Accountable Care Organizations Demystified

Accountable Care Organizations (ACO), what are they?  Will they work?  Can they be trusted in a dark alley?  This model is an increasing focus of many people’s attention.  Medicare, through policy, embracing the ACO signals a cautious return to the era of managed care and capitated payment structures popular in the 1990’s.  The purpose of ACOs are to reign in healthcare costs while maintaining or even improving care quality.  They are, however, contentious and some question the effectiveness of the model to achieve stated objectives.

Consider this an ACO crash course.  Using some of the more recent academic literature¹ I’ll help explain exactly what an ACO is, why it’s different from things we’ve tried before, and why exactly it may or may not do the things that we hope.  The explanation of ACO’s will be structured by the three attributes that distinguish it from other models of care: shared savings, accountability for quality, and free choice of providers by patients.

The basic model

Referring back to prior posts about reimbursement you may remember the term “capitation” or paying a flat rate to take care of a patient.  Capitation can happen in many flavors and on many levels.  For example one can capitate based on a disease (I’ll pay you a one-time fee of $500 to care for a broken ankle), one can capitate care for a person (I’ll pay you $50/month to care for my aunt), or finally you can capitate on a population level (I’ll pay you $1,000,000 to take care of a population of 5000 patients for one year).  ACOs are a form of capitation that occurs on the population level.

Medicare notes how much money an organization spent in a year to care for a large group of patients.  Using that number as a benchmark, Medicare promises to share any savings the organization manages to generate in the following year.  In other words if the organization’s spending for the group is less than the prior year, it splits the difference with Medicare.  In this way an organization can financially benefit from doing less (or being more efficient).  To ensure that organizations won’t “stint” on care, quality measures are put in place.  An organization only gets the full savings if it maintains a specified level of quality.

Shared savings

Many health management organizations (HMO) in the 1990’s approached cost reduction by attempting to control the flow of services.  Insurers required physicians to get pre-approval for more expensive services and also made attempts to control patient’s access to specialists.  This model was often frustrating for doctors who felt that insurers were trying to dictate how to practice medicine.  ACOs place care decisions back in the provider’s hand but construct financial arrangements which encourage judicious use of resources.

The ACO falls into a spectrum of different forms of risk-sharing between payers and providers – risk meaning who will have to pay if a patient gets sick and/or if care is expensive.  To understand better why ACOs are unique, discussing the range of payment approaches is useful.

Fee-for-service: the provider has no risk in this arrangement.  The payer pays a fee to the provider for every service performed.  In this model the provider is incentivized to do more rather than less.

Shared-savings: The provider is given a portion of savings – as measured by performing a service or group of services for less than a benchmark number.  If the provider spends over the benchmark number nothing happens.  In this way providers are incentivized to control costs through the potential of a bonus.

Shared-risk: Similar to shared savings, the provider gets a bonus for performing services for under a certain amount.  In this model, should the provider go over the benchmark amount, she must pay back a portion of the excess spending back to the payer; this is also called “two-sided risk”.  Cost-savings is incentivized through both the “carrot” of a bonus based on shared savings and the “stick” of a fine for overspending.

Capitation: Proper capitation is paying a provider a fixed amount for care before the care is delivered.  If the provider spends less than the amount then she keeps the entire savings – rather than just a portion as in the shared-savings model – and if the provider goes over the amount then she must shoulder the entire excess cost – as opposed to only a portion of the excess in the shared-risk model.

Generally speaking, the closer towards complete capitation a payment model becomes, the more powerful its ability to catalyze cost saving.  Many, if not most, organizations are deeply entrenched in fee-for-service relationships.  Insisting that those providers flip instantaneously to complete capitation is unrealistic – the ACO is a compromise.  ACOs allow various iterations of savings sharing and risk sharing, providing some of the benefits of capitation without being entirely unpalatable to organizations.

 

Both shared-risk and shared savings are layered on a base payment structure of fee-for-service.  Critics claim that the incentives generated by an ACO won’t be strong enough to meaningfully alter spending.  The findings of the Physicians Group Practice Demonstration (PGP), a 5 year ACO pilot program, are mixed.  Some participating providers achieved consistent savings, some only achieved savings a few years of the study, and still others achieved no savings for all five years.

Quality

As mentioned in the previous post, measuring quality was a trait of the leading HMOs but was not ubiquitous; some HMOs delivered low quality care as a way to gain more cost savings.  ACOs are unique in that all organizations are tracked on numerous quality metrics.  Only by maintaining a high level of quality can an ACO reap the “shared savings” manna. 

Most PGP participants consistently improved the quality of care across all measured metrics suggesting that ACOs are effective at improving quality of care.  There is concern that the quality metrics are inadequate to ensure appropriate care is being delivered; though the quality metrics were selected to represent a broad spectrum of care the theoretical possibility remains that providers may be able to selectively improve measured aspects of care while allowing other non-measured aspects slip.  Not enough quality measures and the metrics won’t meaningfully change care, too many quality metrics and providers may find the requirements overly burdensome and take a pass on forming an ACO.  Recently the required metrics were reduced from 64 to 32.

Choice

Patients, if assigned to an ACO, are still free to seek care from any provider they wish.  This is a contrast to a traditional managed care model with “in-network” and “out-of-network” providers – charging more if a patient seeks care out of network.  Why does this matter? 

How physicians are paid can dictate the style with which they practice.  Theoretically, a physician that is a part of a managed care system will be more judicious with resources and may be more likely to recommend against seeing a specialist, getting a high-tech test, etc. 

However, our prevailing cultural perception leans towards the “more is better” side of the equation and some patients want the high-tech test, specialist, etc. regardless of medical necessity.  Patients can seek care directly from specialists, regardless of what their primary care physician suggests. An out-of-network specialist may operate under entirely different financial incentives and lean towards a maximal approach to testing and care – which, though perceived as ideal by the patient, is costly and often inappropriate.

Managed care networks deal with this problem by creating a financial barrier; a patient that seeks care out of network needs to pay more thus encouraging patients to seek care from resource-conscious network providers.  In leading managed care programs, where care is responsive and quality focused, patients are often more than happy to make this trade – limited choice for more affordable care.  However, when HMO’s exploded, the new HMO breed lost the delicate cultural balance of economy and quality and patients felt locked into networks providing inadequate care; it’s one of the reasons why “HMO” is a bad word for some people and also the reason why ACOs have steered clear of limiting patient choice.

The flip-side of this coin is the provider who has no leverage in controlling how a patient consumes health care resources but is financially accountable for any care the patient seeks.  For example, if a patient bruises her knee but, instead of going to her primary care doctor, seeks care directly from a non-ACO orthopedist – who then orders a $1500+ MRI - the orthopedic billing is taken out of the ACOs shared savings potential.

Architects of the ACO fear that limiting a patient’s choice of providers will condemn the model to knee-jerk unpopularity.  Providers are concerned that very little leverage exists to control patient use of medical resources which limits the ACOs ability to significantly change patient use of health care resources.  Both are right.

Conclusion

The ACO, at this moment, is a game of carrots.  Many components are voluntary, the choice of providers to form an ACO and the choice of patients to seek care from their ACO physicians.  Medicare has to ensure that the deal is sweet enough – in terms of potential shared savings and ease of achieving quality benchmarks - to lure provider networks into forming ACOs.  Similarly, it seems that providers must do the same for patients; unable to require patients to see “in-network” providers, ACO’s may need to ensure accessible, efficient, and high quality care to keep patients from straying to other providers.  If done well, care may rise in quality and accessibility.  If done poorly, the ACO name may be tarnished as patients go elsewhere for care.

References

1. Berenson RA. Accountable care organizations in medicare and the private sector: A status update. 2011.