing your insights about our provider pay-for-performance programs ("P4P"). My staff truly appreciate the energy and passion that's been reflected in your many emails and voice mails in which you've vigorously contrasted our modest quality outcomes with the millions of dollars in additional expense that has gone to out to our contracted providers. Like you, we continue to seek better understanding about the role of this important initiative in mitigating our company's continued unfavorable cost trends, especially since this is on the Agenda of an upcoming Board of Directors meeting.

Since we've agreed that there may be one more opportunity to continue P4P, please allow me to share some important insights I've gained thanks to the Disease Management Care Blog. I highly recommend that healthcare leaders such as yourself receive regular doses of DMCB, since scientific studes have shown it can reduce the size of throbbing forehead veins, which has been a concern for you.

Thanks to the DMCB, we now know about this hot off the presses American Journal of Managed Care article titled "Using the Lessons of Behavioral Economics to Design More Effective Pay-for-Performance Programs" by Ateev Mehrotra, Melony Sorbero and Cheryl Damberg. This is an important manuscript, not only because it can help us better incent physicians, but because it also gives some insights about how physician groups and Accountable Care Organizations (ACOs) can screw this up.

First: the lessons:

1. Divide the incentives rather than a single lump payment: repeated payments have more appeal than single ones.

2. Tier the performance thresholds: everyone should have a goal and a reward that's within reach.

3. Reward the providers right away: a lag-time can undermine the priority of today's performance measures versus other concerns.

4. Witholds are bad: it can annoy docs, who are already grouchy. It you must do this, try a "deposit contract."

5. Keep it simple: otherwise, complex decisions will make persons risk averse. Contrast shared savings or quartile scoring with say, simply paying a bonus for every A1c that's ordered.

6. Make it stand alone: it shouldn't be mixed up with other forms of reimbursement, which could dilute its impact

7. "In kind" rewards can work: a "splurge" is more enjoyable than a check.

This is why our Provider Relations personnel have been concerned about the terms of our P4P programs with our larger physician groups. As you are aware, these groups tend to demand that any P4P from our insurance company go to the group and not their individual physicians.

They claim that it is difficult for us to accurately credit their individual physicians, that the group is ultimately responsible for performance improvements and that we can "trust" them. Yet, we believe we have a stake in assuring the physicians are fully and personally rewarded and suspect the group is diverting money toward other capital projects. What's more, every single of the lessons described above are routinely ignored by the large groups.

Which has important implications for ACOs. As we've noted before, there is a long history of physician-hospital distrust. That alone should give pause to the notion of funneling P4P to a hospital dominated organization without careful assurances that the P4P is flowing to where it is intended.

Once again, I want to thank you for your interest in shaping our P4P programs so that our enrollees can achieve the greatest benefit possible. It is clear the science of P4P is still in evolution and your support will be greatly appreciated going forward.

1 comment:

Phil 314 said...

Good stuff. I've always contended its not about short term results but long term behavior change.

and that requires a behavioral approach to physician change. So far our approach has been money and information (and lots of pieces of paper)