The Average Beneficiary Cost Component of Medicare’s VBP Measures

As Medicare attempts to measure and influence provider behavior, most of the programs it creates have flaws, and many of those flaws revolve around penalizing hospitals and physicians for factor out of their control, such as the underlying sociodemographic characteristics of populations they serve.   (HA Article)   The hospital readmissions reduction program has been shown to suffer from inadequacy in adjusting for this factor, for example.  CMS has recently begun including a measure of economic performance in its value-based purchasing modifiers for both hospitals and physicians.  This measure compares the average episode spending by a hospital to national averages and for physicians compares average spending on beneficiaries under their care with national means.  While these measures are adjusted for health status and some other factors, concerns have already been raised about whether they adequately protect providers from things that cause spending to differ, but they have little or no influence on.  In particular, these researchers looked at spending patterns for dual eligibles–people who have both Medicare and Medicaid coverage, a growing population.

Dual eligibles are only 18% of the Medicare population but they account for 31% of all Medicare spending.  They are a similar drain on Medicaid budgets.  If the various CMS reward/penalty programs don’t adequately adjust for the factors that cause higher spending in this group, the result is penalizing the providers who disproportionately serve these populations, taking away resources needed to deliver and improve care to an already underserved set of beneficiaries.  The researchers examined spending differences for dual eligibles, including by category of service and sought to understand if these differences were captured in the value-based cost measure and adequately adjusted for.  Focusing on hospital episodes, they found that 32.5% were for dual eligibles in the study period.  These beneficiaries were younger, more likely to be female, more likely to be disabled and more likely to be minorities.  They had significantly higher health risk scores and were more likely to be treated in safety-net hospitals.  Their spending was 4.3% higher per episode than the spending for non-duals.  This higher spending was across almost all major diagnostic categories.  They were more likely to have had care in the three days prior to admission, less likely to use post-discharge care and more likely to use post-acute institutional care.  Under the value-based purchasing measure, duals had $447 higher spending per episode than “expected” (as calculated under the VBP measure), while non-duals had $223 lower spending per episode than expected.  And it appears that the presence of more dual eligibles in a hospital’s treated population led to a greater likelihood of being found to have “excessive” spending.

One possibility, therefore, is to adjust the measure for the percent of a hospital’s dual eligible population.  Some object to this because they argue it may remove incentives to improve care for this population.  But that assumes that it is poor care and not other underlying factors that result in the higher spending and utilization for this population.  I think this is unlikely.  This is a population with a high rate of irresponsible health behaviors which have resulted in higher medical condition rates and resulting care and cost needs.  The cause of those behaviors is a complex social issue and I am not sure why providers should be held responsible for the results of the behavior or for trying to fix it, which is usually futile for most of these patients.  The diseases are too advanced and the behaviors too ingrained.  And penalizing them does take away resources that are needed to care for this population.