MACRA Payment Regulations Should Be Evidence-Based

Whether applied to policymaking for individuals, large populations, or administration of health services nationwide, it is imperative regulatory decisions be anchored to empirical evidence. The official MACRA rule has now been released.  It is 2,000 pages based on the opinion of many non-practicing physicians, Dartmouth economists, and government administrators with input from a few doctors on the front line. In my opinion, what began as a certain death sentence has commuted us to life in prison; MACRA will regulate physicians without representation. 

Let me acknowledge my opinion is limited by my own “small” practice bias. 380 thousand “small” practices (having 15 providers or less) will be exempted if they have less than 100 Medicare patients.  Your definition of small and mine are strikingly different.  Every single independent practice in my hometown of that “quasi-small” size, has sold to the local hospital already.  The “small” practices remaining in my community have 1 or 2 physicians, so I will refer to those as micro-practices for clarity.  My micro-practice serves more than 400 Medicaid patients, with a waitlist of more than 50.  MACRA rules do not seem to have an answer for when there are not enough micro-practices remaining with which to form a “virtual” group. 

I humbly suggest you expand the options in your “flexible” plan, to include a control group composed primarily of 1-2 physician practices.  Please, do not overlook the importance of tailoring interventions to the unique needs of small communities in order to ensure the existence of micro-practices in the long-term. The fates of millions of Medicare (and presumably Medicaid) beneficiaries is at stake.  It is absolutely essential that new payment plans are evaluated in comparison to a control group prior to arbitrarily being applied across the nation. 

A recent article in the NEJM evaluated early performance of ACO’s by using a control group, which is vitally important to the evaluation process.  Researchers concluded the first year was associated with early reductions in Medicare spending among 2012 entrants (1.4%, P=0.02) but not among 2013 entrants. Performance on quality measures was improved in some areas and unchanged in others.  And surprise, surprise, savings were consistently greater in independent primary care groups than in hospital-integrated groups among entrants in 2012 and 2013 (P=0.005 for interaction).  How on earth can CMS ignore yet another study showing independent primary care groups save money before someone important realizes MACRA (as it stands now) is on the bridge to nowhere?

Policymaking must use scientific research to guide decisions at each stage of the process in every branch of government. According to the Washington State Institute for Public Policy, there are three designations to grade the rigor of research methods and the amount of evidence available to guide sweeping program interventions: Evidence-based, research-based, and promising. 

Evidence-based programs have been rigorously studied; using randomized controlled trials, and found to be effective. Research-based programs have been tested using rigorous methods (studies using strong comparison groups, as I am proposing) but do not meet the evidence-based standard.  Promising programs have been tested using less arduous research designs and typically use well-constructed logic or theories to support ideas. 

Postulating and theorizing by Dartmouth economists has left us all on treacherous ground. These experts assembled data, “interpreted” it creatively, and then drew unsubstantiated conclusions upon which to base recommendations for creation of PCMH’s and ACO’s.  The fruits of their “promising, yet non evidence-based” labor have generated unimpressive outcomes, yet their poor quality decisions will not affect their income.  Culpability must be incorporated in the process this time.  Government agencies, their managers, and those economists now advising them must be held accountable for their outcomes this time before holding physicians responsible for ours.

Confidence in these experts is fading because Patient Centered Medical Homes (PCMH) and Accountable Care Organizations (ACO) are not holding up their end of the bargain, demonstrating miniscule savings at best, while making the life of a physician far more cumbersome.  A thorough critique by Kip Sullivan summarizes the research on three PCMH’s and three ACO’s showing little to any cost savings, further exposing the weak platform on which CMS has built their Quality Payment Program. 

In that same vein, CMS is estimating how much value-based payments will bring down medical costs while guiding patients toward better health.  The word “estimate” appears far too often in the Executive Summary of the MACRA Rule for me to be comfortable with this plan.  CMS intends to impose “promising, albeit not evidence-based” options on all physicians treating Medicare patients in less than 3 months.  Where is the conclusive data demonstrating cost containment and improved quality?  It does not appear to exist. What if your estimates are incorrect?  The consequences will be catastrophic for independent solo practices if your “estimates” are wrong.  Should I be forced to make this blind leap of faith without being certain?

Andy, good science will be good for your conscience. CMS policymaking must be based on rigorous research that is supported by empirical evidence, even if the results are equivocal.  Presuming, opining, and educated guessing are not adequate methods for imposing non evidence-based programs upon large populations.  Before CMS officially implements sweeping payment modifications on January 1st; please consider allowing a control group option, composed of small practices with 1-2 physicians.  I, for one, would like to be at the top of the list.  Do not throw the “fee-for-service” baby out with the bathwater before being absolutely certain your non-evidence-based payment models actually contain costs and are better for patient care quality than what is already in place.