Does Coordinated Care Save Money?

by Linda Gorman, NCPA Health Policy Blog

The claim that coordinated care will save money without degrading quality is a common belief behind almost all mainstream health reform efforts. Yet while the story sounds good, there is surprisingly little evidence to support it. According to the 2012 MedPAC report to Congress:

[F]indings from recent Medicare demonstrations on care coordination and disease management models have not shown systematic improvements in beneficiary outcomes or reductions in Medicare spending.

Of 29 Medicare care coordination programs tested to date, only 1 program reduced overall Medicare spending. Rates of improvement in clinical quality measures were “very low.” The report says that no one knows what the key elements of an effective care coordination strategy would be.

The granddaddy of Medicare coordinating efforts, the PACE program, has been operating since 1990. It integrates Medicare and Medicaid payments for elderly dual eligibles. In 2009, there were 21,000 people enrolled. MedPAC estimates that Medicare spends about 17 percent more on PACE enrollees than on comparable beneficiaries in fee-for-service care, but MedPAC could not “independently verify” the average amount spent on each enrollee. Since PACE quality measures are not publicly reported, MedPAC was “not able to conduct an independent analysis of PACE providers’ quality of care.” It still recommends expanding the program, after reforms to make it pay the same amount that is spent on fee-for-service beneficiaries.

Spending on dual eligible Special Needs Plans (D-SNPs) enrollees is also higher than spending on comparable beneficiaries enrolled in fee-for-service Medicare. MedPAC was “not able to conclude whether …[SNPs]… provide better quality of care” than fee-for-service because “quality data were not available.”

Despite the lack of evidence that care coordination works, CMS is working with states to develop more integrated care demonstration projects. In Oregon, it has agreed to finance a coordinated care experiment that will move all of the state’s 600,000 Medicaid clients into untested Coordinated Care Organizations. While MedPAC “supports the goals of these demonstrations,” it at least has the decency to wonder whether the frail elderly should be used as lab rats in yet another grand social experiment. Or, as it delicately puts it,

[W]hether care management models should be tested on large number of dual-eligible beneficiaries or entire subgroups within a state…the large scope makes the demonstrations appear to be large-scale program changes rather than true demonstrations.