Editor’s note: This post is part of a Health Affairs Blog symposium stemming from “The New Health Care Industry: Integration, Consolidation, Competition in the Wake of the Affordable Care Act,” a conference held recently at Yale Law School’s Solomon Center for Health Law and Policy. Links to all posts in the symposium will be added to Abbe Gluck’s introductory post as they appear, and you can access a full list of symposium pieces here or by clicking on the “Yale Health Care Industry Symposium” tag at the bottom of any symposium post.
The definition of an accountable care organization (ACO) involves two elements: organization and payment. First, an ACO is organized as one entity capable of taking both clinical and fiscal responsibility for care. Second, ACOs operate using a payment model centered on a budget target: if the ACO spends less than the budget target, it shares in the savings, while, in some models an ACO that exceeds its budget target might be required to pay more.
ACOs differ from past models like health maintenance organizations (HMOs) and preferred provider organizations (PPOs) in various ways. Most importantly, they are provider, not health plan focused. In ACOs, people are attributed to the model; they do not sign up as in insurance plan. Furthermore, ACOs, unlike insurance plans, do not set premiums, nor do they control benefit design.
Evaluations of a commercial ACO, the Alternative Quality Contract (AQC) in Massachusetts, found that, several years out, ACOs were able to save about 10 percent, half of which is based on reductions in price, suggesting that physicians will shop if they are incentivized to do so. This illustrates a market-based method of dealing with price concerns, which exists in addition to the regulatory method of addressing price.
In the Medicare program, where there is much less price variation and much less experience with ACO models, results have been more similar to AQC results in its first year, with savings slightly over 1 percent. Admittedly this is likely a small percent of the possible savings, but there is considerable room for improvement. Empirical results show that most of those savings come by reductions in post-acute care and shifting the site of care from hospital outpatient departments, which are relatively highly paid, to physician offices that are paid less for the same services.
Furthermore, the evidence is strong that quality is not diminished in ACOs. In fact, the sickest patients tend to do the best in ACOs. ACOs reach out to the sickest patients and touch them in a greater variety of ways, leaving patients happier with that level of interaction. Thus, both quality and patient satisfaction are enhanced in ACOs.
The financial model of ACOs works because, by reducing utilization, ACOs avoid the variable cost associated with care, which can be shared between the ACO and payers. The result is that, by reducing utilization of services that would otherwise be delivered, both the ACO and payers improve their financial profile.
ACO implementation, however, is not without challenges. Consider a hospital-based ACO. Reducing hospitalization allows the hospital to eliminate variable costs, but does not address the fixed costs of hospitalization. Thus, only the variable portion of the cost of hospitalization is returned to the organization. In a non-hospital-based ACO, on the other hand, a reduction in hospitalization saves the ACO the entire price of that hospitalization, which is much more lucrative. Moreover, saving money by reducing hospitalization is more difficult if the ACO has a lot of hospital dominance in the decision making process.
However, there remain a number of reasons why it is beneficial to include hospitals in ACOs. First, hospitals may improve integration and coordination efficiencies between primary providers and specialists, and between hospital care and post-acute care. Second, hospitals have more access to capital, which facilitates data collection and infrastructure. Third, hospitals are large, heavily managed organizations which may help them develop programs to succeed in the new model.
Another challenge is that saving money in an ACO by reducing specific services will also affect other non-ACO subsets of individuals that those organizations serve, for whom a reduction in use of the services at issue may not be desirable. For example, a hospital that implements a readmission reduction program may make money by reducing readmission of its ACO patients; however, this could also cause the hospital to lose readmissions from fee-for-service (FFS) patients, creating a complicated dynamic that may or may not ultimately produce a profit. Empirically, in the Pioneer ACO program, hospital-based ACOs have savings similar to those in ACOs that are not integrated with hospitals. Therefore, it is unclear whether the aforementioned advantages of hospital-based ACOs outweigh the fundamental incentives in a large physician-based non-hospital-oriented ACO to save money.
Finally, the financial success of ACOs depends on the percentage of shared savings and the cost of operating the ACO. It is not clear whether, with current parameters, that the shared savings will be enough to offset the operating costs.
Given these challenges, it may appear that organizations are likely to shun ACOs. Yet they participate, not because the new model is some sort of panacea or because ACOs are easy to administer: Running an ACO is very difficult, and ACOs may not ultimately prove to be successful. However, organizations continue to gravitate toward ACOs because the alternative looks so bad. For instance, we have placed the trajectory of fees in the FFS system on a very low-growth path. It is not clear how well organizations can succeed with fees from public payers growing lower than inflation.
The strongest case for the ACO model may be that it allows organizations to capture the efficiencies if they can achieve them and thus succeed financially with lower revenue growth. Thus while it is not certain that organizations can make this transformative turnaround, the alternative looks particularly bad.