Getting the Facts Right: Regional Variations in Health Care and the Dartmouth Research

A story in today’s New York Times begins: “For years, health policy experts have said health care spending is much higher in New York City and Boston because doctors and hospitals there provide more services, practicing medicine in a more intensive way. But new government data show that Medicare costs per patient in those cities are slightly below the national average when the numbers are adjusted for the cost of living and other factors.”

The Times goes on to explain that President Obama “says that the nation could save huge sums if all doctors and hospitals were as efficient as those in lower-cost states like Iowa, Minnesota, Washington and Wisconsin. Lawmakers from those states have reached an agreement with House Democratic leaders that would increase federal Medicare payments to health providers in their states. Higher-cost states, which could see their Medicare payments reduced, are fighting back.”

The paper acknowledges that Dr. Denis A. Cortese, president of the Mayo Clinic, based in Rochester, Minn., agrees with the president, saying that Medicare wastes billions of dollars a year because it “pays the most to health care providers and geographic areas that provide the lowest-quality care at the highest costs.” In other words, Medicare is rewarding inefficient hospitals that keep patients longer because it takes them longer to diagnose the patient, because they run more tests, because more specialists are called in to consult, and –because they don’t always consult with each other–  patients suffer complications, or fall victim to medical error.

But, the Times notes, “Dr. Steven M. Safyer, president of Montefiore Medical Center in the Bronx, disagrees, saying : ‘Our Medicare expenditures reflect the low socioeconomic status of the population and the very high cost of doing business here. Many of our patients do not receive regular care before becoming eligible for Medicare and have no one to care for them after they leave the hospital.’”

Moreover—and here’s the eye-opening paragraph in the Times story: “The Medicare Payment Advisory Commission (MedPac), an independent federal panel that advises Congress, has found that geographic variation in Medicare spending is substantial. But it told Congress recently that much of the variation could be explained by local differences in the cost of providing care and in the health status of beneficiaries, as well as by extra payments, authorized by Congress, for hospitals that train doctors or treat large numbers of low-income patients.

After adjusting for those factors, the commission said, Medicare payments per beneficiary were 92 percent of the national average in New York City and 95 percent of the national average in Boston.

Wait a minute—is MedPac saying that more than two decades of Dartmouth research showing that patients receive far more intensive aggressive and expensive care in cities like New York and Boston, is wrong?

No. Looking at the dollars that Medicare spends is one thing. Looking at utilization of health care resources—what percent of patients see 10 or more specialists during their last six months of life, how many days they spend in the hospital, how many tests they undergo—is something else. “When you look at utilization, you don’t have to price-adjust,” Dr. Elliott Fisher a professor of medicine and at Dartmouth Medical School and Dartmouth’s Center for the Evaluative Clinical Sciences explained in a phone interview this morning. “If you look at hospitals in affluent parts of Manhattan, you’ll find that utilization is twice as high as in Rochester New York.”

The table below compares spending and utilization in Rochester and in Manhattan both during the last six months of life and during the last two years for Medicare patients suffering from very similar chronic diseases. How sick were they? They all died within two years. (The data looks back on their care during those years.) The table also zeroes in on two hospitals—NYU Medical Center in Manhattan and Strong Memorial Hospital in Rochester.

Table1*Part B spending for imaging per decedent  (click table to enlarge)

Source: The Dartmouth Atlas of Health Care 

While differences in the dollar amount that Medicare spends is affected by differences in local prices, that doesn’t explain the enormous disparities in the amount of time patients spent in the hospital, the number of physician visits, or the percent who were treated by 10 or more specialists.

Clearly, patients in Manhattan were receiving far more aggressive, intensive care, and inevitably, this winds up being more expensive care.  Meanwhile, reams of Dartmouth data show that outcomes are no better in regions where doctors and hospitals pull out all the stops. Sometimes they are worse. After all, every test, every treatment exposes a patient to some risk. If the test is redundant, if the extra days of hospitalization weren’t necessary, than the patient is, by definition, exposed to risk without benefit. “Hospitals are dangerous places,” says Fisher, “especially if you don’t need to be there.”

Other Dartmouth Studies Do Adjust for Differences in Local Prices, Race, and Socioeconomic Status

The table above does not factor in differences in local prices, race, age, sex, socioeconomic status and severity of illness. But in other studies, when comparing spending and outcomes at academic medical centers in different regions of the country, Dartmouth researchers do adjust for all of these factors.

And, after making these adjustments they find enormous differences regional differences, even at academic medical centers, where one might expect physicians would follow similar evidence- based guidelines. One would  be wrong.

Fisher explains: “In a study published in Health Affairs, we compared outcomes and quality of care for heart attack, colon cancer and hip fracture patients, adjusting for underlying differences in both health and socioeconomic status. We then compared costs of care, adjusting for price differences — and found that at the academic medical centers where patients received more intense care, per-patient costs were still about 60% higher over the long term — with no better outcomes or quality.”

Specifically, “Patients in the higher-intensity hospitals spent more time in the hospital and intensive care unit (ICU); had more frequent physician visits (especially in the inpatient setting); have more specialists involved in their care; and received more imaging services, diagnostic testing, and minor (but not major) procedures.”

When researches compared long term survival, it turned out that mortality rates were slightly higher for heart attack patients and cancer patients who received the  priciest, most intensive care.

In today’s article, the Times quoted Dr. Darrell G. Kirch, president of the Association of American Medical Colleges, saying that the proposed redistribution of Medicare money could have “catastrophic unintended consequences for teaching hospitals and their patients.”

 What he didn’t say is that, in fact not all teaching hospitals provide more costly care. The level of intensity tends to be tied to the region where the medical center is located. Secondly, proposals for reducing Medicare payments in some regions most often suggest cutting payments to hospitals (not individual doctors) that fail to approach benchmark standards in terms of cost and quality.

Why Do Patients In Some Regions Receive More Care?

Why is care so much more aggressive in some academic medical centers than in others?  In the past, some Manhattan hospitals have explained that their patients are simply more sophisticated, and more demanding than patients elsewhere. But researchers point out that in general, patients who want more aggressive management of their illnesses are likely to choose academic medical centers. “Therefore,” Fisher and colleagues note in the Health Affairs article, “we would expect patients’ preferences to differ relatively little on this dimension across these hospitals. Data from other studies also argue against patients’ preferences as a major determinant of the differences in practice we observed across hospitals.”

Another Dartmouth study that compares similar patients at Stanford University Medical Center, to Medicare enrollees who were cared for at New York University(NYU) Medical Center  finds that the NYU patients spent more than twice as many days in the hospital (twenty-seven versus ten) and saw more than three times as many physician visits (seventy-six versus twenty-three) during their last six months of life. Yet it’s hard to imagine the patients at NYU are significantly more knowledgeable, sophisticated  or demanding than patients at Stanford.

The single factor that does seem to explain differences in cost and intensity of care is the supply of specialists and hospital beds in a given region. In areas where there are more doctors and more beds per capita, patients are more likely to find themselves in one of those beds—and they are more likely to wind up with ten or more specialists consulting on their case. The researchers explain: “Given the current culture and financial incentives of U.S. medicine, which favors intervention among the seriously ill (even in situations where the scientific evidence is weak), physicians in a given care system are likely to draw upon all available resources to care for their patients.

“Moreover, it is often easier to manage patients’ care in the inpatient setting; the outpatient setting imposes a greater burden on the physician (for example, to arrange telephone, visiting nurse, or office visit follow- up to ensure that the patient is responding to therapy). The inpatient setting also lowers the threshold for further intervention: It is easier to obtain tests, perform minor discretionary procedures, or request a consultation. Studies of ICU bed availability reveal the same phenomenon: When more beds are available, the average severity of the patients cared for in the ICU is lower and the average length-of-stay is longer.”

In the end, is the Times story spreading misinformation about the Dartmouth research—research that is essential to President Obama’s argument that we can rein in health care spending, and improve outcomes, by rewarding providers who begin to practice higher quality, more efficient medicine? \

Not eactly.. Butt because the Times doesn't ' tell  lthe full story , it muddies the waters. By failing to note that you don’t have to adjust for prices when you’re looking at utilization, and by overlooking the studies where Dartmouth does adjust for local prices, income and demographics, the Times piece creates confusion on the eve of President Obama’s speech. This is truly unfortunate, because tomorrow, the president will probably want to make the case that we can achieve significant savings in Medicare spending—by reducing the amount of needless spending that doesn’t improve outcomes and, in fact, often makes them worse.

“Next week, we have an article coming out in the New England Journal of Medicine addressing these issues, including differences in local prices,  says” Fisher . Factors like socioeconomic status do matter, he adds: “we don’t want to penalize a hospital in the Bronx that cares for much more vulnerable patients. But when you look at hospitals in the most affluent section of Manhattan, it’s hard to explain more aggressive, more intensive, more expensive care” compared to,  say, Johns Hopkins in Baltimore Maryland. Especially because patients don’t benefit from pricier care.

But  how many of the people who listen to the president’s speech tomorrow will read the article in the New England Journal of Medicine next week?  Verf ew. One can only hope that both bloggers and reporters in the mainstream media will  read piece in NEJM,[[ and make a concerted effort to explain the waste in our health care system to their readers.


26 thoughts on “Getting the Facts Right: Regional Variations in Health Care and the Dartmouth Research

  1. Montefiore’s response reminds me of the post on The Healthcare Blog a few weeks ago saying that the reason healthcare in McAllen is so much more expensive is because they have so much more comorbidity in chronic illnesses than any other area in Texas.
    As if diagnosis were some kind of iron-clad science, such that doctors at Mayo would always come to the same conclusion as a doctor at Montefiore.
    As if varying aggression in diagnosis weren’t at heart the same thing as varying agression in treatment.
    They just don’t get it. I’ve heard they don’t teach data analysis in med schools very much, so maybe they don’t understand how unlikely it is that two patient populations with similar socioeconomic and demographic backgrounds (McAllen and El Paso) could have such completely different diagnostic data that just happens to correspond to more aggressive practice patterns even when diagnoses are identical.

  2. Ahh I see, the big medical centers are just as powerful at getting what they want. These are BIG advertisers at the Times. Best not to piss them off.

  3. That there are geographic variations in health care quality is not breaking news. These quality disparaties occur even within cities, and probably exist within single hospitals. It’s not the physicians and hospitals don’t know the right thing to do for patients; it’s that there are competing and distracting incentives and influences that pull us off the quality curve. Some of these include financial conflicts, litigation fear, institutional pressure and a medical culture of diagnostic and therapeutic excess.

  4. I’m sure Pear didn’t write the article in order to please academic medical centers. Why not look at all sides of issue? Why not put the Dartmouth data under scrutiny? Why not validate the Dartmouth data? Isn’t this what evidence-based medicine is all about?

  5. To assess regional practice pattern variations in hospital based care, I think we need to break the care down into four separate but somewhat overlapping buckets as follows:
    1. A definitive diagnosis was established. The patient needs a hip or knee replacement, heart bypass surgery or a stent. My understanding is that famous medical centers like the Mayo Clinic or the Cleveland Clinic charge commercial insurers and wealthy self-payers very high rates compared to other hospitals for the procedures that they perform. There could be material regional differences in the need for heart operations based on varying rates of obesity, smoking or socioeconomic status. Conversely, there could be more need for joint replacement surgery when more of the population participates in outdoor, athletic activities like skiing, golf, tennis, etc. As an aside, procedures like these lend themselves especially well to bundled pricing.
    2. We don’t know what’s wrong and we need to figure it out. Here, there could be significant regional differences in practice patterns due to the supply of specialists and hospital beds, differences in the legal environment or the extent to which the medical culture may be more money driven than elsewhere. I assume factors including age, gender, race, and socioeconomic status were adjusted for in a statistically appropriate manner.
    3. There could be differences in how effective the management of chronic disease is between one region and another. The cost of caring for patients with heart disease, diabetes, congestive heart failure, asthma and depression should probably be evaluated separately, again adjusted for differences in age, gender, race and socioeconomic status.
    4. Finally, there is end of life care. Doctors often cannot accurately predict how long a given person will live. Hospice patients may die within a week or live for more than a year. However, patients with severe dementia or Alzheimer’s or late stage cancer could be treated very differently from one place to another. There could be differences in the percentage of these patients that have executed living wills. There could be differences in the legal environment that compel providers to provide care that even they may not think is useful or appropriate or there may be money driven issues related to medical culture and the supply of specialists and hospital beds.
    Personally, I would like to see CMS provide data that would show Medicare’s spending for the over 65 population that occurred in the last 6 months and the last two years of life as a percentage of total Medicare spending for those individuals since they became eligible for the program. Does this percentage vary by region and has it changed over time?

  6. jms–
    The Dartmouth Data has been scrutinized for more than 2 decades.
    At this point, no one among medical reserachers and informed physicians disputes it.
    I myself have spent months studying it–as have others. See Shannon Brownlee’s excellent book Overtreated.
    Perhaps Robert Pear just heard about it — I don’t know.
    But Pear is not a health care reporter; he is a political reporter.
    The New York Times does have a number of fine health care reporters– people like Reed Abelson who have written about the Dartmouth Reserach in depth and who understand it. She has scrutinized it and reports that it is true–and very important.
    Why the Times doens’t have her write about the Dartmouth research– I don’t know.
    Pear has been at the Times for years and year, and frankly is not considered nearly as good as many of their younger reporters.
    Too often, he gets his facts wrong.
    As a political reporter, he failed to question the facts about weapons of mass destruction and Iraq.
    As a political reporter, he failed to stand up to the Bush administration.
    When it comes ot healhtcare, he just doesn’t knwow the subject in depth.
    But he (or someone at the Times) does seem to have a political agenda regarding health care reform. Someone is not enthusiastic about taking the waste out of the system. People who profit form the waste?
    Affluent middle-aged Times exectuives and editors who feel that they want to have every test, and every pill and every surgery that comest down the pike–even if their if no evidence that it works–because they hope to live forever? I don’t know.
    Why didn’t someone at the Times call Dartmouth to check the story? (They didn’t even try).
    Why didn’t Pear make it clear that this was not a report from MedPac but something put together by a MedPac staffer?
    Why didn’t the MedPac staffer call Dartmouth to see what they had to say?
    (He didn’t)
    I picked up the phone yesterday am. and had no
    problem reaching Fisher.
    And I wasn’t using a private number–just went through the switchboard at Dartmouth.
    Anyone who knows how to dial information could have reached Fisher.
    I also reached Jon Skinner, a Dartmouth economist involved in the reserach.
    To say that utilization in New York is no higher than in the rest of the country is simply untrue.
    Medicare has the numbers.
    Finally- your reference to Pear not trying to please “academic medical centers” –I assume you are referring to Dartmouth.
    If you actually read my post (rather than commenting without reading) you would know that the Dartmouth reserach is very critical of many academic medical centers–like NYU–

  7. It makes you wish someone would compare the case controlled costs of McAllen with Brigham and Womens. If they could find comparably sick patients, it might make McAllen look like a bargain.

  8. You don’t think the local hospitals aren’t big spenders on NYT advertising? Do you think they promised more if the Times was nicer?
    Take a reality check, this is big money.

  9. Barry —
    1.) Unfortunately for us Midwesterners, Minnesotans and others in the area are more,not less obese. Last I saw, Wisconsin had the unfortunate distinction of being the fattest state in the union. Mayo also has a very large Medicaid practice, including Illinois — that’s Chicago, a pretty good match with NYC — Medicaid.
    2.) Mayo, as you probably know, is a world reknowned diagnostic center of last resort, probably facing a larger proportion of “black box” cases than anywhere else.
    3.) The Dartmouth data includes matched data for particular health conditions as well as more general data.
    4.) There is considerable difference in end of life management. That is one of the things that the more efficient centers do much better. Maggie has already cited pounds of data regarding more efficient care in end of life situations.
    The fact of the matter is that there are major differences in cost and quality of medical care as well as outcomes. That has been studied to death for thirty years. Some of the variation is within the same cities — Mayo Scottsdale vs. the Phoenix area, Kaiser Bay Area vs. the SF area.
    This is not a conservative vs. liberal or a market vs. government issue. These are facts. As they say, you can have your own opinions but you can’t have your own facts.
    We can argue about how to fix this problem, but we can’t argue about the existance of the problem.
    BTW, you are right that Mayo, like many famous centers, does extract higher payment rates from private insurers. That is also true of many systems that do poorly in the Dartmouth data. However, many of the other players on the Dartmouth All-Stars don’t have that as a way of fattening their incomes. For example, Kaiser Bay Area, Kaiser Oregon, Kaiser Hawaii, Group Health of Puget Sound, and others are HMO’s that are going head to head with other HMO’s and insurers in very competitive markets. As you yourself have pointed out, Kaiser faces an uphill battle for every contract they get, and does not have any source of money to fall back on.

  10. Christopher George —
    Us non-Harvard doctors should probably hate on Harvard on our own time.
    But yes, the Harvard hospitals are a Dartmouth disaster area, and yes, Gawande doesn’t point that out. Dartmouth Atlas however has a long article documenting Mayo crushing MGH head to head (50% less costly with better outcomes.)
    To Gawande’s credit, however, he does lead his article “The Checklist” by documenting a central line infection disaster on a patient on his own service.

  11. Pat,
    I am sure that Gawande would not improve his chances for tenure by pointing out the obvious.
    It is misleading to leave the public with the idea that McAllen is expensive because of Cecil Rhodes sized greed, but Partners is inexpensive, because everyone is on salary. Both are outrageously expensive. Bureaucratic medicine is just as expensive as Entrepreneurial Medicine! I didn’t read that in the New Yorker.
    It was cheap theatrics to compare with El Paso, which is somewhat medically backward…. and guess what it is cheaper for no less effective treatment!
    I think one of the interesting thing that might come out of real outcomes research is the very small effect medical care has on outcomes, and how hard it will be to demonstrate the improvement.
    Many medically sleepy hollow small cities have been colonized by specialists with modern training, to the towns’ benefit. This has been a method for spreading modern medicine to the hinderlands. I make no excuses for McAllen or Partners.
    The subliminal message is: it is OK for us to waste money like a Congressman, because we’re from Boston… but you guys in TEXAS… why should we spend any money on you.
    Off topic..opps.

  12. Christopher Geroge–
    Let me recommend Dr. Gawande’s book “Complications: A Sugeon’s Notes on a An Imperfect Science.”
    It is all about mistkes and close-calls that Gawande was personally involved in— at the hospital in Boston where he still practices.
    They knew he was writing the book. They didn’t protest. He is still working there– and highly valued.
    Gawande is not a Texas-basher or a Boston promoter.
    As I recall he grew up , the child of two physicians, somewhere in the Midwest.
    Gawande would be among first to say that the DArmtouth reserach shows a large amount of overtreatment in Boston.
    Read “Complications-“-it’s a brilliant book. I think you would like it.

  13. Gawande grew up in Athens, Ohio, with his father a urologist and his mother a pediatrician. I think that means he does understand what it is like to practice medicine in a smaller city in fly-over country.
    However, many of us in the REAL Midwest would consider Ohio to be in the Mideast 😉
    Also, as far as El Paso being backward, their results are very similar to those in McAllen. That is one of the main points.
    Dr. George is right about one thing. Outcomes are much less related to big bang medicine than a lot of people like to think. What they are related to is small-ball medicine: visits with primary care providers who spend time dealing with and improving more basic health and closely managing chronic conditions.
    Not expensive, not glamorous, but it works. That is why the Europeans beat us, not just in overall longevity but also in results with specific disease conditions, including things like coronary artery disease, kidney failure, and so on, despite using a lot less major intervention than we do.

  14. My point exactly, Pat. High tech medicine largely improves at the extreme margin, if at all. Not only does expensive care not improve results in McAllen much over El Paso, it doesn’t improve outcomes much anywhere. It is a lucky thing for us that is true.
    To avoid confusion, lets not mention McAllen without mentioning Partners as expensive places to get care. Or show me the study that compares them.
    If honest outcomes research is actually the guide, we are going to do much much less than we are doing now. There are, of course, a lot of reasons to treat besides improving outcome statistics. Patient expectation for treatment are going to be very difficult to rein in. And we all know what unhappy patients do, rightly or wrongly. They find a new doctor.. they complain.. they make trouble. Don’t hold your breath while you expect utilization to go down.
    Remember, sects that shun all medical care usually live longer than the average American.
    Over-doctored is over treated. Boston is the most over doctored place on earth.
    Maggie, I don’t think the Harvard Hospitals are that much worse than anywhere else. I am just more familiar with them. Trust me, in Cleveland the expert is the one that treats the most; in Boston, it is the one that writes the most.
    It is troubling that several of the nominees for the Medical Central Committee are from Partners. To paraphrase Maggie, Experts are on the bell curve, just like everybody else. You can be on the far left end of the curve in clinical experience and miraculously still be an expert.

  15. Maggie,
    I would advise the readers of this blog to google variations in healthcare expenditures in European countries and within European countries,county by county and institution by instituion. Also Look up the Ave. price medicare pays the Mayo clinic for different operations and procedures at its 3 locations in MN, FL. & AZ. It is eye openning. Regional variation is a world wide phenomenon

  16. jamesd —
    Regional variation is a world wide phenomenon, but spending 17% of GDP on health care and seeing medical inflation of 7% plus each year is not.
    We cannot afford saying “well everyone else does it” anymore.
    We have to fix this, soon, or health care in the US will be priced out of the reach of everyone but the most wealthy, even in Medicare and employer programs.
    The easiest, most painless way, and the way that improves rather than worsens health care for Americans is to learn from the US systems that provide efficient care with best outcomes, and apply those lessons to the rest of US health care.
    If we once get that in order, we can move on to looking at what people do in France, Britain, and elsewhere.

  17. Maggie,
    Richard Cooper from U Penn/Wharton disputes the Dartmouth data. I would hold him up to be a very informed medical researcher/ health care economist. (I must not be informed…since it’s quite obvious you think I’m not.)
    the Dartmouth Atlas maps of the country showing variation in Medicare spending oversimplify a very complicated problem, and they serve no purpose in explaining the causes of local-regional variation. You assume I disregard the Dartmouth data–I don’t. I think the Dartmouth work has been a quite valuable contribution to health policy research. I just think in order for this research to translate into any practical prescription for action, we need to drill down hard into the data at the local level a la Gawande. Generalities get us nowhere. The research needs to be as local…as doctor-patient encounter specific…as possible.
    And I agree with Barry and Pat: it is not meaningful to use Mayo and the Cleveland Clinic as the gold standards. Much more meaningful to use for a standard a non-academic, non-mecca type health care provider IN AN URBAN AREA, such as Group Health of Puget Sound, or Kaiser.

  18. Jamesd & Pat S., Christopher G., Pat S. Barry & Pat S., Joe Says,
    Christopher George, Mike C.
    James d & Pat S.
    James, sorry, I guess I wasn’t clear.
    Medicare pays Mayo differnt prices in differernt regions in part because the cost of living –and so the cost of a doctor’s or hopsitals’ overhead (price of real estate, wages, etc. varies so widely.
    But, as I pointed out in this post, if you simply look at utilization of resources for very similar patients (the number of tests a patient unedergoes, the number of
    days he spends in the hospital, the number of procedures–you see vast differences in how aggressive and intenisve treatment is in different locations, and inevitably, this translates into much more expensive vs. less expensive treatment.
    As Pat S. says, we are spending far more than other countries–about twice as much per capita as the average developed coutnry becasue we over-use advanced medical technology (which includes cutting edge drugs, procedures and tests.)
    Patients in the U.S. under many more diagnostic tests and take more medications. We undergo more surgeries and when we are hosptialized more happens to us while we’re there– more procedures, more tests, than would during an episode of hosptial care in Euopre.
    We see more specialists, and receive less primary care.
    Yet overall, outcomes are no better. In many areas (chronic diseases management, heart disease, etc.) they are worse.
    But as Pat points out there are medical centers in the U.S.–and communitites in the U.S.–that have figured out how to simulaneously raise quality and lower costs, by doing less.
    We need to look at how they do it and apply those lessons to the rest of U.S. health care.
    Christopher G.
    Once again you are spreading misinformation. You write .
    “It is troubling that several of the nominees for the Medical Central Committee are from Partners. To paraphrase Maggie, Experts are on the bell curve, just like everybody else. You can be on the far left end of the curve in clinical experience and miraculously still be an expert.”
    You offer no evidence that the doctors nominated to serve on a Independent Medical Advisory Commission (IMAC) or on those named to the Comparative Effectiveness Research panel are not superb clinicians.
    There are too many unfounded claims ciruclating in the health care debate–sowing confusion and suspiecion.
    I can’t let HealthBeat be part of it, or be used as a plastform for unfounded assertions.
    And frankly, I don’t have the time to present the research to counter your claim every time you choose to make a wild assertion.
    At another point on this thread you declare that doctors in Europe have “absolute protection” against malpractice suits. What does that mean? How? Evidence?
    So from now on I’m going to ask you to provide a link to evidence whenever you make an assertion.
    You are right, however, that healthcare in Boston is more expensive than in any other spot on the globe because there is more overtreatment there: patients spend more days in the hospital, see more specialists, undergo more tests and procedures—and yet outcomes are no better, Sometimes they are worse.
    One if the earliest Dartmouth studies compared care at Yale-New Haven in New Haven, CT to care at Harvard. Far more over-utilization of resources at Harvard. No better outcomes.
    Pat S.—
    Thank you for facts. And thank you for this suggestion to Dr. George: “Us non-Harvard doctors should probably hate on Harvard on our own time.”
    Not a lot of room on HealthBeat for professional jealousies. We have too many important things to talk about.
    Barry & Pat S.
    Barry- Please read Pat S’s reply to your comments.
    He knows the Dartmouth research well—and he knows Mayo from the inside.
    Barry, often you demand that I, or another reader, or the government (Medicare or someone else) provide evidence when the information is already out there—in the Dartmouth research and elsewhere.
    The problem is NOT that no one has studied these issues.
    As Pat points out they have been studied to death. Barry, read the Dartmouth research ( Or the many links that I, Pat and others have provided to research showing what the problems are.
    It’s not that we don’t know what needs to be done to reform healthcare. Conservatives groups such as Cato like to pretend that we don’t really know—it’s a delaying tactic, designed to obstruct health care reform. “Well . . . we’re going to have to study this . . .. Obama’s rushing blindly ahead. (Another aspect of the conservative fear-mongering.)
    See my recent post quoting Dr. Geroge Lundberg on Low-Hanging Fruit.
    If we know what to do , why haven’t we done it? The lobbyists who represent the for-profit corporations that make a fortune on healthcare in this country (and the wealthy Wall Street investors who line their pockets with a the profits made while gouging patients) don’t want to see those revenues and profits reduced.
    Lobbyists also represent some doctors’ groups and some hospital groups that also don’t want to see comparative effectiveness research used to reduce hazardous overtreatment by lower ingfees and raising co-pays for treatments and products that provide little benefit to patients.
    So, please, stop suggesting that we don’t know what to do. This is simply disingenuous.
    Finally, in a recent response to one of your comments, I provided abundant evidence that non-profit medical centers (where docs are on salary and profits are not the goal) are providing better care.
    As Pat puts it, “facts are facts.”
    Joe Says,
    Yes, I do think that at least some at the Times are responding to pressure from Manhattan’s most prestigious hospitals. This doesn’t necessarily mean its all about ad sales.
    NYT executives and editors are treated by doctors who practice at these hospitals. Some mix socially with these doctors. If they need to be hospitalized, they will wind up at these hospitals. They want to believe the doctors who tell them that care is more aggressive and expensive at these hospitals because sophisticated New Yorkers (“people like you Dan”), won’t settle for anything less than the best. New Yorkers are more discriminating than other patients.
    “That’s why we put you through so many tests. You know it’s not the money, Dan. Heck, there’s hardly any money left in this business. Medicare won’t pay us enough. At one time, you could count on a seven-figure income, now . . .”
    Finally, if they or their loved ones become seriously ill, these executives, editors and reporters are likely to wind up at NYC’s brand-name hospitals. They desperately want to believe that these are the safest, the most efficient hospitals in the country—that they provide the best care. They need to believe that the most expensive care is the best care. Because that’s what they’re getting.
    Bottom line, the Times and the people who work there are part of NYC’s culture. And NYC is a city that is all about money. As Dr. Atul Gawande wrote in his June 1 New Yorker article, every town and city has its own medical culture.
    This is not to say that every NYC physician or every hospital administrator is money-driven. I know quite a few who are not—and who I greatly admire. I’ve written about people like Dr. Diane Meier on this post. A Manhattan physician has written a superb guest-post for HealthBeat. When I wrote the book I interviewed a very powerful physician at NYU Medical Center who was very honest about over-treatment.
    But, both as a patient, as the wife of a patient and at medical forums, I have encountered NYC hospital administrators, and some doctors, who become very angry and very defensive, if you suggest that we need to cut back on medical spending.
    Christopher George— You write: “It makes you wish someone would compare the case controlled costs of McAllen with Brigham and Womens. If they could find comparably sick patients, it might make McAllen look like a bargain.
    As Mike C notes in the comment below yours, the appropriate comparison, which Gawande did, is between McAllen and El Paso where the demographics are nearly identical.
    Mike C.–
    As far as I know, they don’t teach data analysis in med school—at least not the type of data analysis that would question the received wisdom about more care. . . .
    What I think we need in medical schools is one course on how physicians can help reduce unnecessary spending, to make sure that we are getting value for our healthcare dollars.

  19. jms–
    Richard Cooper’s work has been quite thoroughly debunked by Dartmouth Reserachers in Health Affairs here:
    I’m afraid that Cooper a shill for the American Association of Medical Colleges and AAMC an organizaion that is lobbying that the U.S. shouldl fund residency slots for More Specailists- at a time when it is very clear that we have too few primary care docs–and too many specialsits– making
    health care more expense.
    But Cooper is part of a lobbying effort that argues that we need to increase medical spending and that Americans should be seeing more specialists more often.
    Thus, once Cooper knew that he had a paper accepted by Health Affairs (which was trying to represent all points of view in a issue that offered varoius perspectives on spending and heath ccare,) Health Affairs editor-in chief Susan Dentzer reports that he
    “sent copies of the accepted papers in advance of publication to an official of the Association of American Medical Colleges (a fact confirmed by both Cooper and the AAMC). Copies of electronic correspondence also show that the papers went from Cooper through a separate channel to officials in leadership positions at the American College of Surgeons, among others, as well as to U.S. congressional staff. ”
    Dentzer points out that “This advance distribution came at a time when the AAMC is urging a 30 percent increase in medical school enrollment, and when it and other physician groups are seeking federal legislation to increase the number of Medicare-supported residency programs at the nation’s teaching hospitals. ”
    In other words, Cooper was trying to use the fact that his paper had been accepted by Health Affairs as a tool to push legislation–before
    others in the medical community had a chance to respond to the paper, and point out the many flaws in his research and reasoning (See below)
    Dentzer goes on to explain that Cooper violated Health Affairs’ “embargo”policy: once a paper has been accepted for publication, it should not be circualted as something about to appear in Health Affairs until it has been published. Dentzer explains why this policiy is important::
    “Embargos, meanwhile, serve several critical purposes. At the most basic level, they allow the editors of Health Affairs sufficient time to make certain that papers are truly ready for publication. (For example, we were still correcting errors in Cooper’s papers just days before publishing them.)
    “Equally if not more important, an embargo allows us to release papers in a coordinated fashion that upholds the journal’s editorial mission of providing an array of views on one subject. In our experience, there is seldom, if ever, a single body of received wisdom on a particular topic in health policy. We often seek to publish simultaneously a broad array of research and perspectives on a controversial issue, so that arguments on various sides can be presented and assessed together. Our ability to publish a range of views is a central reason that readers value Health Affairs, and why the journal is of particular importance to policymakers.”
    But Cooper, and the groups he was shilling for, wanted Congressoinal staffers to see his paper in isolation–not in the context of other perspectives which might well lead readers to see the faults in his reasoning.
    Because he violated the scholarly standards that virtually all medical journals adhere to–after being warned on four occasions– Dentzer writes:
    “For violating Health Affairs’ publication policy, we will not ask Cooper to serve as a peer reviewer for journal articles and will not accept any submissions from him for five years, until 4 December 2013. This decision follows the lead of other journals, such as the New England Journal of Medicine, that have responded to similar breaches of their standards.”
    To be banned from publishing in a highly-respected journal like Health Affairs tells you something about how serious Cooper’s breach was as he attempted to use Health Affairs brand name as a lobbying tool.
    Finally, a summary of what was wrong with Cooper’s paper from the Dartmouth reserachers:
    “First, he ignores statistical inference, reporting statistically insignificant correlations as if they were more definitive. Second, he confuses correlations (which tell whether or not two variables are related) with the magnitude of the effects (which tell you how big the relationship is), and thus fails to recognize that his exhibits suggest that generalists have an effect on quality that is ten times larger than that of specialists. Third, as we illustrate with an example, his rejection of multiple regression in favor of considering each variable only in isolation results in misleading inferences. Fourth, his research fails to meet scientific standards in other important dimensions.”

  20. For those mystified about McAllen vs Partners (BWH), here you go–right from Atlas (MMC/BWH). Yes Virginia, there is a difference:
    McAllen Medical Center (McAllen, TX) Brigham and Women’s Hospital (Boston, MA)
    HCI Index
    HCI Index Percentile compared to all U.S. hospitals
    96.0 70.8
    Medicare Spending Report:
    Medicare spending during last two years of life per decedent
    Total Medicare spending
    $79,650 $87,721
    Inpatient site of care
    $45,214 $50,156
    Outpatient site of care
    $12,068 $14,518
    Skilled nursing/Long-term care
    $9,087 $13,633
    Home health care
    $5,436 $4,943
    Hospice care
    $390 $1,302
    Durable medical equipment
    $2,295 $1,012
    Resource Allocation Report:
    Resource inputs during last two years of life per 1,000 decedents
    Hospital beds
    84.34 72.42
    Intensive care (ICU) beds
    28.50 23.24
    High intensity
    22.31 11.87
    Intermediate intensity
    6.19 11.37
    FTE physician labor
    All physicians
    38.87 29.26
    Primary care physicians
    17.86 9.82
    Medical specialists
    16.46 12.31
    Ratio of medical specialist to primary care labor inputs
    0.92 1.25
    Required RNs (proposed federal standard)
    81.54 66.05
    Patient Experience Report:
    Care during last six months of life
    Hospital days per patient
    17.73 16.13
    Physician visits per patient
    65.37 37.06
    Percent seeing 10 or more physicians
    59.62 55.01
    Percent of deaths in hospital
    47.52 39.27
    Percent of deaths with ICU admission
    34.66 26.18
    Percent enrolled in hospice
    12.20 28.25
    Average co-payment for physician care per patient during last two years of life
    $6,095 $3,729

  21. Maggie,
    Point taken about Cooper. We need more Gawandes however doing the trench work to get at root causes locally. McAllen does not explain all of America.

  22. jms–
    Thank you.
    And thank you for callig Cooper to my attention. The next thing I knew, there was his op-ed in the Washigton Post.
    I’m now doing a post about him.
    I spent quite a bit of time researching a response to your comment: I sensed this guy could be dangerous and do quite a bit of harm and wanted to warn you and other HealthBeat readers.
    I agree that McAllen’s money culture is just one part of the problem.
    The doctor-owned hospital in McCallen and what sounds like excess capacity
    (beds, diagnostic equipments, etc) is another part of it.
    Dr. Gawande didn’t highlight that part, but nationwide, excess capacity is at least as important in driving over-treatment as a “money culture.”
    I’ll explain more in my post about Cooper.

  23. Brad, you have to use “case-control” methodology to have any validity. Essentially you match a McAllen patient with one at BWH who is equally old, equally obese, has had their diabetes ignored for an equal interval, etc. before you can make a meaningful comparison. Having driven through West Texas, I am not sure you could actually do that.
    This was not done in the comparison with El Paso either, though we are told that the patients are from similar sleepy towns in far off Texas. The article was published in the New Yorker, and not in a peer reviewed medical journal.
    Money driven McAllen care, or status driven Harvard care are both extremely expensive. The point is we can’t afford care at either of these places.

  24. Chris
    I am well versed in biostats and epi methodolgy, and understand data quite well thank you.
    This data did not come from NY Mag, but from the (adjusted) Dartmouth Atlas itself. You should visit.
    If you dont buy these numbers, please contact Fisher et al, not me.
    That is all i will say on the matter.

  25. Brad, as you are familiar with statistics then I am curious why you would confuse lumped data from the Dartmouth atlas with case controlled data.
    You are most welcome.
    This is the sort of analysis that made newspapers believe that President Dewey won in 1948. The Dartmouth data is very useful, just not for this purpose.

  26. Brad F. and Christopher George–
    As regular readers know, I am concerned about the amount of misinfromation about healthcare circulating out there, and don’t want HealthBeat to be part of that. .
    For the record, BradF. has the facts.