Wednesday, March 11, 2009

Variations in Medicare Care 

Maybe it's the patients

By Greg Scandlen

Filed As:  Health Care

Here is one of the most thought provoking articles I’ve come across in a long time. It’s by Elliott Fisher, MD, and colleagues at Dartmouth and published in the New England Journal of Medicine. As you know, Dartmouth is the home of the Dartmouth Health Atlas developed by Dr. Jack Wennberg, who pioneered the research on “small area variations” in medical practice. The Health Atlas has been zeroing in on variations in the intensity of use between regions of the country under Medicare.

These discrepancies are well-known by now. People on Medicare in New York City consume twice as much care as those in Minneapolis, and this is independent of any differences in local prices or differences in demographics. It also seems to be independent of any differences in outcomes.  

It is fascinating work. The usual knee-jerk response from policy people is to blame Medicare’s fee-for-service payments for not rewarding efficient care and paying for whatever services are provided. In effect, paying more for more services.

But that has never been a satisfactory explanation. Sure, incentives drive behavior and people will do more of what they are rewarded for. But the payment system is the same in Miami as it is in Seattle. If that was the cause, the results should be the same everywhere.

This article agrees. It tracks the rate of increase in per capita spending in five regions since 1992 and notes vast differences between Miami and Salem, Oregon. But, it says, “it is difficult to blame regional differences entirely on the current payment system, since all our evidence on regional growth comes from populations in the fee-for-service system.”

It also suggests that managed care is not likely to be the solution (as many have advocated) because about half of Medicare enrollees in Miami are in Medicare Advantage plans.

The article says that physicians in some areas are simply more aggressive than in other areas and more likely to recommend “discretionary services, such as referral to a subspecialist for typical gastroesophageal reflux or stable angina or, in another vignette, hospital admission for an 85-year-old patient with an exacerbation of end-stage congestive heart failure.”

It goes on to suggest more use of “integrated delivery systems,” though it is not clear to me how that is different than managed care.

I would like to propose to the authors a different cause, and that is variations in patient demand caused by cultural differences. I was working at Blue Cross Blue Shield of Maine when Wennberg was starting out. He used our claims data for much of his early work, including an early study that found women in Lewiston, Maine were something like three or four times more likely to get a hysterectomy in their lifetimes than were women in Wiscasset, just thirty miles away.

Wennberg concluded that physicians in Lewiston were scalpel-happy, and more inclined to jump into the OR than their brethren in Wiscasset. But that wasn’t it. Fact is, Lewiston was heavily French Canadian and Roman Catholic, while Wiscasset was almost entirely Protestant. Women in Lewiston were using hysterectomies as a form of birth control that was acceptable to the Church after having six or more children. It was a small price to pay to avoid Eternal Damnation.

Statistical analysis is fine, but understanding is something else. Understanding requires personal contact and a love of people. It requires patience and listening. In the case of differences between Miami and Minneapolis, it would be worth asking if there is something about Miami’s Jewish culture that results in a different use of health care than the largely Lutheran culture in Minneapolis.

It may be that it isn’t doctors who are causing the differences, but patients.

• posted by Greg Scandlen
• 3:36 pm
• 0 comment