Skip to content
Opinion Column

Attacking Dartmouth Atlas Is Off-Target

By now, every health care reporter in the country should have at least heard of the Dartmouth Atlas, and those who haven’t should probably go to Wikipedia or the Atlas website itself before committing another word to the topic of health care reform. The Atlas’ notoriety among reporters and the general public is only recent, despite the fact that the researchers involved have been compiling their data for nearly 20 years. But their central message has been absorbed by health care policy wonks and many of the members of Congress who are involved in crafting health reform legislation: Our health care system delivers a huge amount of unnecessary care, maybe as much as 20 to 30 percent of every health care dollar.

Many researchers and providers agree with this assessment, and I’ve had individual physicians tell me that they know for a fact that patients are being put at risk in their hospitals by all the unnecessary care they receive. So why are a handful of administrators and doctors at some of the nation’s most prestigious academic medical centers trying to discredit Dartmouth’s findings? The answer, in a word, is money.

Using Medicare claims data, the Dartmouth Atlas shows that different hospitals deliver wildly differing amounts of care to patients with similar conditions and similar levels of illness. In two seminal papers published in 2003, Dartmouth’s Elliott Fisher showed that patients who had suffered a hip fracture, heart attack, or had colon cancer that needed surgery, got very different care, depending upon where they lived. For example, a heart attack patient who lives in, say, Los Angeles, will receive about $7,000-worth of extra care over the course of a year than a similar patient in Salem, Oregon. That $7,000 of extra Medicare spending in LA wasn’t due to higher prices, it went towards more services – more hospitalizations, more drugs, doctor visits, tests, and procedures.

Nobody disputes these variations in spending or utilization. The part that makes doctors and administrators in places like LA see red is Dartmouth’s conclusion that all that extra care doesn’t necessarily result in better outcomes. It’s mostly waste.

They worry that the Centers for Medicare and Medicaid Services (CMS) will use Dartmouth data as a benchmark for measuring health care efficiency and try to wring savings out of the most inefficient regions — Los Angeles, for instance, down-state New York, all of New Jersey, Miami and McAllen, Texas (which was profiled as the most expensive town in America for health care in a New Yorker article). The Baucus health care reform bill would create and fund an “innovation center” within CMS, with broad powers to experiment with different payment options aimed at “bending the curve,” reining in the rate of growth of health care spending. According to Fisher, 20 to 30 percent of inpatient hospital days and 30 to 40 percent of specialist visits could be eliminated in the most inefficient hospitals without hurting patient outcomes.

To do that, CMS could employ a range of incentives such as a shared savings program, which would reward administrators for emulating more efficient hospitals – places like Intermountain Healthcare, in Utah, and the Geisinger Clinic of Pennsylvania, which were recently cited by President Obama as medical exemplars. Another way to control costs and encourage better care would be to spare providers in more efficient regions from across the board cuts in Medicare payments.  

Rather than stepping up to the plate and vowing to get to the bottom of their spending problem, some academic medical centers are instead attacking the Dartmouth findings. They explain away the data by saying their patients are poorer and sicker, and therefore need more care. In Los Angeles, for example, where 15 percent of the population is below the poverty line, the average Medicare recipient costs the federal government $10,810, compared with $6,705 per recipient in Minneapolis, where 10 percent of citizens are below the poverty line. 

This claim, that poverty and illness lie at the heart of variation in Medicare spending,  makes intuitive sense, but it doesn’t hold up to careful scrutiny. Poverty does influence health, but it can’t account for the extraordinary differences in the volume of care delivered by academic medical centers – even when they are located within a few miles of one another. In Manhattan, for example, chronically ill Medicare recipients, whether they are poor or not, are given about 40 percent more services at New York University Medical Center than nearby Columbia Presbyterian. In Chicago, African-American Medicare patients suffering from chronic illness, who are generally poorer than white beneficiaries, spent 46 percent more days in the hospital at Rush-Presbyterian than African-American beneficiaries using the University of Chicago. There’s no evidence that patients at Rush-Presbyterian and NYU are enjoying better outcomes.

Academic medical centers are supposed to lead medicine toward better, more effective care. They are typically considered centers of excellence, beacons of sound medical science. Trying to discredit the Dartmouth data is a distraction from the real work that’s needed to understand and remedy the extraordinary amount of money spent on care that does not appear to make a difference in health.

Shannon Brownlee is a Senior Research Fellow at the New America Foundation and author of Overtreated: Why Too Much Medicine is Making Us Sicker and Poorer.

Related Topics

Cost and Quality Health Industry