In Medicare, states are far from equal

In 2015, Hawaii spent $6,608 per traditional Medicare beneficiary, but Mississippi spent $10,700. Nationwide, Medicare per capita costs amounted to $9,582, CMS data released Tuesday show. Why the difference?

Medicare spending, utilization and quality vary widely across states, a phenomenon that is longstanding but remains poorly understood. These fluctuations by geography do not seem to reflect beneficiaries' health, the CMS noted in its methodology document for the data, which are released annually in what is formally known as the Geographic Public Use File.

This latest update includes data for the period 2007 to 2015, in seven categories, including geography variables, utilization, quality and study population.

The data revealed significant variation in many areas of Medicare spending, utilization and quality, such as within Medicare Part B. The actual (non-adjusted, non-standardized) costs per user of Part B drugs in 2015 ranged from $381.20 in Vermont to $978.75 in Florida. Maine had the lowest proportion of beneficiaries using Part B drugs — 33.52% — whereas Tennessee had the highest proportion at 60.78%.

Quality varies substantially too, at least as gauged by hospital readmissions. Utah's 30-day inpatient readmission rate of 12.74% ranked it the lowest; Washington, D.C.'s readmission rate of 21.39% made it the highest, followed by Florida at 19.43%.

The CMS released the data to make analyzing differences in healthcare spending and utilization easier. The data are grouped into three types of geographic units: hospital referral regions, which delineate regional U.S. markets, states and counties. The data include Washington, D.C., Puerto Rico and the Virgin Islands.

In order to facilitate the most meaningful comparisons possible, the data exclude some Medicare beneficiaries, including those in Medicare Advantage and those enrolled only in Part A or Part B. The study population included 34.1 million out of 58.3 million total Medicare beneficiaries.

To measure utilization, it looked at metrics such as the number of stays and days of care in healthcare facilities, home health episodes and visits, and numbers of tests. It also calculated all-cause hospital readmissions, emergency room use, and admission rates for medical issues including diabetes, hypertension and congestive heart failure, to measure quality. These measures can be used to analyze the relationship between healthcare utilization and quality of care in Medicare, the CMS said.

Experts and policymakers disagree over what accounts for this variation and how to implement payment reform accordingly.

"A sizable literature documents that dollars spent per Medicare beneficiary vary greatly" among regional markets, noted Joseph Newhouse and Dr. Alan Garber in an editorial in the New England Journal of Medicine in 2013. "However, the consensus ends there. There is controversy about the variation's causes, its effects on quality and outcomes, and what, if anything should be done about it." They noted that a committee convened by the Institute of Medicine attributed most of the variation to differences in the use of post-acute care and inpatient services.

Experts and policymakers have been grappling with this issue for decades, to little avail. Data analysis published in a 1999 article in Health Care Financing Review showed variations in per capita spending measures that were notable for their parallels to today, two decades later. Hawaii spent $2,289 per Medicare enrollee in 1995, the least of all states. Louisiana -- as opposed to Mississippi in 2015 -- was the top spender with nearly twice the per capita expenditures of Hawaii, at $5,481 per fee-for-service beneficiary, in 1995.