It is a well-recognized phenomenon that a relatively small subset of patients can account for a high percent of all health care utilization and spending in the United States. The same appears to be true for most payer sub-populations. The Government Accounting Office took a look at the concentration of Medicaid spending, particularly on non-dual eligible beneficiaries, as of 2009. (GAO Report) (As an aside, why are these programs so slow at getting good data available. It is 2014 and the last year of complete data is 2009?) For 2009 there were 64.2 million covered people with total spending of $314.3 billion. Of the 5% who represent the highest per capita spending, 4.3% were Medicaid only, the rest are dual eligibles. Those 4.3% of beneficiaries accounted for a third of all Medicaid spending. On a per capita basis, 18 times more was spent on these Medicaid-only recipients than on the rest of that group. The annual per person average was $35,983 in the high-spending group and $1,989 for the rest. But for comparison, the dual eligible population is even more expensive, with a per capita average of $89,440 in the top 5% and a$7,762 across the rest of the population. There is a very wide geographical range in spending, however, even on the high-cost Medicaid-only recipients, from a low of $20,896 in one state to a high of $83,365 in another. The South, Southwest and West tend to have lower per capita spending on these high cost patients while the Northeast and the Midwest tends to have higher spending.
Hospital and nursing home spending was the primary difference in category spending for these high-cost beneficiaries. As might be expected, nursing home residents are more likely to fall into the top 5%, as are those with AIDS, and people with disabilities. Mothers and infants comprise part of this population, but this is likely a one-time occurrence. Factors like age, race, and characteristics of the state where the patient lives were less associated with high or low spending status. Understanding on which beneficiaries Medicaid spending occurs, the categories of utilization in which the spending occurs, and factors associated with beneficiaries who have high spending is useful for guiding care management efforts, but it is a snapshot, often taken retrospectively. More valuable is understanding the persistence of high spending–which beneficiaries not only have high spending but have it over an extended period of time. Many high-spending patients are at the end of life, many have a year or so of high spending and then regress toward the average spending. The subset that has persistent high spending is where the greatest potential return is on the spending investment it takes to manage care and health more effectively. And one thing this report confirms is that states have a huge budget issue with their role in financing long-term care.