The answer: none of the above.

It’s your zip code.

Everything’s bigger in Texas

In an influential article in The New Yorker magazine called “The Cost Conumdrum,” author and physician Awtul Gawande described McAllen Texas, the square dance capital of the world, as one of the most expensive Medicare health care markets in the country. In 2006, Medicare spent $15,000 per enrollee in McAllen. That’s almost twice the national average and nearly three times the amount spent in Grand Junction Colorado. Was it because the residents of McAllen are sicker than other seniors, resulting in higher Medicare costs? Again, the answer is “no.” El Paso, Texas has similar demographics to McAllen, yet their Medicare expenditures were half as much as McAllen’s.

Measuring local use

Gawande based his article on the Dartmouth Atlas Project, a program that documents variations in how medical resources are distributed and used in the United States. The Atlas depicts marked differences in health care spending within small regional areas roughly approximating zip codes. It shows that the use of health care resources in the United States is highly localized. Overall, the highest areas spend 2.7 fold more on Medicare than the lowest areas. And utilization, not price, demographics or health status, drives the differences.

The Dartmouth Atlas flows from the pioneering work of John Wennberg, the founding editor of the Atlas, to understand “small area variation” in the utilization of health care services. In his book “Tracking Medicine: A Researchers Quest to Understand Health Care,” Wennberg chronicles a lifetime of discoveries. His work grows in importance as the national deficit increases and our national debt skyrockets (driven, in no small part, by health care spending). And it’s variation in costs that presents such a complex challenge. If the entire country spent the same as the lowest cost regions, we would not need a national debate about how to solve the deficit and debt problems.

Three kinds of care

Wennberg describes three categories of care: effective, preference-sensitive and supply-sensitive.

Effective care, or necessary care, represents services that, on the basis of sound medical evidence, are known to work better than any alternative, and for which the benefits of treatment far exceed the side effects. Examples include immunizations for young children, surgical repair of hip fractures, certain medications for heart attacks and colon resection for colon cancer.

While some critics say Medicare services are overused, the problem with effective care is the opposite. Effective care is underused, resulting in a failure to provide care for patients who did not get their required treatment. Spending for effective care accounts for only 15% of total Medicare spending with little variation from region to region.

Preference-sensitive care refers to interventions for which there is more than one option and where the outcomes will differ according to the option used. This category accounts for about 25% of Medicare spending and includes elective surgery. One example is treatment of an enlarged prostate in men. Options for treatment include living with the symptoms, taking drugs, undergoing surgery or other invasive procedures. As you can see, preference-sensitive care involves making trade-offs in quality of life. 

Examples of regional variation in preference-sensitive care include a 6 fold difference in local rates for lumbar discectomy (“back surgery”) for chronic low back pain and 5 fold difference for coronary artery stenting for non-acute coronary artery disease. Wennberg cites evidence that opinions of physicians, rather than preferences of patients, drive local variation in this category. When physicians engage patients in formal “shared decision-making,” patients choose less invasive procedures more frequently than when physicians’ opinions drive choices.

Supply sensitive care accounts for 60% of Medicare spending and results in big differences in health care costs from area to area. Supply sensitive care is not about specific treatment per se; rather it is about the frequency with which everyday medical care is used in treating patients with acute and chronic disease. These include physician visits, referrals, home health care, imaging exams, hospitalizations and use of intensive care units.

Regional differences in rates of these services appear to be guided by, as economists say, an “invisible hand” of supply. In other words, the greater the supply of hospital beds and physicians (particularly specialists) the more they are utilized. Although Americans like to believe that “more care is better care,” Wennberg provides overwhelming evidence that higher utilization does not correlate with superior clinical outcomes or patient experiences. Indeed, the evidence is highly suggestive that areas with high utilization tend to have inferior outcomes and less satisfied patients — probably because the added services results in more uncoordinated care.

Park Nicollet’s care model

What does this mean for the Park Nicollet care model? Quality care entails eliminating unnecessary variation in all three categories. We also must approach each category differently. First, we must encourage patients to receive all their effective care. Second, we must use formal shared decision making tools for preference-sensitive care. Third, we must develop protocols to reduce supply-sensitive care.

The Park Nicollet plan for 2012 and beyond will include reducing unnecessary variation in each of these categories. The evidence is compelling. It is the right thing to do and it will lead to higher quality care that is more affordable.

Please feel free to use the comment box below to share your thoughts. We encourage a free exchange of ideas, but we reserve the right to remove comments that make personal criticisms or attacks on individuals or specific businesses.