Dartmouth Revises Medicare Spending Variation Report

Dartmouth Atlas researchers have taken to heart last year's heated criticism that they hadn't adjusted for variations in regional cost of living differences when they published statistical maps showing wide regional variations in Medicare per capita spending. This week, statisticians with the group released a new set of tables that do just that.

Most of the 306 geographic hospital referral regions moved a few points up or down, but most noteworthy were downward shifts in spending for New York City and San Francisco, Dartmouth statistician Jonathan Skinner said in a telephone interview.

"We published these measures for years where we didn't adjust for the fact that Medicare pays more in San Francisco, Chicago and New York than in other places," Skinner said. "And over time, people said, 'Oh, our city costs (Medicare) a lot more because everything costs more in our city.'

"So we decided to redo the whole thing where we did it both ways – price adjusted and unadjusted, and with a larger Medicare sample size, 20% (of Medicare fee for service beneficiaries) instead of 5%," he said. "We felt it was important for researchers and policymakers to have this data available in case they want to know how much cheaper is San Francisco really."

The Dartmouth Atlas had been a relatively obscure set of maps showing the amounts providers in various regions of the country submitted in healthcare claims for Medicare beneficiaries. But during the health reform debate, the maps came alive, showing how providers in some regions of the country were providing patients with a lot more expensive care, but with apparent negligible benefit to their survival.

According to an article in the New York Times June 2, 2010, "the maps made reform seem relatively easy to many in Congress, some of whom demanded the administration simply trim the money Medicare pays to hospitals and doctors in the (more expensive) zones. The administration promised to seriously consider doing just that."

The new tables, which look at just Medicare spending for 2003 to 2008, use relative value units to measure cost of labor to reflect the fact that it's more expensive overall to provide medical care in New York City than it is in, say, Fargo, North Dakota.

The Dartmouth team's new tables also reflect the removal of differences in reimbursement rates that allow Medicare to pay hospitals with medical and surgical residency training programs more than hospitals that don't have those academic functions. They also leave out higher payments to hospitals that serve a high percentage of low-income patients, called DSH or disproportionate share adjustments.

The major spending anomalies or outliers, Miami and McAllen, TX, which became synonymous with enormous Medicare spending during the healthcare reform debate of 2009, continue to maintain their first and second places in healthcare spending.

But major shifts were seen in New York City, Skinner said. Bronx and Manhattan, which were fourth and fifth highest out of 306 areas in the country in Medicare spending in the original report, dropped so much that they are no longer considered outliers.

"It makes them look better because they're spending less," he said.

That's not the full story, however. Looking at how New Yorkers use healthcare, the picture changes. "If you look at hospital days near the end of life, and doctors visits in New York, they're still really, really high," Skinner said. "And here's another paradox, when people are admitted to a hospital with a given DRG (diagnostic related group) in New York, they stay in the hospital longer and they're seen by more doctors."

According to an executive summary on the new calculations, in general urban regions tended to exhibit lower spending rates after price adjustment, while rural regions exhibited higher spending.

In addition to downward spending adjustments for New York and San Francisco, expenditures were somewhat higher for certain rural regions such as Bismarck, ND; Jackson, MS and Monroe, LA.

Expenditures still varied greatly across the country per capita, from less than $7,000 per beneficiary in 18 hospital referral regions – including Rapid City, SD ($6,264); Honolulu ($6,653); San Mateo County, CA ($6,723) and Portland, OR ($6,971) to more than $12,000 per beneficiary in Miami ($15,571); McAllen, TX ($14,529); Harlingen, TX ($12,231) and Monroe, LA, ($12,027), the summary said.

Skinner said the Dartmouth team is working on correcting another aspect of its earlier published data, namely that it did not risk adjust for disease status in various parts of the country, because some areas are known to have greater percentage of people affected by obesity, diabetes, hypertension and cardiovascular disease.

There is the factor of "reverse causation," which is the phenomenon by which Medicare beneficiaries who move from low intensity spending and healthcare utilization to areas of higher spending end up with additional diagnoses "so much so that their HCC (hierarchical condition categories) codes would predict 19% higher spending per patient than for their counterparts who moved to low intensity hospital referral regions.

"Thus there are biases in using HCC-based adjustments that tend to artificially diminish true geographic variation in both spending and risk-adjusted mortality" that can be seen most graphically in McAllen, TX.

By the looks of it, it is one of the most expensive healthcare areas in the country and should be deemed to have a population that is the sickest.

"Yet the 2007 age, sex and race-adjusted mortality rate for McAllen in the Medicare population (4.28 per 100 beneficiaries) was well below the national average (4.82) and even below the mortality rate in Provo, Utah (4.67)."

The summary concludes that clearly, some regions of the country are less healthy than others, and the Dartmouth Atlas Project's findings "are likely to overstate spending in unhealthy regions and understate spending in healthy regions."

That's why the team is developing new measures "based on cohorts of Medicare beneficiaries who have had a hip fracture or heart attack in order to compare, for example, how hip fracture patients are treated in Minneapolis compared to those in New York."