Looking at Bundled Payments From an Anesthesiology Perspective

Bundled payments are coming.  Are there any readers who have not heard?  But do we know what “bundled payments” might mean for anesthesiologists and pain physicians?

Although there are certainly anesthesiologists participating in bundled payment systems—common examples include global surgical packages or OB packages—no one has written a field guide for the specialty.  The chapter on bundled payment for post-acute care strategies in MedPAC’s June 2013 Report to Congress provides a thorough and up-to-date review of general bundled-payment design issues.  More important, Congress will consider MedPAC’s recommendations in any new legislation on the topic, much as it did when it wrote and adopted the Patient Protection and Affordable Care Act of 2010, including a provision requiring Medicare to test a bundled payment approach.

In a bundled payment methodology, a single, bundled payment covers all of the services delivered by two or more providers during a single episode of care or over a specific period of time.  In an orthopedic case such as a total knee replacement, for example, a bundled payment could cover the hospital charge, the surgeon’s and the anesthesiologist’s fees, and, if applicable a stay in a rehabilitation facility and physical therapy.

In designing a bundled payment system, or evaluating a proposed design, the parties need to address the scope and duration of the bundle of services to be provided, the form that financial risk-sharing will take, the benchmark against which spending will be measured, how the payments will be allocated between the providers, and quality metrics including those that will prevent under-treatment. 

For each of these variables there is a potential trade-off between exposing the providers to financial and clinical risk for care furnished by other providers and increasing the opportunities for coordination of care.  Payers (and policy-makers) will want to foster more coordination of care.  Providers will want to limit their risk by keeping as much control of the processes of care as possible, e.g. by limiting the scope and duration of the bundle.  Anesthesiologists may not want to assume responsibility or risk for anything that happens to the patient after discharge from the recovery unit, but the payers and the hospitals could seek to include pain management services for the duration of the bundle. 

Although the MedPAC discussion focuses on bundling post-acute care with inpatient care, raising many issues that do not involve anesthesiology, the trade-off analysis for the variables studied in the Bundled Payments chapter is instructive.  

1.  Scope of Services

The greater the number and types of services included in the bundle, the greater are the opportunities for coordination.  Conversely, if a narrow set of services is bundled, the individual participating provider or physician has greater control and lower financial risk.

The Illustrative broad-scope bundle discussed by MedPAC begins with an initial hospital stay; spans 90 days after discharge; and includes any potentially avoidable readmissions, post-acute care and physician services furnished during the hospital stay and during any institutional post-acute care (skilled nursing facilities, inpatient rehabilitation facilities, and long-term care hospitals).  This bundle would tend to appeal to payers because “providers would have strong incentives to coordinate care across PAC settings, carefully manage care transitions, and refer beneficiaries to providers that minimize the risk of readmissions.”  (Report to Congress, p. 69.)

The illustrative bundle tracks the services that make up one of the models in which 69 entities are participating under the CMS Bundled Payments for Care Improvement Initiative.  A second CMS bundled payment model is narrower, excluding inpatient care and covering only  post-acute care, physician services, related readmissions and other Medicare Part B services. 

While the CMS Bundled Payments initiative allows the participants to select any of 48 clinical condition that make up an ad hoc collection of MS-DRGs, the MedPAC analysis focused on just 10 conditions with high rates of post-acute care utilization.    One of the oldest bundled payment systems in the country, launched in 1984 by the Texas Heart Institute and Cardiovascular Care Providers Inc., was initially limited to cardiovascular surgery.  It has since expanded to invasive cardiology procedures.  Services of the following specialties are included in the bundle:  cardiovascular surgery, cardiology, anesthesiology, pathology, radiology, pulmonary medicine, infectious disease and nephrology.

A bundle could theoretically be as narrow as anesthesiology and major joint replacement surgery.  It might not get much attention from payers, however, because the orthopedic surgeons and anesthesiologists on their own—not including even the hospital—would have limited incentives and ability to achieve better outcomes at lower cost through coordinated care.

2.  Duration of Services

Setting the duration of the bundle invokes the same trade-offs as the definition of the services to be included.  Long bundles—90 days in the case of the illustrative MedPAC bundle including post-acute care and readmissions—hold the providers responsible for more services and entail the likelihood that services furnished at the end of the period will be unrelated to the original hospital stay.  Short bundles, e.g. 30 days, have less variation in the services utilized, are more predictable, and can therefore be more attractive to the providers.

Payers will want to discourage providers from delaying care until after the bundle period has ended.  In CMS’ bundled payment initiative, spending during the 30 days after the end of the period will be compared with aggregate historic spending trended forward. Providers will be at risk for spending above a set threshold.

Anesthesia care traditionally ends upon discharge from the recovery unit.  One way to lengthen the service is to bundle post-operative pain management, both inpatient and outpatient.  Because anesthesia can play an important part in reducing the number of surgical and ICU readmissions (see ABC Alert, Monday, July 1, 2013), groups may want to consider participating in certain bundled payments that extend for 30 or more days after discharge.

3. Risk Adjustment

Risk adjustment is still an underdeveloped science.  MedPAC used an additive approach involving three types of risk adjusters—MS-DRGS; the patient’s comorbidities recorded during the year before the trigger hospitalization (using clinical risk groups); and, for patients with patient assessment information, functional status at admission—which together explained only 36 percent of the variation in resource use (as measured by charges) across bundles.

Anesthesiologists considering bundled payment methodologies should likewise benefit from additive approaches to risk adjustment.  We expect that the research community has already delved into the role that ASA patient status modifiers might play.

4. Payment

There are two basic options for distributing payments to the participating physicians and providers:  one is the payment of an all-inclusive amount to a single entity (such as the hospital, or a third party escrow agent) and the other would continue to pay the participants individually on a fee-for-service basis, with a withhold or some other sort of incentive.  Payment to a single entity offers simplicity to the payers, but it requires a complex and expensive infrastructure for allocation of payments on the provider side—not to mention a good deal of trust.  Fee-for-service methods (frequently referred to as “virtual bundling”) are likely to be more attractive to anesthesiologists and other physicians for now.  The advantage of virtual bundling is that:

[I]t does not require a single entity to receive payment for a collection of services, establish rates for other providers, and administer payments to them, thus making it practicable for most providers. It also avoids the thorny policy issues of how to attribute responsibility for episodes (since all providers share it) and which provider would receive the bundled payment.

(Report to Congress, p. 76, citations omitted).

The most common payment strategy is a shared savings model that offers a financial incentive for the provider to reduce healthcare spending to a percentage of the negotiated bundled payment rate by allowing the provider to share in the savings realized. The provider is not at risk for costs that exceed the bundle rate. Many providers test the waters by first entering shared savings models where they are protected against downside risk.  If and when the providers gain confidence in their ability to manage costs, they may progress to a full risk model, in which the provider is responsible for all of the costs above the negotiated rate, but can also retain all of the savings realized as a result of their efforts.  Given the immature state of the market, anesthesiologists will probably prefer the one-sided, shared-savings model.

5. Setting the Spending Benchmark for the Episode

The spending benchmark will be in effect a discount from current cost or spending levels.  The benchmark should fall somewhere between the point where it is too difficult to achieve the savings targeted and the point where the potential shared savings amount creates little incentive to coordinate care.

Two approaches to setting the benchmark so as to create incentives to increase efficiency involve looking at spending levels across providers, or across different geographic areas.  In MedPAC’s model, the Commission found that “Given the variation in readmission rates across providers within and across sectors, a 10 percent reduction in spending associated with readmissions might be an appropriate initial value for the episode benchmark.”  Alternatively, the benchmark could be set at the mean of the national average spending level and the average for low-spending metropolitan areas.  It could also be set at a given level below the mean, e.g., the 40th percentile.

An anesthesiology group contemplating a bundled payment would need to begin by determining its own average costs as well those of the other participating providers and making a judgment as to potential savings.  

6. Quality Measures for Bundled Payments

Bundled payments shift some of the financial and clinical risks of care to the providers—and the providers, in turn, can shift some of those costs back to the payer, or to the patient, by such means as stinting on services provided, cost shifting outside the bundle period, and increasing the volume of bundles.  The drive toward saving costs can also impinge on the quality of care coordination and transitions and on patient experience.  Payers will therefore seek to include quality measures that counter such efforts.

The set of measures proposed by MedPAC to measure and maintain quality are designed for payment bundles covering inpatient services and 30 days of post-acute care.  Many of them can be adapted to anesthesiology, including unplanned readmissions (and unplanned admissions to the ICU or returns to the OR) and patient surveys on pain management, provider communication and shared decision-making.  The full list from the Report to Congress (p. 79) is reproduced here to stimulate our readers’ imagination.

 

Conclusion

Bundled payments are more theoretical than real for most anesthesiologists.  As a way to restrain the costs of fragmented care and as a stepping stone to future accountable care organizations, bundled payment systems’ appeal to payers—commercial as well as Medicare—is undeniable.  MedPAC’s initial study of design issues in creating bundles that span inpatient and post-acute care points to some of the major questions that anesthesiologists will need to resolve to prepare for successful participation in bundled payment packages.

Other important questions include attribution of patients, addressing variable patient acuity levels, dealing with multiple payers and different bundled payment systems, and, more fundamentally, “whether the physician will be given information sufficient to enable the physician to calculate independently bundled payment amounts and determine whether or not those amounts are appropriate given the enrollees the physician is responsible for treating.”  (AMA, Evaluating & Negotiating Payment Options, Bundled Payments).  We would very much like to hear from our readers as you face these and similar questions.