Geographical variations in health care costs: Roger Lane’s Rules of History in medical practice
Forty years ago I had a wonderful history professor at Haverford College, Roger Lane, who would begin the courses he taught by giving us “Lane’s Rules of History.” The first one was: “things are complicated.” Others included “things change,” which had a corollary: “and nobody knows how they will turn out.” What he meant, of course, is that generally the closer you study something involving human behavior, the more complicated it becomes. And predictions about the future are always problematical. Professor Lane has always had a healthy and amusing respect for the vagaries of human behavior, as you can see in the above picture of his often-smiling face.
Now we are finding that Professor Lane’s rules apply to medicine, too. This should not be surprising, since medicine is practiced by people on other people. Many years ago the Dartmouth Atlas of Health Care, a still ongoing project, noted there were astonishing differences in the cost of heath care across geographic regions of the country. Some of these differences were between places quite nearby to each other, and correcting for such things as differing costs of living and doing business did not affect the results. Most importantly, high cost regions did not deliver better health care — it just cost more overall.
What appears to be the most important difference between high and low cost areas is medical culture, mostly physician culture, in “the way we do things here.” Patterns of practice, of how doctors evaluate and treat the same condition, vary quite a bit. In places where doctors do more, costs are higher, but with no discernible improvements in patient outcomes. As a case study, Atul Gawande wrote a fascinating essay about the inexplicably high costs in McAllen, Texas — the most expensive medical market in the country. But people are no healthier there.
These disparities have not gone unnoticed by those who pay the bills, primarily Medicare. A recent editorial in the New England Journal of Medicine describes the results of a government study that tried to figure out what was going on. What they found is Lane’s First Rule of History in action — things are indeed complicated.
The Dartmouth Atlas divides the country into 306 hospital referral regions, among which there were these major differences in costs. That’s a lot of regions, so you would think it slices up the country into fairly small bits. But it is even more complicated, because the variations within some hospital referral regions were just as great as between them. Even more complicated was the finding that extensive (and expensive) use of treatments for one condition did not translate to others; regions were high cost in some things but low cost in others.
Some have argued the variation is primarily patient-related. That is, regions vary in the health of their respective citizens, so it’s not just doctors and hospitals in one place being more inherently aggressive than those in others. Some places, the theory goes, have sicker inhabitants than others. I’m sure there are examples of that. But refuting this theory is the finding that in many cases people who move from a high cost region to a low cost one, and vice-versa, immediately experience whatever the costs are in their new region. It seems unlikely they got sicker or healthier just from moving.
What conclusions did the blue-ribbon committee come to? What should be done? Anything? They did offer this:
In sum, the committee found that most of the variation among geographic areas is attributable to variation in the use of post–acute care and inpatient [hospital] services. Moreover, within any area, provider behavior varies substantially, so increasing reimbursement for all providers in an area would unfairly reward poorly performing providers, and reducing reimbursement for all providers in an area would unfairly penalize high-performing providers.
My conclusion is that sociology is at work here. I don’t think physicians and hospitals in one place are especially more venal than they are in another place. I think physician behavior varies according to the local physician culture. I have observed physician routines in multiple places around the country — Minnesota, Colorado, South Carolina, Kansas, Missouri, Arizona — and I have seen rather large differences in the nitty-gritty at least in how my branch of medicine is practiced.
We certainly can improve things with “best practices” research and training physicians to use them accordingly. Yet that approach only applies to some of what we do. Medicine resists standardization because it remains in some irreducible ways a black art — a mishmash of science, near-science, intuition, guesswork, and blind luck.
Things are complicated.