Why “More is Not Better”: Patients Bring a New Perspective to the Discussion

Kenneth Raske, president of the Greater New York Hospital Association…attributed the aggressiveness of private hospitals in New York simply to the sophistication of the patients and their families…"that’s a reflection of the New York culture that we have.” -The New York Times

All together now: when it comes to medical care, more is not necessarily better. This is a point we’ve driven home here on Health Beat, but a JAMA study published this week adds a new dimension to the discussion by showing that even patients feel that more is not better.

The research, authored by experts from UMass Boston, the Foundation for Informed Medical Decision Making, and Dartmouth, surveyed 2,515 Medicare patients across the nation to find out how they felt about the medical care that they received over the past year. The survey specifically asked whether patients felt that their needs were met, how they perceived the quality of ambulatory care, and how they perceived the  quality of overall care.

The researchers took responses to these questions and matched them up
with per capita Medicare expenditures for 306 hospital referral regions
(HRRs). Here is what they found: where Medicare spent more patients
did, indeed, receive more care, but they didn’t appreciate it. Indeed
“the perceived quality of care was higher in the lower-expenditure
areas
[i.e. those that provided relatively less care].”

How much more did Medicare spend in the high-spending regions? "The mean
per capita expenditure varied from approximately $5200 in the lowest
quintile to $8500 in the highest quintile.” It’s worth noting that
workers in low-spending regions (like Iowa) contribute the same
percentage of their paycheck to Medicare as workers in high-spending
regions (like Manhattan).  So folks in Iowa are paying for that
Cadillac care in Manhattan—and New Yorkers  don’t even appreciate it!

As the researchers point out:  “…spending more on medical care does not
improve patients’ perceptions of the medical care they receive. On
average, those in the lower-expenditure areas reported no more unmet
need” and “they rated the overall quality of their health care at
least as highly as those in high-expenditure areas.”

Regular Health Beat readers know that the evidence shows  that more
spending—and thus, more treatment—can lead to worse outcomes. But it
turns out that over-treatment isn’t just undesirable in an academic,
technical sense. It also has a real impact on the experience
of being a patient. It’s a reminder that spending on health care isn’t
just a game of numbers—it’s real life, lived by real patients.
Given what we know about high spending and health outcomes—much of it
from the Dartmouth researchers—it’s actually not surprising that
patients in high-expenditure HRRs are less than thrilled about their
medical care.

A 2003 study
led by the Dartmouth team showed that higher end-of-life spending in an
HRR is actually associated with increased mortalities. Writing in the
Annals of Internal Medicine, researchers reported that when they looked
at “variations in Medicare spending” that are  not  connected to
regional differences in pricing or regional differences in the overall
health of patients,  they found that while patients in hospital referral regions
with higher end-of-life spending received 60 percent more care than
lower-spending areas, each 10 percent increase in regional end-of-life
spending was associated with a higher relative risk of dying among
patients with hip fractures, colorectal cancer, acute myocardial
infarction—and even among the broader cohort of all Medicare
beneficiaries.

So why the excessive spending  in some regions? This dynamic is due to what the 2008 Dartmouth Atlas
calls supply-sensitive medicine, or “care that varies with the local
availability of such medical resources as physicians, hospital beds,
intensive care unit (ICU) beds, and diagnostic imaging equipment.”
According to Dartmouth research, the supply of physicians, beds, ICUS and
equipment is  “by far the most significant factor associated with how
much Medicare spends in any given region.”

Back in 1961, Milton Roemer, a professor at UCLA, codified this dynamic
through “Roemer’s Law.” The law states that, in a fully-insured
population (like patients on  Medicare), the more hospital beds a
community has, the more hospital care patients will receive. When
everyone has access to care, empty beds will be filled simply because
they are there. Availability predicts treatment and spending—which is
inversely related to health outcomes and, we now know,  patient
satisfaction.

How does over-treatment harm patients? ”Hospitals “are good places to
acquire antibiotic resistant infections,” which means the more time a
patient spends in the hospital—“as they do if they live in regions that
deliver more care”—the greater chance there is that they’ll be exposed
to such infections. Further, “because most treatments pose some risk,
providing those treatments to patients who do not need them could cause
harm. And as care becomes more complex and as more physicians get
involved in an individual patient’s care, it becomes less and less
clear who is responsible, and miscommunication—and medical
errors—becomes more likely” (for more on this last point, see my
previous post on the “hand-off”).

In an editorial
accompanying this week’s JAMA study, Gerard Anderson (from Johns
Hopkins) and Kalipso Chalkidou (from NICE in the UK) make clear that
this holds true in an international context as well. The authors note
that an OECD study found that “the correlation between level of health
spending and scores on most indicators was relatively low. The United
States, by far the biggest spender on health care, scored high on some
indicators, such as breast cancer survival, and average or low on
others, such as in-hospital mortality rates following myocardial
infarction or mortality from asthma.”

Patient satisfaction at the national level is also not related to
health care spending. Anderson and Chalkidou note that a “surveys of
satisfaction conducted by the World Health Organization and by the
Gallup Organization  across a wide range of countries have found only
minimal correlation between higher levels of health spending and levels
of satisfaction with care or between health spending and the broader
notion of system "responsiveness," a measure of how well individuals
believe the health care system responds to their own needs.”

Of course, not everyone agrees with this analysis. Today the New York
Times
reported that New York’s private hospitals are among the most
aggressive of about 3,000 hospitals studied across the nation, ranking
in the 94th percentile as a group. (New York’s  public hospitals landed
in the 69th percentile.) In private hospitals like NYU and Lenox
Hill elderly patients in their last two years of life had more
intensive care and spent more days in the hospital.

The level of care was most intensive at New York University Medical
Center, where 65 percent of patients saw 10 or more physicians in their
last six months of life. NYU ranked in the 99th percentile
nationally.

Kenneth Raske, president of the Greater New York Hospital
Association,  attributed the aggressiveness of private hospitals in
New York simply to the sophistication of the patients and their
families and the desire of doctors to give them the best possible care.

“The patients and physicians and their families, of course, are trying
to live longer and beat whatever malady they have,” Mr. Raske said,
“and that’s a reflection of the New York culture that we have.”
But if New Yorkers are just more sophisticated than everyone else, why
is it that at Bellevue hospital, a public hospital that is  just a
stone’s throw from NYU, only 7 percent of patients see 10 or more
specialists during the final six months of life?

Dr. Eric Manheimer, who is the medical director at Bellevue and on the
faculty at NYU, told the Times that “ ‘having a foot in both the
public and private systems gave him a unique perspective on the
discrepancies.’ He said that care was less aggressive at public
hospitals because most of their doctors — he estimated 75 to 85 percent
— were salaried physicians with little financial incentive to order
tests or other interventions. At private hospitals, he said,’supply
can create its own demand: There is often an abundance of beds and an
endless list of specialists who can be called.’

“’You end up with the phenomenon of specialists referring to other
specialists, with nobody coordinating, which results in confused
messages, more referrals, more hospitalizations, deterioration in
health care and a more anxious patient," Dr. Manheimer said.

Finally given the results of the JAMA study regarding spending and
patient satisfaction, one can’t help but wonder: Is it the patients or
their families who are demanding more care at NYU?  How many
patients really want to be poked and prodded by ten or more
specialists? How many ask to die in an ICU?

19 thoughts on “Why “More is Not Better”: Patients Bring a New Perspective to the Discussion

  1. >> “It’s worth noting that workers in low-spending regions (like Iowa) contribute the same percentage of their paycheck to Medicare as workers in high-spending regions (like Manhattan). So folks in Iowa are paying for that Cadillac care in Manhattan—and New Yorkers don’t even appreciate it!” <<
    But on the flip side… shouldn’t Iowans appreciate the fact that Manhattanites pay a lot more into Medicare on a per capita basis, simply because the New Yorker’s average income is much higher?

  2. Tim–
    Not really. Manhattanites with high salaries lead a much richer lifestyle than Iowans. . .
    If we were going to be fair, Manhattanites would pay a significantly higher Medicare tax–or Medicare would simply pay hospitals and doctors in Manhattan less than they pay them in Iowa (which they have talked about doing, but will never do.)
    The high-spending states are more powerful in Congress (NY, Calif, Mass, Florida etc) and so this is not something that Congress likes to talk about.
    But Congressional Budget Office Director Peter Orszag (a good guy and very smart) keeps bringing it up.
    The irony is that the higher spending in NY, etc. is doing the patients any good . …

  3. There is a methodological fault with studies like this. They only study patients who die. Put another way they only study medical failures.
    We don’t know how many people who saw seven specialists found one who successfully treated what the other six had missed. Seeing specialists and having extra tests is the cheapest part of medicine, especially when a course of chemo for cancer can cost $100K or ongoing treatment for kidney failure can also run into the tens of thousands.
    I know there are studies which show that there is over treatment and unnecessary treatment, but I haven’t seen any measures of successes. Using the death rate or the number of months to death is a poor indicator.
    We apply a strict moral measure for trials (in principle): “better 100 guilty people go free than an innocent man be convicted.”
    What measure should we use for treatment? If a therapy has a 10% chance of a cure but costs $100K should it be tried? The UK has an economic measure it uses in such cases, but by making it a strict financial measure they have ducked the moral issue.
    They wouldn’t disallow paying for such a treatment if it cost $1000 and had the same odds of success. All this goes to show is that society’s priorities have been distorted. We see nothing wrong with spending $1 trillion on warfare, but balk at $100K for health care.
    To say that health costs are too high or are projected to become “unaffordable” when military spending continues at its present rate is a sorry comment on our national sense of values.

  4. Robert–
    Actually there isn’t a methodological problem here.
    They are studying Medicare records of patients who are already dead, looking at what happened to them in the final six months of life.
    They are not studying patients who are alive and then only reporting on the ones who died.
    The reason to look at patients after they are dead is that you can do an apples to apples comparison. In a very, very large sample (which this is) you are looking at cardiac patients, diabetic patients etc. who were more or less equally sick. HOw do we know that? Becuase they all died six months later.
    There is now general agreement within the medical establishment that we over-treat and that patients are harmed as a result.
    It’s not just that we are spending too much. The over-spending and over-treatment is hurting patients.
    I’ve written about this here–http://dartmed.dartmouth.edu/spring07/html/atlas.phpit
    and here http://dartmed.dartmouth.edu/winter07/
    will give you a better sense of the research and the harm done.
    Think of tonsillectomies in the 1950s. We’re doing a lot of procedures like that today.

  5. Hi there,
    I loved your blog! I wanted to extend an initiation to blog at our site. I think you could offer some very interesting content to our blogs. Plus, it would be a great opportunity to gain exposure. Our site currently receives approximately 1.2 million unique visitors a month and we are one of the fastest growing online health communities. New blogs are featured on our home page, as well as recent posts and we also feature the blogger with the most quality content on our homepage with a mini bio, which changes each week.
    Or you could feed existing posts to our site as a means to drive additional traffic to your site, if you are too busy to write an additional blog.
    If you are interested, you can sign up here http://blogs.healthcare.com otherwise feel free to email me.
    Have a great day!
    Alexandra Snyder
    Content Editor
    HealthCare.com
    1749 NE Miami Ct, Suite 604
    Miami, FL 33132
    asnyder@healthcare.com
    305.371.9724 x224

  6. Sorry, I still disagree. There is an implication that because there is excessive end of life treatment that there is also excessive treatment at other times. We know some of this is true, but as I said above, there is no way to measure the successes using this technique.
    Over treatment is not excessive when it works, even if it turned out to be the third or fourth thing tried. All that is demonstrated is that medicine still has to rely on trial and error more than we would wish.
    In addition there is still the issue of cost. Many heroic treatments are quite expensive. If they weren’t we wouldn’t be having these sorts of discussions.
    Obviously the public is still unwilling to face the “pull the plug” discussion head on. I think this is because people are having inner conflicts over religious teachings and their own experiences seeing others die unpleasantly. We still have laws against suicide, which make no sense from a practical point of view, but are strictly based upon religious precepts.
    If we could get the excessive profit out of health care, perhaps we could then focus on the moral issues more clearly.

  7. It’s Friday, so some slightly less serious — although informative — comments:
    In one of the neonatology areas at Georgetown, a vendor’s poster said “The first five minutes are the hardest.” Someone had written underneath, “the last five aren’t any picnic, either.”
    It is true, Maggie, that I had a tonsillectomy, guessing the early sixties. At the time, the idea that a patient, much less a child, would know anything about the procedure was beyond the comprehension of many, if not most, physicians. My tonsillectomy, however, brought a measure of justice.
    Having read up on the procedure, I knew the most common cause of postoperative hemorrhage was coughing. Through immense self-discipline, I don’t think I coughed more than once.
    The attending ENT, however, thought my progress was due purely to his surgical skill, and was waking me constantly to show off his work to students and house staff. Not only did he forget the patient might have opinions in the matter, he forgot that to show the site of a tonsillectomy, one must place one’s fingers in a certain area.
    Waking me for the fifth or sixth time, and inserting his fingers, he started his self-congratulatory lecture, interrupted by
    *C*H*O*M*P*
    Very deliberately, I relaxed my bite until he stopped screaming. Making eye contact, I very deliberately held gently until I was certain I had his attention.
    With all the dignity I could summon at age 12 or so, I croaked out, “No, Doctor. It was not your unilateral contribution. It was a factor about which you didn’t bother to warn me, and I took the trouble to do.
    “Now, go away until you do something useful for my throat. I don’t care which you bring back, but when I next see you, I expect ice cream, morphine, or jello.”

  8. It’s so simple ask the patient if they are going to die in the next 6 months. If they say yes then cut all of their benefits.
    We’ve had this problem since the beginning of time. The problem with death is, you usually don’t know when its coming. It’s a bummer we can’t plug people into our little charts and graphs so our statistics look better.

  9. Robert-
    Not only have the Dartmouth researchers looked at what happened to patients during the final 6 months and during the final two years by looking at records of patients who are already dead, it has also spent years tracking patients receving treatment for chronic diseases who are alive.
    And as the reserachers explain in the 2008 Dartmouth Report on treatment of patients with chronic diseases:
    “The most surprising and singficiant difference between regions is that mortalities are higher in high-spending regions. In other words, your chances of dying increases in regions that deliver more care.”
    They have shown the same thing at the hospital level, comparing hospitals that provide less care and less intensive care (fewer days in the hospital where the patients sees fewer specialists and has fwer procedures) to hospitals that provide more intensive care. More pateitns die in teh hospitals when they receive more intensive care.
    I know it seems counterintuivie that more care consistenntly kills more patients. But this isn’t just my opinion. And it’s not based on one set of studies. If you looked at the links to the articles I included in my last response, you’ll see that these findings are now widely accepted by phyisicans and physician organizations around the country.

  10. ASnyder, HCBerkowitz, Jenga,and Robert
    ASnyder– thank you! I’ll definitely look at the site.
    HCBerkowitz–good story –
    I also had my tonsils out–as did virtually everyone back then. Very unpleasant and, we know realize, completely unncessary in most cases.
    Jenga–
    Exactly– we don’t know when people are going to die. So we can’t cut back on their care simply because we think they are going to die (or appear to be in the final months of life.)
    As Diane Meier, the palliative care specialist I’ve written about puts it: “You can’t say ‘we’re spending too much on the dying’ becuase you can’t know who will actually die. And these are the sickest people in the hospital. So of course we should be giving them the best possible care.”
    But, she adds “we have to ask is it APPROPRIATE care?
    For example if someone appears to be dying of cancer and also has some other problems, do you operate to try to correct the other problems?
    Probably not. You don’t want to torture the patient needlessly. At most, you explain to the patient that he has this other problem, and that you could do an operation, but, given how sick he is, he might not survive it.
    Palliative care specialists don’t give up on the patient–in other words they don’t stop treating him for the disease that appears to be killing him.
    But they do make it clear that the patient has options as to which treatments he wants and which ones he doesn’t–explaining the odds of benefits and risks with each treatment.
    And in the meantime, the palliative care team (which consists of a doctor, nurse and pscyhologist specifically trained in palliative care) makes sure that the patient who appears to be dying isn’t in pain. And they talk to him about how much pain-killer he does or doesn’t want.
    This seems to me the gold standard for end-of-life care–treatments are offered, someone spends time sitting down and talking with both the patient and the patient’s family about the potential benefits, risks and side effects of these options.
    And the palliative care team protects the patients from others in the hospitals who, instead of asking what the patient wants, simply say :” Okay, here’s what we’re going to do now.” So the kidney specialist puts the patient dying of cancer on dialysis; another specialist decides to perform another procedure;
    a neruologist decides to order an MRI to try to figure out what the cancer patient is experiencing vertigo– you get the picture.
    Everyone is working fee-for-service, paying attention to that part of the patient that he specializes in. The palliative care team looks at the whole patient–mind as well as body–and respects his wishes.
    Robert-
    I responded to you separately a little while ago-see below.

  11. key question, apparently unaddressed, is whether Raske is correct when he says that NY patients and their families demand more. fact that rich get more than poor doesn’t seem to address that. poor could ask for T-bone steaks, too (with resulting cardiac risk), but they wouldn’t get them and their resulting hunger wouldn’t tell us much about dietary habits of the rich. is there any real data on whether patients in heavily served areas demand more?

  12. key question, apparently unaddressed, is whether Raske is correct when he says that NY patients and their families demand more. fact that rich get more than poor doesn’t seem to address that. poor could ask for T-bone steaks, too (with resulting cardiac risk), but they wouldn’t get them and their resulting hunger wouldn’t tell us much about dietary habits of the rich. is there any real data on whether patients in heavily served areas demand more?

  13. Jim–
    A good question. The Dartmouth folks have wondered about this, and what they have discovered is that if people move out of an area where there were a lot of hopsital beds and specialists (and consequently patients were getting a lot of expensive, aggressive care) and new people buy their houses and move into the area–the hospitals and specialists continue to provide more aggressive, more intensive care, even though this may not be what the new people are used to–or demand.
    So, in healthcare, the doctor, not the patient, usually drives demand.
    (The Dartmouth reserachers figured this out by looking at places where something like a defense plant closing caused a huge exodus, many home sales, and new people coming in.)
    The other thing we know is that, in hospitals where there are more beds and more specailists per patient, the expensive, aggressive services they provide are not things that most people would ask for: the chance to die in an ICU,
    a third round of chemo, a second operation in a ten-day period because due to a lack of co-ordination of care (too many specailists not communicating with each other) the first operation wasn’t really that urgent, but the second one is; a chance to spend twice as many days in the hospital as you would in a hospital where Medicare spends less; the opportunityto pick up a hospital acquired infection while spending more time there . . .
    Wealthier patients probably do demand more attention–ring for the nurse more often, complain about the food, and want to see their doctor frequently.
    But it’s not as likely that they want 10 or more specialists poking them and prodding them. If you’re very sick (and these people are) that’s just too many doctors to feel comfortable with.
    Many wealthy people have someone they call “my doctor” (often a cardiologist or orthopedist rather than a primary care physician) and that’s the person they want to see. They may also want a second opinion–especially if there is someone known to be “the Best” in that particular specialty.
    So I guess what I’m saying is that the stuff wealthy patients demand (attention, pillows, an extra consult) is not what drives the big-ticket spending that makes care twice as expensive in some hospitals.

  14. I also thought the NYT article was very interesting. What particularly struck me, however, was that Bellevue scored in the 69th percentile for aggressive treatment despite most of its doctors working on salary and despite treating large numbers of low income people. I wonder how many for profit hospitals elsewhere in the country and/or hospitals where doctors work on a fee for service basis and not salary scored well below the 50th percentile.
    In could well be that doctors in NYC, whether they are paid on a salary or on a fee for service basis, just have a more aggressive practice pattern that defines their community standard. Some of that aggressiveness may be a function of a surplus of hospital beds, specialists that can be called in, and other resources. Some might relate to the large number of Academic Medical Centers that train some 16,000 new doctors each year or about 15% of the nation’s total and twice what the local market requires. It could also be that patients, including those from lower socioeconomic groups have higher expectations than elsewhere and are quicker and more willing to demand aggressive treatment including at the end of life.
    If regional differences in local physician practice patterns and culture, along with differences in patient expectations are a significant driver of regional cost differences that are unrelated to outcomes, just moving more doctors from fee for service reimbursement to compensation via salary may not be as effective as many people (including myself) might have thought it would be.
    I think over the long term, we are going to need a credible and reasonable approach to rationing that has what Normal Daniels and James Sabin in their book, Setting Limits Fairly, call “accountability for reasonableness.” The body ultimately charged with setting those limits will have to be perceived by the public as having the moral authority to do so. Unfortunately, we’re a long, long way from that point.

  15. Barry–
    Bellevue’s patients probably do need more care because Bellevue sees many poor people.
    The poor know where they are and are not welcome. Word-of-mouth tells them that if they try going to the ER at NYU Medical Center down the street they will be turned away. (Legal Aid in New York documented this a couple of years ago.)
    A couple of residents who worked at both hospitals told me– go into the ER at Bellvue, and you’ll heart 4 or 5 languages being spoken, see people of all colors–many obviously very poor.
    “Go into the ER at Bellevue and who’ll see guys in shirts and ties lying on gurneys with someone apologizing that they’ve been kept waiting.”
    Bellvue’s ER is actually very good Ed Koch always said he would go there if he had a heart attack.
    My point about Bellvue is that, once a poor person lands there, they probably are suffering from 4 or 5 other conditions that need treatment.
    The Dartmouth study does adjust for race adn underlying health of the population, but it’s hard to adjust for differences in the population that two hopsitals a stone’s throw from each other attract.

  16. More is not necessarily better when it comes to healthcare

    The rising cost of health care, slumping economy and increasing numbers of uninsured Americans continues to bring health care issues into the spotlight.  In addition, the upcoming presidential election feeds the rhetoric for health care reform.  As pol

  17. Congratulations Maggie on this blog post! Not sure if you noticed, but it won an Honorable Mention on our Redscrubs.com site for this weeks Scrubby Awards! Keep up the good work.

  18. Reforming health care in the current economic climate

    Some are saying that the economy will prevent any drastic health care reform.
    In an op-ed (via the WSJ Health Blog), Ezekiel J. Emanuel argues that major reform will be chump change compared to the $700+ billion spent last week.
    He asserts that recent e

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>