The largest single problem with flow-based quality concepts in medicine is that neither patients nor providers consider process quality measures to be relevant to their personal concept of quality. Instead, they focus on the practical concepts of provider skill and experience – the results of constant training and procedure volume. On the other hand, doctors and hospitals are paid according to Relative Value Units (RVUs), which have nothing to do with volume, experience, or the resulting improvements in quality. Instead, RVUs create a system where specialized procedures are paid far more highly than primary care. This lures many struggling hospitals into offering high RVU procedures, regardless of their ability to attract sufficient patient volume to develop the experience and skill required for a high quality outcome. Legal restrictions on referrals compact the problem by blocking the ability of smaller hospitals to direct patients to high-volume centers without losing all of the potential revenue. To achieve an acceptable level of quality improvement in healthcare, we must first find a solution to this value/volume conundrum.
There has been a tremendous amount of work in clinical and academic circles addressing medical quality as a process-related function. As we addressed in our August paper, Medical Quality Systems: the Elusive Goal of Quality in Complex Systems, there are considerable practical problems with this approach, mostly centered on the non-linear process flow in the vast majority of medical clinical systems. The largest single problem with flow-based quality concepts in medicine, however, is that neither patients nor providers really consider process quality measures to be at all relevant to their personal concept of quality.
Patients and doctors focus on the practical concepts of provider skill and experience as the key determinants of quality, and with good reason. The literature is very clear that, more than any other single factor, experience – of both the provider and the medical center – determines the success or failure of a medical procedure (Birkmeyer et al). Specifically, in their seminal paper Hospital Volume and Surgical Mortality in the United States, the authors conclude, “In the absence of other information about the quality of the surgery at the hospitals near them, Medicare patients undergoing selected cardiovascular or cancer procedures can significantly reduce their risk of operative death by selecting a high-volume hospital.” (N Engl J Med 2002;346:1128-37)
This is pretty strong stuff, and published in a top journal. Staying on the same theme, Birkmeyer et al concluded in a paper in the New England Journal of Medicine the following year, “For many procedures, the observed associations between hospital volume and operative mortality are largely mediated by surgeon volume. Patients can often improve their chances of survival substantially, even at high-volume hospitals, by selecting surgeons who perform the operations frequently.” (N Engl J Med 2003;349:2117-27)
The racetrack metaphor that pops to mind is, if the horses are all pretty much the same, bet on the jockey. Clearly patients are encouraged by the academic literature to select busy jockeys at busy tracks.
The academic literature about this subject is voluminous, straightforward, and unassailable: practice makes perfect. The Leap Frog Group is a Washington advocacy group sponsored by major corporate employers; the kind who have great (and very expensive) healthcare plans for their employees. These folks are uninhibited by the limitations of peer reviewed publication and state with no mistaken purpose:
Tens of thousands of Americans die every year undergoing elective surgery. For many high-risk procedures, surgical patients can reduce their risks considerably by having their procedures performed at hospitals with low risk-adjusted mortality rates and/or sufficient experience with those procedures.
For some procedures, the right surgeon may be even more important. In particular, high-volume surgeons often have markedly lower mortality rates than surgeons who perform those procedures infrequently.
Dr. Atul Gawande is perhaps the most prolific and successful advocate for process control from the active ranks of medical professionals. Dr. Gawande is an endocrine surgeon at Brigham and Women’s Hospital in Boston, an associate professor at Harvard, and author of the enormously popular book (at least among quality wonks) The Checklist Manifesto, in which he analyzes the impact of checklists on process control and quality outcomes in medical applications. Dr. Gawande put an exclamation mark on the concept of the duality of quality in medicine with his recent New Yorker article, Personal Best (October 3, 2011), in which he describes how and why he decided to hire a personal coach to help improve his surgical skills. He addressed head-on the importance that technical skills and technique mean for the practicing surgeon. Clearly process control is necessary but not sufficient for a quality outcome. Constant practice and skill refinement are at least as important as any process refinement in achieving quality outcomes in medical procedures.
The message for healthcare system design is clear: Specialization, volume, and constant training are key determinants of outcomes in healthcare. From a practical perspective, a move to increase specialization and volume always tends to increase linearity in a production system, which directly addresses the issues of complexity inherent in medical delivery systems.
RVUs and the Centrifugal Forces of Procedure Venue
The Fairfax Medical Dictionary defines Relative Value Unit as a comparable service measure used by hospitals to permit comparison of the amounts of resources required to perform various services within a single department or between departments. It is determined by assigning weight to such factors as personnel time, level of skill, and sophistication of equipment required to render patient services. RVUs are a common method of physician bonus plans based partially on productivity. RVUs were adopted in 1989 as part of Omnibus Budget Reconciliation Bill of that year, and replaced the overly vague and wickedly expensive “reasonable and necessary” language that was part of the original Medicare payment method.
“…all RVUs are created equal…the person who performs 100 procedures per year gets paid the same as a provider who does just one…”
As of 2010, about 7,000 different procedures are paid by RVUs, which are based around the Current Procedural Terminology Codes (CPT Codes) owned and controlled by the American Medical Association (AMA). All RVUs are determined by the AMA’s closely controlled Specialty Society Relative Value Scale Update Committee (RUC) and are comprised of compensation for three components: physician work, practice expense, and malpractice insurance. (For a detailed description of how RVUs are calculated, see www.acro.org/washington/RVU.pdf.)
From a practical perspective, all RVUs are created equal, with no specific differentiation for the volume practiced by a hospital or physician. In the same geographic region and type of facility, the person who performs 100 procedures per year gets paid the same as a provider who does just one. This creates a system with enormous incentive for struggling hospitals to try to fix their fiscal woes by undertaking new lines of business that have high RVUs, whether or not the hospitals have the potential to become high volume – and therefore high quality – centers for that procedure.
Another important consideration is that not all procedures are created equal in the eyes of the RUC. As Dr. Pauline W. Chen put it in her recent New York Times article on the subject (Sept. 22, 2011):
Put simply, our payment system pays more for procedures performed by specialists. Specialists, therefore, have greater earning power, so more doctors choose to train to be specialists. Careers in specialties like radiology, dermatology, and neurosurgery offer lifetime earnings several million dollars higher than those in primary care. It is no surprise that medical students emerging from the educational mole hole saddled with hundreds of thousands of dollars of debt choose more lucrative fields.
The fact is that the RUC is dominated by specialty medical societies (23 of 29 seats) and operates with all the openness of England’s notorious Star Chamber, yet it directly controls the relative disbursement of $60 billion paid by Medicare to physicians, and effectively directs another $60 billion from private payers who use Medicare standards.
There is a Logical Solution, but…
Barring a coup d’etat in the RUC, disproportionate incentives for specialty procedures are here to stay. The current RVU/RUC system will continue to provide incentives for dispersion of care that work to the detriment of quality in every case delivered in a low-volume environment. Higher volumes produce better physician skills and better teams, along with delivery systems that become more linear and easier to monitor and control. Low-volume environments inhibit physician skill development, preclude specialty team formation, and are always more expensive to operate.
The logical solution to this situation is to:
- Direct procedures to centralized facilities for execution.
- Allow RVU-based definitions of clinical care delivery to begin with screening and end with rehabilitative/home healthcare.
- Treat the episode of care across both time and venue.
- Require all parties to the care to share in the revenue, so all can benefit financially from patient-centric collaboration and good outcomes.
Unfortunately, until very recently what is most logical has also been illegal.
Logic and Reality Collide at the Federal Anti-Kickback & Stark Laws
From a practical point-of-view, the Anti-Kickback and Self-Referral legislation is referred to today as part of the Stark Laws; although they are different legislative undertakings, they are directed at the same sorts of problems. The Stark Laws are named for their sponsor, Rep. Peter Stark of California, and are really targeted at the issues surrounding self-referral. The Federal Anti-Kickback laws were first passed in 1972, prohibiting the payment for referrals. As the Office of the Inspector General states in its fact sheet, first published in November of 1999:
… the federal anti-kickback law’s main purpose is to protect patients and the federal health care programs from fraud and abuse by curtailing the corrupting influence of money on health care decisions. Straightforward but broad, the law states that anyone who knowingly and willfully receives or pays anything of value to influence the referral of federal health care program business, including Medicare and Medicaid, can be held accountable for a felony. Violations of the law are punishable by up to five years in prison, criminal fines up to $25,000, administrative civil money penalties up to $50,000, and exclusion from participation in federal health care programs.
Because the law is broad on its face, concerns arose among health care providers that some relatively innocuous – and in some cases even beneficial – commercial arrangements are prohibited by the anti-kickback law. Responding to these concerns, Congress in 1987 authorized the Department to issue regulations designating specific “safe harbors” for various payment and business practices that, while potentially prohibited by the law, would not be prosecuted.
Like so many things in Federal Legislation, the law of unintended consequences came to play full force in this case. The implications of volume and quality were not understood at the time the legislation was passed, and the legislation has not kept up with the reality of medical practice. None of the 13 Safe Harbors address the concepts of optimized care based on quality/volume. On the face of it, the law makes the concept impossible. This misdirected impact of the Stark Laws has created a value/volume conundrum that must be deciphered if we are to achieve an acceptable level of improvement in medical quality.
Bundled Payments: A Viable Alternative for High-Volume Care Delivery
“…here at last is a way to implement the four-part solution to the value/volume conundrum…”
The recent Bundled Payment Initiative (BPI) launched by the Centers for Medicare & Medicaid Services (CMS) in August of 2011 offers a unique and direct opportunity to address the perverse financial incentives that disperse high-RVU procedures to low-volume facilities. In this initiative, groups of providers can share the provision of services and fees related to specified medical procedures. Here at last is a way to implement the four-part solution to the value/volume conundrum, allowing best practices and patient-centric care to bridge the legal chasm created by the inhibiting Stark Laws.
In rural environments with large numbers of critical access hospitals (CAHs), this is a particularly interesting alternative. CAHs are limited to 25 beds or less, and are designated as critical to care in geographically isolated communities. In return for the limitations on size, the CAHs are allowed to charge Medicare “cost plus” for the rendering of services to Medicare beneficiaries. These hospitals tend to be the effective “medical home” for their patients, providing primary care and related services. Depending upon specialist availability, these small facilities sometimes enter into the delivery of specialized (i.e., high RVU) medical procedures – almost always in a low-volume environment. This makes it virtually impossible to achieve the high procedural volumes necessary for high-quality outcomes.
In the Bundled Payment Initiative, these facilities can team up with larger high-volume acute care hospitals (ACHs). They can direct specialty procedures to those facilities, while sharing in the diagnostic, patient prep, and post-procedure care. They can also share in the fees associated with the procedure. This allows the patient to receive care close to home for as long as possible, and vastly improves the potential for integrated care between the primary care physician (PCP) and the specialist performing the procedure. Importantly, it eliminates the need to have a surgeon at the CAH perform the procedure in order for the CAH to benefit financially.
Still One Stumbling Block
The greatest problem with the implementation of the CAH–ACH pairing is the fact that the existing BPI plan is limited to expenses incurred under the Prospective Payment System (PPS) for CMS. The PPS establishes rates for payments to most hospitals, but does not include the cost-plus payments made to CAHs. To be truly cost effective and patient centric, the BPI must be modified to include historical levels of cost incurred at the CAH related to a single episode of care for a patient, when those costs can be tied directly to an integrated episode of care.
Consider, for example, patient John Doe (or JD), who presents at his CAH medical home for a diagnosis of hip pain. His PCP, in conjunction with the hospital staff, does a series of radiographic images – including an X-Ray and MRI – and determines that a hip replacement is called for. Accordingly, the PCP makes a referral to an ACH to perform the surgery. The ACH performs exactly the same radiographic examination in deriving its diagnosis and surgical plan.
In an ideal bundled payment, the duplicate imaging would be eliminated, and the savings would accrue to the participants. This would require some coordination and protocol adherence, but it would be better for JD and would save several thousand dollars in redundant tests. In order to accomplish this, however, the BPI package must be expanded to include the cost-plus payments from the CAH, as well as the PPS payments from the ACH. It would seem that the immediate improvement in patient care (through reduction in exposure to radiation) and the savings of thousands of dollars per case would make this well worth the additional effort.
New Hampshire and Vermont as a Case Study
On November 4th of 2011, a group of 15 healthcare providers, including acute care hospitals, critical access hospitals, and home health/visiting nurse groups from the states of Vermont and New Hampshire, submitted a letter of intent to Medicare to start the data analysis necessary to develop a BPI that would include all of the parties. Our firm, PCD Partners, Inc., acted as convener in the process, arranging the data analysis, program design, and implementation.
New Hampshire and Vermont are known locally as the Twin States. While they are quite similar in size and history, they have quite different approaches to state government and healthcare. Vermont has had a long commitment to state-directed control of healthcare payment and planning, while New Hampshire has had a very hands-off policy, with little state involvement. Summary statistics for the two states are shown below.
New Hampshire and Vermont have relatively large and geographically isolated rural populations. In both states, the rural populations are older and poorer than in urban areas. These rural populations depend on their local CAHs for the delivery of the preponderance of their medical care, but they tend to migrate to higher volume, larger ACHs for specialty procedures – amounting to an estimated 20-40% of the total annual service provided to the served populations. This is a critical loss of revenue for these small facilities, and there have been attempts over the years to provide some of these services locally. Unfortunately, with their small size and a difficulty in recruiting physicians, this has not proved to be a viable alternative in most cases.
The Bundled Payment Initiative for the CAHs in New Hampshire and Vermont offers a unique opportunity, both to improve care for their patients and to improve their own financial circumstances. Any reasonable level of revenue participation should ensure long-term financial improvement, and the integration of care back to the medical home will greatly facilitate post procedure care. Finally, the cases could be referred to ACHs in such a manner as to maximize the case loading of those facilities and physicians. The increased volumes would clearly improve care and lower the cost per procedure.
Directing patients who require specialized procedures to high volume facilities is the only way to improve care and lower cost in those procedures. Smaller hospitals, such as the critical access hospitals in New Hampshire and Vermont, have never had an alternative that allowed them to do this legally and still participate in the patient care and the financial benefits. The Bundled Payment Initiative, though not yet perfect, is a unique opportunity for small hospitals in New Hampshire and Vermont – and anywhere else in the country – to resolve the conundrum of high RVU financial benefits and high volume quality.
There is an old management saw (incorrectly attributed to Edwards Deming) that states, “you can’t manage what you can’t measure.” Measurement is a critical part of any management system – both of the process and the outcome. In the effort to improve quality in healthcare, regulators, practitioners, payers, and patient advocates have all agreed on one thing: you must have some way to measure the effects of what you are doing, and the results. The CMS, AHRQ, NCI, and a long list of other acronymic agencies have specified thousands of metrics to measure quality in healthcare. Almost all of these – and certainly all of the metrics and measures accepted by regulatory agencies and payers – are process measures. There is a certain logic to this: in the confusion of procedures and protocols currently implemented in the nation’s hospitals and clinics, it is impossible to compare outcome measures in any meaningful way. You must define the process before you can understand the outcomes. Over the past decade, literally thousands of these process metrics – normally associated with the concept of best clinical practice – have been promulgated by these groups. Making sense of who has been doing this, what the metrics are, which really matter, and why is quite a puzzle to sort out and will be the objective of most of our analysis. Not too surprisingly, outcome metrics have lagged behind process metric development, as they depend frequently on idiosyncratic patient interpretation. A patient’s own measure of satisfaction can be so entirely subjective as to be anecdotal, and is very dangerous as a comparative metric on which to base either claims of healthcare quality or claims for payment. Unfortunately, all healthcare providers, from the largest acute care hospital to the smallest walk-in clinic, must do both in order to stay certified, attract patients, and get paid. How then do we choose the best metrics and measures to satisfy the regulators and payers, and still get the data we need to improve processes and achieve a level of quality that will also satisfy the patient?
- http:// www.acro.org/washington/RVU.pdf
- High Volume and Surgical Mortality in the United States, Birkmeyer et al, N Engl J Med, Vol. 346, No 15, April 11, 2002
- Surgeon Volume and Operative Mortality in the United States, Birkmeyer et al, N Engl J Med, Vol. 349, No. 22, Nov. 27, 2003
- Trends in Hospital Volume and Operative Mortality for High-Risk Surgery, Finks et al, N Engl J Med 364, No. 22, June 2, 2011
- Outlier Payments for Cardiac Surgery and Hospital Quality, Baser et al, Health Affairs, 2009; 28(4), 1154-1160
- Operative Mortality and Procedure Volume as Predictors of Subsequent Hospital Performance, Birkmeyer et al, Annals of Surgery, 243, No. 3, March 2006
- United States rural hospital quality in the Hospital Compare database – Accounting for hospital characteristics, Goldman et al, Health Policy, 87(2008), 112-127