Why Value-Based Purchasing will put at least one-fourth of hospitals out of business
Remember your first day of college orientation? Do you remember when that college president or dean gathered all of the incoming freshman into an auditorium and said these words “Now, look to your left and look to your right - one of you won’t be here in 4 years … because our graduation rate is 2/3 of our matriculation rate.”? Of course we all thought, that poor sucker to our left doesn’t have a chance. Well, The Center for Medicare and Medicaid Services (CMS) is proposing something similar for all hospitals receiving medicare risk-pool dollars.
CMS is considering adopting Value-Based Purchasing (VBP) as part of its reimbursement model and VBP is also a core cost-control component in many of the healthcare reform proposals snaking their way through our capital’s corridors. On it’s face, VBP seems like a great idea, namely that reimbursement levels should be tied to objective measures of value such as quality of care, quality of outcome and quality of patient experience. After all, why should a hospital with a high rate of complications during coronary-artery bypass graft (CABG) surgery, a high rate of 30-day readmits after CABG surgery and/or a consistently aloof and condescending staff before, during and after CABG surgery be paid as much per procedure as a hospital with a low-rate of complications during CABG surgery, a low rate of 30-day readmits after CABG surgery and/or a consistently caring, responsive and communicative staff before, during and after CABG surgery?
Intuitively, I think that we would all agree that the first hospital should be paid less, if at all. More importantly, the vast majority of the public believes so and hence our representatives will vote so. But let’s take a deeper look. Specifically at the Senate Finance Committee’s Description of Policy Options Transforming the Healthcare System: Proposals to Improve Patient Care and Reduce Health Care Costs, April 29, 2009 wherein existing metrics such as the Department of Health and Human Services (HHS) core measures as presented on the Hospital Compare website and the CMS-mandated Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) measures will be combined in a non-obvious, yet-to-be-defined way in order to yield a percentile score for each hospital which is relative to all hospitals in the country. In other words, just like college students in Intro to Econ, each hospital will be graded on the curve and someone, at least one, must get a zero. The good news of course, is that someone, at least one, must get one hundred. Those hospitals scoring 76 and above will receive all CMS risk-pool dollars plus a modest bonus of perhaps 3 percent. Hospitals scoring between 26 and 75 will receive CMS risk-pool dollars on a linearly pro-rated basis. Hospitals scoring 25 or less will receive no risk-pool dollars. Let’s consider 3 hospitals, each with $4 million in risk-pool funds eligible for refund. The first hospital scores a 90, the second hospital scores a 50 and the third hospital scores a 10. The first hospital will receive $4.12 million over the withholding period, the second hospital will receive $2 million (note that is a net loss of $2 million) over the withholding period, the third hospital will receive no risk-adjusted dollars at all, a net loss of $4 million (for more details see Hal Andrews and Gunter Wessels article in August’s issue of HFM).
Now, before you stab your finger in the air and say the third hospital deserves to lose $4 million, consider that for some metrics there are only small differences between hospitals; while for other metrics there are large differences between hospitals which differences are almost entirely dependent on the differences between the patient populations served by those hospitals. Consider also that hospitals will rapidly converge on the mean score for all metrics which are straightforward process improvements and that hospitals will avidly “game the system” for all metrics which are not straightforward process improvements and you have a recipe for massive hospital attrition where the attrition is no longer tied to a failure to meet absolute performance measures, but is rather tied to the relative performance of how well a hospital can game the system and to statistically random fluctuations in hospital performance on specific metrics. In other words the relative VBP system mandates that one-fourth of all hospitals, regardless of what they do or do not do, will have all of their risk-pool dollars withheld. Now, in some fields of endeavor, a relative VBP-type system is a model for high-performance success. I remember hearing of one database vendor who fired the lowest performing 10% of their sales force every year. It didn’t matter if every account executive beat their quota that year, it didn’t matter if an individual account executive had been in the highest performing 10% the previous year. This model worked very well because it strongly incented every account executive to peak performance and because this particular company knew that every year there was a new and eager crop of sales talent to recruit. In the field of quality incentivization for hospitals, relative VBP is a model for disaster. Given that half of all U.S. hospitals are in the red (at least as of March of this year), that the medicare inpatient margin percentages have been increasingly negative since 2006, that the ranks of the uninsured (who either simply don’t pay their hospital bills or who rely on medicare to pay their hospital bills) have significantly swollen with the downturn in the economy and that rates of elective and self-pay (and therefore highly profitable) surgeries has dropped in 2009 (albeit anecdotally), can one out of four hospitals stand to lose all of their risk-pool dollars? Can three out of four stand to lose at least 25% of their risk-pool dollars?
Disclaimer: The opinions expressed herein are my own personal opinions and do not represent my employer's view in any way.