Physician Pay-for-performance: Does it make sense?

Physician compensation formulas that account for performance on a range of quality metrics have become commonplace, including at our hospital.  But questions remain:  do they work now?  If not, could they?  And finally, what are the unintended consequences of P4P?  Here are the arguments in brief.

P4P is a rational approach to physician compensation whose time has come.

The history of physician compensation, speaking especially of private practice models, has been all about "production."  More visits, more encounters, and more procedures meant more compensation.  Neither the quality of the work nor outcomes mattered.  It is as though a builder constructed a house with the expectation that he would get paid the same regardless of whether the supporting structures were correctly put together, or even whether it happened to collapse in a year.

The processes in Medicine, of course, relate to a myriad of interventions from primary prevention strategies in the outpatient world, to clinical practice guideline adherence on the Medicine service.  Outcomes are, as we know, a more complicated issue, but at least in principle, some kinds of complications and even patient-reported experience of care are promising.  If the organization wants to send the message that performance to particular specifications matters, then physician compensation should reflect this.

"Bonusing" physicians for providing the right care is ineffective and misguided; it's time to pull the plug.

P4P makes sense in theory.  But as the old saw goes, "in theory there is no difference between theory and practice, but in practice there is."
I was speaking with a physician from another health system about P4P not long ago. Her compensation formula has sections in it for achieving extra credit for meeting more than a dozen quality benchmarks.  There was also a component for resource utilization and one for several patient "experience of care" measures.  She admitted that she could not actually cite what all these measures were, specifically.  And with that in mind, I wondered out loud how they could possibly, even in principle, be influencing her clinical performance.  Then there is the matter of evidence.  A Cochrane database review cited in a 2013 Wall Street Journal article revealed no evidence that P4P programs improved performance at the individual physician level.

Additionally, there is the question: what should we choose to measure? Notwithstanding the foregoing issues, economics CAN drive behavior.  And the question here is whether measuring a few things distorts the process of delivering medical care by crowding out numerous, other, unmeasured clinical activities.

And finally,  there is a culture question.  What IS the message when organizations introduce P4P?  Is it that doing the right things should be rewarded?  Or is it that professionalism is no longer seen as the critical driver it has always been, and what makes us pause to be sure there isn't something we forgot to remember, something more we should be doing for our patient, the one in the bed who is our first and only concern?  Please share your thoughts.