By Jane Norman, CQ HealthBeat Associate Editor
January 5, 2011 -- A Medicare Payment Advisory Commission (MedPAC) study found that regional variation in Medicare use is particularly high for services associated with post-acute care—including care at home, in skilled nursing facilities, and in long-term-care hospitals.
And, once again, Miami and McAllen, Texas, top the list of high Medicare users.
Big variations in overall Medicare spending across regions of the country intrigued lawmakers as they shaped the health care law (PL 111-148, PL 111-152) because there didn't seem to be much evidence of a relationship with quality of care or outcomes for patients.
The issue made its way into a speech by President Obama to the American Medical Association (AMA) and provoked widespread discussion when surgeon Atul Gawande wrote an article for New Yorker magazine examining why health care costs in McAllen are so much higher than in the rest of the country.
But the question is how to measure the differences in a meaningful way, and policymakers are continuing on that front. The new MedPAC study, which covers data from 2006 through 2008, drilled down even further than the agency has in the past into not just regional variation in overall spending but in the levels of use of specific services covered by Medicare.
To pin down the variations in use, the study adjusted Medicare spending to account for geographic differences in: wages; special payments to certain hospitals and physicians; beneficiaries' health status; and demographics.
That's because Medicare spends more in some geographic areas. Medicare accounts, for example, for the differences in nurse's wages between New York and North Dakota. Without removing those factors, a study could "wrongly conclude that an area had high use when in fact it had high prices, a very ill population or both," the study said.
MedPAC didn't recommend what policies Congress should pursue to try to even out these variations in use. "However, we do note that at the extremes, there is nearly a twofold difference between the metropolitan statistical area (MSA) with the greatest service use (the Miami, Fla., MSA) and the MSA with the least service use (the LaCrosse, Wis., MSA)," the report said.
"Extremely high levels of service use in certain areas may be driven by overuse and, possibly, fraud and abuse," it continued. "Additional policy measures may have to be taken in those areas beyond those used to address variation in general."
Miami has been cited in the past by federal officials as a particularly ripe area for Medicare fraud and abuse. In October, for example, a Department of Justice strike force alleged some $200 million in fraud in a case filed there.
The study also found that geographic areas that are high-use in one sector such as ambulatory care, inpatient care or post-acute care tend to be high-use overall. And areas with high service use among Medicare patients who died tend to have high service use for patients who survived as well. "In short, the pattern of high use often extends across different services and different groups of beneficiaries," the study said.
The lowest-use areas exhibit very low use of post-acute care services, about half the national average, the study said. Service use in LaCrosse is moderately low for inpatient and ambulatory services and very low for post-acute care, for example. Hawaii is similar.
On the other hand, the McAllen MSA has very high service use in the post-acute care sector, 3.2 times the national average, the study said. Home health use in McAllen is almost seven times the national average. Gawande's article pointed to an explosion in the use of home health services in McAllen, the study noted. Very high use of durable medical equipment and home health care is also apparent in four counties in south Florida.
"Variation at the level found in Miami-Dade cannot be explained by simple differences in practice patterns, price or health status," the study said. "Rather, it is likely evidence of fraud, as is suggested by the many Office of Inspector General reports on the topic."
The study also looked at regional variations in Medicare enrollees' use of prescription drugs covered under Part D, based on 2007 and 2008 data. It found that regional differences in drug use is less than the variation in drug spending, though some differences remain.
MedPAC released a somewhat similar study in December 2009. The new study was different because it looked at use of specific services and also prescription drug use. Data used in the study was also updated and refined, the authors said.