Saturday, February 08, 2014

Query: will the doc fix bend the healthcare cost curve?

[Program note: the lede is somewhat buried here, in that the post goes from exposition to open query 5 paragraphs down. Also, see update at bottom re doc fix pay-fors]

The doc fix is in, as Sarah Kliff pithily put it. That is, bipartisan legislation unveiled on Feb. 6 would repeal the broken Medicare payment formula that has had to be patched every year -- and transition essentially all Medicare payment to doctors away from fee-for-service and into (alleged) payment-for-performance.

Under the proposed legislation, three existing merit-based incentive payment programs would be consolidated. Payments to doctors will be adjusted based on a composite score of their performance compared to peers comprised of  scores in four categories: 1) quality of care; 2) resource use; 3) "meaningful use" of electronic health records (EHRs); and 4) improvement over their own past performance.* The quality measures, to be updated every year with input from "eligible professional organizations and other relevant stakeholders," are in five categories: clinical care, safety, care coordination, patient and caregiver experience, and population health and prevention.

If it passes and works more or less as designed -- two enormous ifs -- the doc fix, as I have noted before, could eclipse in impact all the toxically politicized budget wrangling of the past five years.  "Healthcare reform is entitlement reform": if Medicare cost growth is slowed significantly, our fiscal future is likely secured  If it works (or, for that matter, if it fails spectacularly), the doc fix could have more impact on the healthcare cost curve than the Affordable Care Act, since its package of consolidated performance incentives, as Sarah Kliff points out, "constitute a more significant move toward pay-for-value than is the Affordable Care Act, where those efforts are either limited to pilot programs (like the Accountable Care Organizations) or, if they are system-wide, tend to limit providers' risk to two percent or three percent of their reimbursements. Going up to nine percent [as the doc fix does by 2021] would step up the incentives in a pretty major way."

It's a supreme irony that this potentially radical long-term budget reform is strongly bipartisan -- put forward by the outgoing chair of the Senate Finance Committee, Max Baucus, and the Republican Chairs of the House Ways and Means and Energy and Commerce Committees, Dave Camp and Fred Upton. I wonder if it can stay that way. Partisanship will be in force on the unresolved matter of how to pay for the fix, which needs to be paid for only because the status quo baked an unsustainable "Sustainable Growth Rate" into long-term budget projections.** The collaboration on new payment mechanisms, however, is evidence that either a) Republicans are capable of rational thought and action when an issue is somehow sequestered from ideological dispute, or b) that both parties are in the grip of untested shibboleths regarding the potential of pay-for-performance and currently prevalent principles of its design. Probably a bit of both.

A program note: as I mentioned in a prior post, I am branching out from amateur blogging to amateur reporting.  On the prior subject on which I mentioned I would seek expert input -- "what will Republicans do to the ACA if they win both houses of Congress and the presidency by 2014?" --I have completed an article based on in-depth discussions with healthcare economists on both sides of the political spectrum that will either appear on this blog or elsewhere early next week. I plan to do the same with respect to the potential impact of the doc fix (that is, the SGR Repeal and Medicare Provider Payment Modernization Act). Let me again go open source here (my prior query drew a blogged response from Duke's Don Taylor) and pose the following questions:

1) The legislation's performance measures are to be based on self-reporting and created with extensive physician-group input.  Based on past experience, how susceptible to gaming are they likely to be -- as the AMA's Specialty Society Relative Value Scale Update Committee (RUC) has gamed current fee-for-service pricing? Are they at present designed, or can they be designed, to be reasonably proof against manipulation?

2) Like the current ACA pilot programs, the doc fix encourages formation of ACOs, medical homes, and other structures perceived to foster "coordinated care."  To what extent will the incentives further stimulate provider consolidation, and how can the tendency of consolidation to increase providers' pricing power and so raise costs be offset?

3) The ACO set up a bunch of pilot programs to test various ways to bend the cost curve. Does the doc fix move too fast in raising the stakes and extending the reach of untested cost control methods? What do you make of this complaint by Jeff Goldsmith of Health Futures, Inc. and the University of Virginia?
The SGR reform legislation is yet another triumph of slogans over evidence.  It builds upon a modified version of the current Physician Quality Reporting System (PQRS) system.  Critics have commented on the heavy reliance of the current PQRS and other federal value-based initiatives on process measures that are poor surrogates for actual patient risk reduction.  The Congressional plan actually ups the ante over the current PQRS and value-based modifier (VBM) programs by basing as much as 10 percent of an individual physician’s compensation on it. The legislation mandates that GAO evaluate the new Value Based Payment System in 2018, two years after it is implemented.  Is it too much to ask that we actually validate that a payment approach actually delivers claimed societal benefits before we enshrine it in legislation?
 Or this one? (by Robert A. Berenson, M.D., and Deborah R. Kaye, M.D.,cited by Goldsmith):
Although we agree that value-based payment is appropriate as a concept, the practical reality is that the Centers for Medicare and Medicaid Services (CMS), despite heroic efforts, cannot accurately measure any physician's overall value, now or in the foreseeable future. Instead of helping to establish a central role for performance measurement in holding providers more accountable for the care they provide and in informing quality- and safety-improvement projects, this policy overreach could undermine the quest for higher-value health care. Yet the medical profession has been remarkably quiet as this flawed approach proceeds...

physicians simply do not respect the measures, and for good reason. PQRS measures reflect a vanishingly small part of professional activities in most clinical specialties. A handful of such measures can provide a highly misleading snapshot of any physician's quality. Research shows that performance on specific aspects of care does not predict performance on other components of care. Primary care physicians manage 400 different conditions in a year, and 70 conditions account for 80% of their patient load. Yet a primary care physician currently reports on as few as three PQRS measures.
The authors provide some detailed examples of how proliferating quality measures fail to capture the most relevant factors in physician performance.  Like Goldsmith, they believe that performance level makes more sense at the practice level than at the individual physician level. Is this a focal point of debate?

4) What evidence exists that pay-for-performance models in medicine work? What are the best existing prototypes?  Ditto for "coordinated care," i.e., various degrees and incarnations of managed care. Austin Frakt here glances at a pair of literature reviews indicating that "Medicaid managed care is unlikely to lead to savings, in general." To what extent is that conclusion, as far as it's true, relevant to the formation of ACOs, medical homes, etc.?  Are there shining examples of savings in at least isolated Medicare Advantage programs?

5) Is there any correlation so far between performance on current CMS-administered quality measures and patient satisfaction or outside assessments?

Underlying these questions are a few underlying, and to some extent contradictory, hypotheses on my part (or say rather that they're in tension):

1.  All performance incentive systems are subject to gaming, i.e., to Campbell's Law.
2.  Government payment systems in the U.S. are especially likely to be corrupted by lobbyists, particularly in healthcare.
3. On the other hand, we can't do without statistical analysis, and fee-for-service badly skews doctors' incentives.
4. As a safeguard, the doc fix does stipulate constant adjustment of the performance measures and methods, with stakeholder input.

I welcome all comments, responses, invitations to talk, and referrals to experts or written sources.


* As far as I can determine, that last category (improvement) has two components: first, credit simply for engaging in approved "clinical practice improvement activities, i.e., credit for showing up; and second, improvement over past scores.

** The recent slowdown in healthcare cost growth has radically reduced CBO's estimate of how much the doc fix would cost. Those "costs" are in a sense imaginary, since they're relative to a Medicare payment schedule that would require a 23% cut this year -- a rate that by unanimous agreement will never happen. Thus the idea that a doc fix must be "paid for" is notional but also rests on our political third rail -- "entitlement cuts" vs. new revenue. 

UPDATE: Re paying for the doc fix, Austin Frakt flags a CBO menu of 16 cost-saving options to help get there. Tucked among the various squeeze-blood-from-a-stone possibilities is adding a public option to the ACA exchanges, which by CBO's reckoning would more than pay for the fix (saving $158 billion over 10 years; the doc fix will "cost" an estimated $150.4 billion). If both were to work as designed, that would be two major, major whacks at the healthcare cost curve, neither resulting in any cuts to benefits. Too bad Republicans, and a handful of Democrats, will never agree to a public option. Not in the foreseeable future, anyway.

An "organic" option working on Medicare itself is "Change the Cost-Sharing Rules for Medicare and Restrict Medigap Insurance," a sample mix of which would save an estimated $114 billion over 10 years, or about three quarters of the savings required.  That sample mix includes combining and raising the deductibles for Medicare Parts A and  B, adding a catastrophic cap to out-of-pocket costs (somewhat limiting the savings), and barring "first dollar" coverage in Medigap policies, leaving all Medicare beneficiaries on the hook for some costs beyond their premiums:
The third alternative combines the changes from the first two. Thus, all medigap plans would be prohibited from covering any of the new $550 combined deductible for Part A and Part B services, and the annual cap on an enrollee’s out-of-pocket obligations (including payments by supplemental plans on an enrollee’s behalf) would be limited to $5,500 in 2015. For spending that occurred after meeting the deductible but before reaching the cap, medigap policyholders would face a uniform coinsurance rate of 10 percent for all services, whereas Medicare enrollees without supplemental coverage would face a uniform coinsurance rate of 20 percent for all services. Those provisions would limit the out-of-pocket spending of medigap enrollees (excluding medigap premiums) to $3,025 and the out-of-pocket spending of Medicare enrollees without supplemental coverage to $5,500 in 2015. 
Economists generally favor such checks on uninhibited healthcare consumption (combined with catastrophic protection), and Obama's 2014 budget included some premium hikes for wealthier Medicare beneficiaries and a nibble at Medigap policies -- "a Part B premium surcharge equivalent to about 15 percent of the average Medigap premium for beneficiaries that purchase Medigap policies with particularly low cost sharing requirements" (p. 5).  He would probably be amenable to a plan something like the sample outlined by CBO -- if Republicans agreed to raise equivalent revenue.  That has always been the fulcrum of polarization: Obama will only cut entitlement benefits if Republicans raise taxes, which Republicans refuse to do.


No comments:

Post a Comment