Are ACOs A Financial Suicide Pact?

One of the key themes of The Road to Hellth is how complexity and trying to outsmart market forces always seem to lead to economic inefficiency and poor choices in the U.S. healthcare system.  The Center for Medicare and Medicaid Services (CMS) must have known that we needed a new example for today’s post, because they were good enough to release their brand-new 429-page proposed rule regarding Accountable Care Organizations (ACOs) just before press time.  Despite all the high hopes, there are clear indications that, at least for healthcare providers, ACOs are going to look, smell and taste more like cold, refreshing Jonestown Kool-Aid than “Quality Care”™-brand champagne.

But first a brief introduction for our readers who may not know what all this is about.  Interest in ACOs is one result of the Affordable Care Act “ObamaCare” legislation that passed into law just a little over a year ago.  Simply put, an ACO is a collection of healthcare providers who contract with Medicare to take the responsibility for taking care of the needs of 5,000 or more patients under the auspices of Medicare’s new ACO payment program that will begin on January 1, 2012.  Under the terms of this program, the providers (who may or may not be owned by hospitals) agree to adhere to certain “quality of care” requirements while they’re taking care of these patients for a three year period.  For each year of the period Medicare will compare the amount that they actually spent caring for these patients with some predicted amount that they were expected to spend.  If less is spent on these patients than expected – while still adhering to government-mandated quality standards –  the government would share the savings with the members of the ACO.  The expected result: high quality, happy patients an oodles of new money for doctors, hospitals and other healthcare providers.  After all, with a plan like this, what could go wrong?

Well, one thing that could go wrong is fiscal reality.  Because if we were betting people, we’d like to short the stock of just about all of those ACO-participating doctors and hospitals right now.  It’s probably the only way that anyone other than the government is going to make a profit on this.  Here’s why.

It turns out that we already know something about how ACOs might be able to perform financially.  As a recent article in the New England Journal of Medicine points out, Medicare has been testing the ACO concept since April 1, 2005 when it launched its Medicare Physician Group Practice  (PGP) Demonstration program.  PGP was Medicare’s first pay-for-performance program to incorporate all of the elements that the ACO initiative is supposed to include: having groups assume responsibility for the care of large numbers of Medicare patients, coordination of care, quality measures and rewarding the minimization of total costs by sharing “savings” with the participating physician groups.  For the purposes of this demonstration, ten highly-motivated, highly-prepared physician groups were selected on a competitive basis.  Seven were located in the Northeast and Midwest; with Montana, Washington state and North Carolina thrown in to give it a little geographic diversity.  As Medicare’s fact sheet on the program explains, “Multi-specialty physician groups with well-developed clinical and management information systems were encouraged to apply since they were likely to have the ability to put in place the infrastructure necessary to be successful under the demonstration.”  In other words, all of the participating sites were meant to be ringers, thereby providing the highest possible likelihood of success for the program as a whole.  Not that there’s anything wrong with that.  That sort of selection bias is perfectly fair for a demonstration project.

But even those top-flight physician groups weren’t really ready for all that CMS was going to demand.  On average, each of these groups had to install an additional $1.7 million in infrastructure in the first year alone.  That was an average of about extra $737 per provider, but each program undoubtedly thought that it was going to be a good investment.  They were thinking about all of the extra money they would make by sharing the cost savings they would achieve for Medicare.  And Medicare was feeling mighty generous.  Based upon a combination of cost savings and meeting or exceeding goals for 32 different “quality” measures, participating physician groups could earn performance payments of up to 80% of the total cost savings, with Medicare getting the balance.

(Unfortunately, the term “cost savings” in this case was a bit of a misnomer, since the physician groups would normally have earned guaranteed extra money from their performance on quality measures under the completely separate pay-for-performance (P4P) Physician Quality Reporting Initiative (PQRI) program.  But for the PGP demonstration projects these funds were thrown into the “savings” pot and placed at risk depending upon whether the groups hit their new PGP “quality” targets.  If their patients were harder than expected to care for in the course of any given year, then the physician groups would essentially be footing the bill with its own PQRI earnings.)   Despite this, each of the ten selected test sites bravely forged ahead.  Overall, the effort represented some 5,000 physicians and 220,000 Medicare fee-for-service beneficiaries.

And the results?  They were great!  At least for Medicare.  By the end of the fourth year, all ten of the physician programs achieved benchmark performance on at least 29 of the 32 quality measures.  Two of the groups achieved benchmark scores on all 32/32.   Most programs improved on their own baseline quality scores for such tough and expensive diseases as diabetes, high blood pressure, heart disease and cancer screening.  America has good reason to be proud of those doctors and their clinical performance.

On the other hand, the financial results weren’t nearly as inspiring for the participating physicians.  In Year 1, just two out of the ten provider groups (20%) got to share in any financial “savings”, including the PQRI money that had been put into the risk pool.  In Year 2, four out of the ten groups (40%) got to share in performance payments, while in Years 3 and 4, just half of the ten groups (50%) saw any money at all out of the combined “savings” and “quality” performance fund.  Considering the $1.7M Year 1 infrastructure upgrade spending plus whatever additional monies were spent on additional infrastructure, administration, software and reporting costs in Years 2 & 3, the net result for most of the participating physician groups appears to have been a substantial financial loss when compared to what they would have earned from straight fee-for-service practice.  And as anyone who has ever dealt with government P4P programs can tell you, that does not even begin to take into account the psychological and stress costs of program reporting and administrative overhead.  Someone worked some very hard hours to generate those quality score improvements.

Based upon these results, Drs. Trent Haywood and Keith Kosel built a financial model to estimate the kind of operating margin that a future ACO might need in order to simply reach breakeven in a setting identical to that presented by the PGP.  Their conclusion: each group would have to generate a whopping 20% margin over a three-year period – and that’s only considering the additional infrastructure costs for the first year alone.  Extending the program pay-off period to five years would reduce the operating margin needed to 12.7%, while going all the way out to 10 years would bring it down to a mere 7.6%.  To put that in perspective, the average U.S. hospital operating margin for the period 1997-2004 was only 4.31%, about one-fifth the level that would be conservatively needed to reach breakeven for a typical 3-year ACO contract.  With some understatement, and without knowing the actual proposed terms of ACO cost sharing that have only recently been announced, Haywood and Kosel called the likelihood of achieving a 20% operating margin for ACOs “unlikely”.

Fortunately, demonstration projects like the PGP are done for a reason.  With these results in had, it was possible for CMS to go back and craft a proposed rule for ACOs that would allow hospital systems and other physician groups to have a fighting chance of tasting both medical and financial victory as a result of their participation in this landmark cost-saving program.  Let’s see how that turned out by exploring the exclusive Road to Hellth PGP-toACO Program Comparison Matrix shown below:

(Click On Table to Enlarge)

Oops.  It looks as if all of those pioneering providers participating in the new cost-saving ACOs may need to tighten their belts.  Far from making the requirements easier and financial incentives more favorable for participants, the rule-makers at CMS seemed to have pulled a switcheroo.  Compared with the PGP demonstration program, it’s going to be much riskier and tougher to see a positive financial return.

• Instead of four years to recover any initial infrastructure development, ACOs will have only three before their contract term expires.  This means that the average operating margin that they see as a result of any cost savings to Medicare is going to have to be up at that “unlikely” 20% level.

   

• Saving just a little bit of money is not going to do you any good.  To ensure that any savings are real and not just the result of random fluctuations, Medicare is setting a lower “savings” limit of at least 2% before they start sharing any benefits with the ACO.  There was no similar minimum for PGP project participants.

   

• Instead of ACOs getting a maximum 80:20 split of the savings generated, Medicare is going to hang onto a minimum of 50-60% of all proceeds, depending upon how much additional financial risk each ACO is willing to take on.  This would be done by sharing not only any savings, but also any “losses” that may occur when an ACO’s patients cost more to care for than the amount Medicare had originally predicted.

   

• Loss sharing is going to be mandatory for all ACO participants.  The only question is how big their exposure is going to be.  The old PGP demonstration had no such provision for having doctors help Medicare cut its own losses if its patients happened to become financial black holes.

   

• Finally, the number of mandatory “quality” measures that ACOs must meet and report on has more than doubled, from 32 measures to 65 measures.  Simply tracking and reporting on these is going to cost far more time and overhead expense when compared to the PGP program.

Even the most casual observer can look at the right-most matrix column and see that the financial effect of each of these program differences is going to be: (a) cumulative; and (b) bad.  Start-up costs for most prospective ACOs will be higher than for PGP participants given the added complexity of the program and the number of quality measures included.  While only half of all PGP participants were ultimately able to share in any cost savings by the end of Year 3, the new program’s stiffer requirements mean that the typical ACO’s chances of securing a financial windfall are likely to be substantially lower.  Even worse, the financial up-side of any savings that do materialize will be smaller, while adding a brand-new risk of actually helping to insure Medicare against its own unexpected losses. And there is one final risk that we’ve not yet discussed, but should be on everyone’s radar screen.  Medicare’s track record so far has been one of systematically turning “pay-for-performance” into “perform-or-punish” programs.  (The financial incentives for purchasing electronic medical record systems in the HITECH Act are a prime example.)  If ACOs actually do achieve substantial financial savings in their first three-year contract period, in the next contract period Medicare can be expected to either reduce the expected level of expenditures or increase the percentage of savings that it retains for itself.  In either case, the shared savings realized by participating ACOs will fall over time, while their corresponding risk of unexpected loss increases or remains the same.

The bottom line?  Dr. Berwick and his colleagues at CMS appear to have taken the ACO concept and made it into a financial program that only delusional practice administrators, or physician organizations bent on financial self-destruction, could love.

No related posts.