Consumer Reports on Health: Worse Than Average


Maybe no one at Consumer Reports has a mother.

The first rule of effective consumer information is “tell it to Mom.” That is, explain why something is important in the kind of language you would use when speaking to your mother. Unfortunately, the folks at Consumers Union have now, for the second time, put out purportedly pro-consumer health care information that no one’s mother could love. Their latest offering is at best mildly helpful and at worst seriously misleading. The only explanation I can think of is that the CU folks believe so firmly in their own good intentions that they ignore the impact of what they are actually doing.

And as long as we’re awarding demerits to the self-delusional in the
Liberal Do-Gooder category (Conservative Market-Worshippers richly
deserving a category of their own), save a sigh of exasperation for the
creators of something called the Healthcare Equality Index.

A quick disclosure: I am a subscriber to Consumer Reports online. I have
known some individuals there for years and did some brief consulting to
them some years back. Therefore, let me be clear that my criticisms are
based exclusively on public information.

CU’s latest offering is a link to the Dartmouth Atlas of Health Care
and its analysis of hospital practice variation. The CU imprimatur
suggests this is Major League quality information. The truth is closer
to the story of the baseball manager who blurted out about one of his
players, “While he can’t hit, neither can he throw.”  In that same
vein, while the Dartmouth information is not easy to use, neither is it
easy to understand.

The Dartmouth Atlas is a health services research tour de force. It
takes an incontrovertible outcome (death from a serious chronic
illness) and then applies a complex algorithm to account for all
medical resources used during the two years before that outcome. The
result is more evidence of practice variation and that “more” care
isn’t always “better.” However, expecting this insight alone to prove
useful to someone making a decision about their own or a family
member’s medical treatment means you’ve spent way too much time in an
ivory tower.

How should a CU subscriber interpret the fact that one hospital is
“aggressive” and another “conservative” in its resource use? In trying
to reassure us that we can choose “conservative” and less costly
treatment, CU explains
that “aggressive care does not necessarily improve patient outcomes and
can sometimes shorten life." But wait: doesn’t that mean aggressive care
sometimes does improve patient outcomes and lengthen life? If I listen
to CU, there’s no doubt I’ll save money for the federal government (the
Atlas uses Medicare data), but if I don’t, I may save my life.

Maybe the folks at CU do have mothers. They just don’t like them very much.

“We’re not at a point yet where we can say which hospitals are bad or
good," Dr. John Santa, director of Consumer Reports’ new Health Ratings
Center, acknowledged to Chicago Tribune reporter/health blogger Judy Graham.

Whatever happened to, “First, do no harm?”

By the way, the Chicago hospital most aggressive in its use of
resources before the patient died was a Catholic one, and the hospital
with the lowest score was the public hospital. However, the Dartmouth
Atlas makes no allowances for the impact of religious beliefs in regard
to life extension or the impact of patient socio-economic status (or
hospital financial status) on intensivity of treatment.

This is the second time CU has been unable to distinguish between
information that sets the blood of researchers racing and information
real people might reasonably rely upon. Its ConsumerReportsHealth site debuted with “Best Treatments” from the Cochrane Collaboration evidence-based medicine reviews.

This carefully constructed tool from the folks at Oxford University
gives us epistemological categories such as “treatments that work but
whose harms may outweigh benefits.”

Come to think of it, that’s a logical precursor to the Dartmouth
University rankings of hospitals where the harms (an adverse event that
kills you) may also outweigh the benefits. Unless, of course, they

The good news is that CU plans to publish genuine cost and quality
information on hospitals by the end of 2009. I hope no one tells them
that some of those evil for-profits have been doing this since the late
20th century.

Now on to the Healthcare Equality Index.  You might assume this is a clever way to call attention to the disparities in care that we know cause serious harm to African-Americans, Hispanics
and other minorities. It is not. Instead, it constitutes the responses
to a survey designed to measure “how equitably hospitals in the United
States treat their gay, lesbian, bisexual and transgender patients and

The survey was designed by the Human Rights Campaign Foundation Family
Project and the Gay and Lesbian Medical Association. A whopping 88 out
of roughly 5,000 U.S. hospitals voluntarily responded and agreed to
make their names public. Perhaps some folks have had too little
exposure to the ivory tower and the idea of “statistical validity.”

Yes, discrimination against GBLT patients and employees can be a problem, and GLBT individuals’ special health care needs certainly can get short-shrifted
by uncomfortable straights. But if you want to talk seriously about
health care equality, then get serious. The well-educated,
well-insured, solidly upper-middle-class white gay couple is far more
likely to get the health care they need than the working-class,
dark-skinned, Spanish-speaking couple with no health insurance.

Even if they have a subscription to Consumer Reports.

Michael L. Millenson is a consultant, author and policy analyst based in suburban Chicago.

9 replies »

  1. I agree with you, Michael. Much of the so-called consumer information out there is impenetrable by average “moms.” Jane Sarasohn-Kahn in her recent Calif. HealthCare Foundation issue brief on health care and social media, quotes Jude O’Reilley, an expert in health information exchange online, “A common part of the consumer’s health experience is to face a health challenge, Google it, spend 20 minutes getting totally overwhelmed, and then do what they did in the 1980s: Call a friend and work through their offline
    social networks.”
    It gets back to carefully thinking about your real audience, not the initiated, inner-circle of colleagues with whom the developers of these tools interact with daily.

  2. As the writer of the piece in AARP, trust me, we are writing a very oversimplified version of the Dartmouth data. The hope was indeed that it would begin to get patients/consumers thinking about the possibility that more isn’t always better, which is the prevailing view. I think CU did a pretty good job, given the how counterintuitive the evidence is.

  3. Having covered health policy for three decades for newspapers, radio and now an online news service, I know how incredibly difficult it can be to explain the quirks of the system in terms that make sense to the average person — or even to brainy Ph.D’s. So I applaud CR for making a stab at translating the Atlas to the public. Even if readers don’t use it for comparison shopping, the way they would a CR issue on cell phones, they should catch on to the central message, that an excess of medical treatment wastes money and time for those who have precious little of it to spend with loved ones. Not to mention that it creates LOTS of paperwork to cope with. It will take years for that counter-intuitive point to sink in, but only then can we rein in the juggernaut. By lending its credibility to the issue, CR is providing a public service. — Carol Gentry, editor, Florida Health News

  4. For 20+ years I have advocated making data about our health care system publicly available knowing that the first attempts would be difficult even clumsy. And they are, no surprise.
    The benefit to this public disclosure is not that consumers aka patients use it but hospitals (and eventually) physicians use it and change (hopefully improve) their processes and outcomes if for no other reason than they can’t hide belove average performances anymore.
    Thanks for all the dialog on this — it helps improve and makes data into information patients will find user friendly. You shouldn’t have to be a policy wonk or professional health researcher to understand this data.

  5. For 20+ years I have advocated making data about our health care system publicly available knowing that the first attempts would be difficult even clumsy. And they are, no surprise.
    The benefit to this public disclosure is not that consumers aka patients use it but hospitals (and eventually) physicians use it and change (hopefully improve) their processes and outcomes if for no other reason than they can’t hide belove average performances anymore.
    Thanks for all the dialog on this — it helps improve and makes data into information patients will find user friendly. You shouldn’t have to be a policy wonk or professional health researcher to understand this data.

  6. To my friends Jim and Maggie:
    I have been writing about practice variation and money wasted by ineffective and inappropriate medical care well before either cause was a glint in the IOM’s eye, much less on the radar screen of Consumer Reports.
    Yet there is a large difference between information that people should have and be concerned about and the kind of actionable information that can help an individual decide about a particular hospital for a particular condition. If your mother had chronic heart disease, would you deliberately pass up “aggressive” care, or would you say, “Thank God these doctors care so much”? Even those of us who are died-in-the-wool wonks, those of us who stay out of the health care system unless truly necessary, those of us who believe that less is more — even we might hesitate when talking to our siblings, Dad and Mom herself. How about your spouse? Or your child?
    I love Elliott Fisher’s work and am delighted that the Dartmouth site is available for use by all. I applaud Consumers Union for giving publicity to the Atlas and to evidence-based medicine and the Cochrane Collaboration. However, as I said above, CU has seriously overplayed how usable the information is in real life for real decisions by real individuals.
    Meanwhile, Marilyn, thank you for your feedback. In my review of the site, I found information that was outdated in certain instances and information that contradicted information posted at a different time on different CU sites. This is distressing.
    I respect CU enough to judge them on what they’ve done, not what they intend to do now or in the future.

  7. I subscribe to both Consumer Reports and Consumer Reports on Health. I am thinking of cancelling my subscription to CR on Health because the information it presents is oversimplified and unreliable. In addition, the list of “consultants” they publish in each issue is misleading because there is no information tying any particular “consultant” to any particular piece of information or even any particular article. In many cases, the experts CR spoke to would not endorse some of the content of CR on Health. I know this for a fact, because in February 2008 CR on Health had an article entitled “The No-Drug, Low-Drug Approach to Heart Health.” On the cover page there was a box that recommended that people lower their doses of statins or blood pressure drugs, apparently without consulting their physicians first. There were several cardiologists listed as “consultants” for that issue. I spoke to two of them (Elliot Antman, M.D. and Howard S. Weintraub, M.D.) and they were quite unhappy with being (by implication) associated with this advice, which they in no way shape or form endorsed. In addition, the article’s recommendations with respect to specific dosages of statins or blood pressure-lowering drugs seem inappropriate, since such matters should be determined based on each patient’s risk factors, medical history and other relevant considerations. I wrote a letter to CR on Health regarding this article. I received a form letter in reply.
    The article discussed both prevention and treatment of heart disease. The recommendations discussed above seem particularly irresponsible in the case of people who already have heart disease, since those people are at very high risk of heart attack, stroke and death.
    Maybe CR should stick to rating lawnmowers and washing machines.
    Marilyn Mann

  8. Michael–
    I’m a fan and usually agree with your work.
    But this analysis of the Dartmouth reserach isn’t quite on target. Their decades of reserach shows that, overall, outcomes are no better when care is more intensive and agressive–and often,they are worse.
    In addition to worse outcomes, both patient satisfaction and doctor satisfacation is lower at places like UCLA Medical Center than at UCSF or Mayo or Intermountain.
    At medical centers where care is more aggressive (and Medicare pays twice as much) patients consistently spend more days in the hospital, undergo more procedures and are more likely to be readmitted.
    I agree that the Dartmouth Atlas info can appear overwhelming–that’s why, when I blogged about Consumer Reports’
    hospital report, I told my readers how to get to “Help” on the Atlas (it’s just two clicks.) Then, if you know your zip code,you can begin to compare hospitals in your area.
    There are a few reasons to find out which hospitals are more aggressive and which practice more conservative medicine: a) if outcomes are overall no better, and often worse at the more aggressive hospital, why take the extra risk? And why undergo the pain of unncessary procedures and the risk of infection if you spend more days in the hospital?
    Finally, these days, even if you have pretty good insurance, if you’re hospitalized, you’re likely to find that “insurance doesn’t cover everything”–and you come out owning $15,000 after a 6-day stay. Why go to a hospital that consistently costs Medicare twice as much–for very similar patients?
    What’s important is that Dartmouth finds such consistent patterns, over a period of years, that are very clearly linked to too many hospital beds and too many specialists leading to unncessary care. The folks who run UCLA Medical Center privately acknowledge this. They just don’t know what to do about it. They can’t close down wings of the hospital or fire specialists–they need the income stream to pay off their debt. (And they have said as much)
    Of all of the things we know about how to rate the quality of hospitals the Wennberg resarch really is the gold standard. As Consumer reports makes clear: this is not the definitive “report card” that we’ve all been waiting for. But for the well-insured, over-treatment is one of the biggest–if not the biggest–hazard in our healthcare system.
    Consumer Reports readers are lucky to have the link to the Atlas.

  9. Hard to argue with Millenson’s argument, but my perspective is different. I see the latest CU efforts, including this month’s magazine article( and similar piece in the new AARP magazine) as blunt tools that make a very important point the public seems generally unaware of — more care doesn’t yield better outcomes and a lot of people appear to be getting more care than they need. Once the public digests that basic point — and it hasn’t yet — it will be possible to move forward toward explicitly defining and attempting to squeeze out the surplus care. without that realization, it won’t. obviously, it would be neater to have the luxury of moving in a straight line, but clearly isn’t an option here.