Several weeks ago I attended an HFMA seminar where yet another speaker presented a “financial analysis” of projected Medicare ACO performance. This was repeated at another recent seminar at which the presenter commented that there was “no downside“ to ACO participation, and again in a major financial magazine that presented “savings” as if it occurred outside of the ACO’s group of providers. All of these use contrived data, of course, since there are no actual results yet available, and for some reason they all seem to show the ACO “making money” in the end, with “making money” being defined as receiving a shared savings payment from CMS.
I’d define “making money” a little differently – to me it means that the participating providers realize a higher net income from participating in the ACO than from not participating, holding all other factors – especially market share – equal. This means that shared savings from the ACO must be balanced with the reductions in payments to providers that created those savings.
The only recent recognition of this is this Health Affairs article ("The ACO Race Is On: Navigating The Terrain”) . Here authors Lawrence Casalino and Stephen Shortell note that “… the shared savings “bonus” is not really a bonus. Under the program as designed even an efficient, high quality ACO will gain less money from sharing in savings than it would have earned if it had simply continued with business as usual.” Let’s look at the total financial picture of providers and ACOs from that point of view.
The Shared Savings Model
The shared savings model for ACOs is diagrammed below. Savings accrue through reductions in fee for service payments to providers to a level below the amount targeted by the payer’s savings model. Some savings are shared with the payer with the remainder going to the ACO. The ACO will then distribute the remaining amounts to the ACO members according to the distribution formula developed by the ACO. The ACO members may vary between organizations; the example below assumes that the ACO includes all providers.
The opportunities to create shared savings are generally related to the relative spending amounts by provider type. From the chart below, hospital inpatient care represents the largest share of spending, making it a primary candidate for reductions.
As an oversimplified example, consider an avoided admission resulting in a $4,000 decrease in payments to the hospital and a $1,000 reduction in physician payments. In a Medicare ACO, the ACO organization would receive approximately $2,500 of that amount, with the remainder going to CMS. But the ACO would distribute that savings among the ACO’s participants, including physicians and possibly post-acute providers. Assuming a 50/50 split of ACO savings between the hospital and other providers, the hospital would receive $1,250 of the $4,000 in FFS payments that it had lost. The physicians as a group would receive $1,250 having lost $1,000 of FFS payments.
While this arrangement is disadvantageous to the hospital, it is advantageous to physicians as a group, who had lower reductions in payments but enjoyed a higher percentage of the hospital’s payment reductions. The actual physicians whose admission had been avoided would be out the $1,000 FFS payment, however.
In a physician-only ACO, the model would be significantly different. Here the hospital savings would be shared only with the physicians, resulting in a $2,500 payment.
Thus, a physician-owned ACO could create significant shared savings by leveraging reductions in the hospital’s FFS payments. The problem with this model is that few physician groups have the capital and infrastructure to manage the health of a population-based healthcare system. CMS estimates the infrastructure cost of an ACO at $1.7 million; other estimates are much higher. In addition, the “leverage” works both ways – the shared loss from an increase in hospital admissions would be borne entirely by the physicians in a physician-only ACO.
Opportunities for Net Savings
Given that the shared savings accrue from incurring costs below a target that’s based on actual historical provider payments, it appears to me that there are only three ways in which an ACO could “make money” according to this definition:
1 – The ACO participating providers are somehow able to reduce payments to other providers that are outside of the ACO. This would result in lower provider payments and thereby might trigger a shared savings payment to the ACO, but the ACO’s providers themselves wouldn’t have been the source of that savings. This could occur, for example, if the ACO didn’t include specialists and the PCPs and hospital were able to significantly reduce specialist utilization in a way that didn’t reduce hospital utilization. This seems a little difficult to achieve…
2 – The inflation rate used to increase the cost targets is disproportionately higher than the actual rate at which the ACO’s provider payments would increase over time. Therefore, the “shared savings” that were calculated would be illusionary – those costs wouldn’t really have been incurred – and therefore there would be no real decrease to payments to the ACO’s providers. These types of data anomalies have occurred in the past (I think they may have caused some of the PGP demonstration’s results) but certainly couldn’t be counted on for many ACOs.
3 – Payments to the ACO’s providers really do decrease, but the providers are able to decrease their costs proportionately, so that the effect on net income is zero or positive. This would require the providers’ marginal costs to be higher than their marginal revenue, which is unlikely.
Casalino and Shortell propose several other reasons for hospitals to form ACOs. They note that hospitals that are operating at capacity may be able to fill the beds resulting from the decreased Medicare utilization with more profitable commercial patients, and that building cooperation with physicians may assist in achieving reductions in readmissions and increases in quality measures, which can affect payment from government and commercial payers. An ACO also provides a vehicle for funneling resources to physicians; for example to purchase EMR systems.
And there are many other non-financial reasons for ACO formation. But providers contemplating forming ACOs need to look at the complete financial picture, including the loss of FFS revenue.