ACOs set to expand cost savings

The rap on accountable care organizations (ACOs) is that there is no evidence to demonstrate they will control costs. I’d say the evidence is right under our noses – right in an analysis by the Congressional Budget Office (CBO).

Earlier this year, the CBO published an assessment of the Centers for Medicare & Medicaid Services (CMS) demonstration projects on disease management, care coordination and value-based payment. This demonstration project provided a venue for disease management suppliers to demonstrate their programs’ ability to impact the costs of Medicare beneficiaries. The CBO was unimpressed with the value of these programs for Medicare beneficiaries. But medical professionals engaging in the physician-directed population health management typical of ACOs will see a different message in this same experience.

One sub-group of the CMS demonstration pilots produced dramatically better results than the group as a whole: the pilot programs that included substantial direct interaction between care managers and physicians. With these types of programs, patient outreach – both the timing and the condition focus – was determined by treating physician discretion rather than automated system prioritization. These programs, in fact, produced impressive results. Hospital admissions declined 7 percent and overall costs declined 6 percent among Medicare beneficiaries enrolled in pilots that included this type of collaboration. That’s big. Using the projected $8,600 average cost per fee-for-service Medicare beneficiary in 2010, that’s a savings of more than $500 per person, per year.

Of course, this will be no surprise to most physicians, and it is exactly the essence of the ACO-based population health management approach. Yet this success was hardly noticed. Why?

The CBO valued the pilot programs by savings net of program costs. That’s the right way to do it, except when you are measuring pilots designed to identify new strategies for further exploration.

These pilots were implemented under very tough conditions. There was little data shared on the Medicare population being managed. Only very small portions of assigned Medicare beneficiaries were included in the pilots for each disease management program supplier. Finally, there was no assurance provided that any up-front investment might be rewarded by a longer program engagement.

No data, no scale and no meaningful client commitment to program duration leads to fees in excess of value. In short, program costs were far higher in these demonstration projects than they would be in an ongoing program, such as in a well-functioning ACO.

The CBO indicated that program costs averaged at least 11 percent of beneficiary costs – meaning the pilot programs on average cost more than $900 for every Medicare beneficiary enrolled in the plans annually (based on our same 2010 Medicare cost estimates). Of course, this is an incredible amount of program expense, much more exaggerated than real program expenses in similar, more mature programs, such as Medicare Advantage programs.

What this CMS experience shows is that aligning care management programs to physician care planning lowers costs. The evidence already exists. Accountable care organizations are set to expand that proof.