At a recent meeting of the American College of Physicians covered in MedPage Today, Neil Kirschner, PhD reported on the growth of ACOs. Dr. Kirschner is the ACP’s senior associate of regulatory and insurer affairs. He reported that in March, 2012, there were only 136 ACOs that had ACO-style gain-sharing contracts with Medicare, a commercial payer or both. Those ACOs included 91 “hospital-based ACOs” and 45 that were described as “physician-led ACOs.” In the last year, the number of ACOs has almost tripled from 136 to 391. And, significantly, the 202 physician-led ACOs now outnumbers the 189 hospital-based ACOs.
Why are physician-led ACOs growing faster?
The rapid growth of physician-led ACOs makes sense because ACO-style gain-sharing reimbursement arrangements are more favorable to physician organizations than hospitals. When a hospital-based health system enters into an ACO contract, it basically says “if you succeed in reducing the total cost of care (primarily by reducing your own hospital revenue), we’ll give you back a portion of your lost revenue.” That’s reminiscent of the old business joke “we lose money on each unit sold, but we’ll make up for it in volume.” But, from the point of view of a physician organization, particularly a physician organization that emphasizes primary care, the ACO contract basically says “if you do more primary care work to avoid the need for hospitalizations and emergency department visits, we’ll pay you for that extra work and we’ll give you a portion of the money we no longer have to pay to the hospitals.”
So, why would hospital-based organizations ever enter into an ACO-style gain-sharing agreement? Two main reasons.
- First, as a defensive measure. By forming an ACO in which they are involved, they can avoid or delay the formation of a physician-led ACO. It is better to be “doing” than to have it done to you.
- Second, as a way to prepare for a potential future state where providers are bearing far more risk, such as in capitated reimbursement arrangements. It takes a very long time to develop effective population management capabilities, including establishing effective governance, building trust, deploying the right health information technologies and analytic systems, and, most importantly, recruiting and developing human resources that can effectively use data and improve care processes.
Ambiguous ACO Terminology
Finally, a note about the use of the term “ACO” and ACO growth statistics.
Many people get confused about the definition of ACO. Some argue that an ACO is, by definition a specific type of Medicare reimbursement arrangement, the Medicare Shared Savings Program (MSSP), created by the Affordable Care Act (ACA). But, I don’t agree that the term “ACO” defines a type of contract, and that the term ACO only refers to Medicare. The “O” is for organization. On its face, the term refers to a type of organization, not to a type of contract. When people generate statistics about ACOs, they should be counting organizations, not gain-sharing contracts.
I consider a provider organization to be an ACO if:
- It includes primary care physicians, and
- It has or plans to enter into at least one contractual arrangement where the provider organization takes responsibility for an attributed population of patients, and bears at least some risk for quality and economic performance
In common use, people are most likely to refer to a provider organization as an ACO if at least one such contractual arrangement includes a fee for service component with some risk sharing that includes at least some up-side opportunity to share in savings. The Medicare MSSP and Pioneer ACO contractual arrangements are just prominent examples of such contracts, not limiting to the definition of ACO. Because of that fee-for-service connotation, people don’t tend to describe a staff-model HMO as an ACO, even though they meet the definition I proposed above.
Furthermore, when categorizing ACOs, it is important that the categories are mutually exclusive. The term “physician-led” seems to refer to an attribute of the governance of the ACO, such as whether the person serving as the CEO is a physician or whether the majority of the shares or votes on the board of directors are controlled by physicians or by people that are employed by physician organizations. The term “hospital-based” seems to refer to whether the provider organizations that own the ACO include at least one hospital or hospital-based health system. An ACO can include a hospital as a co-owner and still be “physician-led.” Therefore, there is inherent ambiguity in the statistics comparing the number of physician-led to hospital-based ACOs.