The recent RAND study that suggested that there were few or no differences in the quality of treament of minorities when they got into the US health system has not been taken lying down by those who believe that there are great differences, and that ignoring them to look at the bigger picture, as the RAND researchers suggested, is not the way to go. Brian Smedley, the research director at The Opportunity Agenda wrote this opinion piece criticizing the RAND study for THCB:
Ask anyone who’s worked in, received treatment in, or studied American hospitals and health care systems, and you’ll find broad agreement: U.S. health care systems are beset by quality problems. Information systems don’t "talk" well with one another, medical errors remain all too common, and many patients don’t receive the types of treatments and services that they should.
Recently, a new study expands on these problems, finding that treatment is mediocre at best for all patients, regardless of race, ethnicity, gender, or income. This study, published March 16 by physician Steven Asch and colleagues in the New England Journal of Medicine, finds that, on average, patients receive a little more than half of the care recommended by a set of "gold standard" guidelines. And even though few patients are well-served, women and minorities were found to fare better than whites and men in receiving recommended care.
This finding wouldn’t be so shocking to most Americans, who tend to believe that health care is fair for all groups, even if less-than-stellar. Many Americans tolerate (and therefore tacitly accept) that fact that minorities are more likely to be uninsured or underinsured, or to live in communities that lack high-quality primary and specialty care, access-related problems which have profound implications for quality of health care. But once patients are in the health care system, we believe, race or ethnicity doesn’t matter.
This view, however, squarely contradicts what the vast majority of research studies have found for decades – that some patients, most notably African Americans, Latinos, those who don’t speak English well, and in some cases, women – receive a lower quality of health care than their counterparts, even when they have similar health insurance and are treated for the same health conditions in the same hospitals. This applies across the gamut of health care, ranging from basic services such as screening and immunization, to primary care, to more expensive, high-tech, specialty procedures.
These are the conclusions of literally hundreds of studies published in peer-reviewed journals over the last two decades. And while a few studies, such as the Asch study, find that disparities are diminishing or that all groups receive equal (albeit poor) treatment, their findings must be considered relative to the massive volume of evidence to the contrary. Even the U.S. Department of Health and Human Services’ National Healthcare Disparities Report, released in January and which represents the most comprehensive survey of its kind, finds that, despite some areas of improvement, racial and ethnic healthcare disparities persist, and are worsening in some areas. For example, the NHDR found that Latino patients with diabetes are receiving poorer quality care today than they were even a few years ago.
Importantly, the Asch study confirms that quality problems abound. But the authors’ conclusion – that "quality-improvement programs that focus solely on reducing disparities among sociodemographic subgroups may miss larger opportunities to improve care" – unfortunately presents a false choice. Policymakers are not confronted with the question of whether to prioritize efforts to reduce inequality, or efforts to improve overall quality. Rather, these problems are inextricably linked. The persistence of healthcare disparities is a clear warning that systemic problems plague our health systems. Moreover, many of the same interventions that will reduce disparities – such as promoting the broader use of evidence-based guidelines and public reporting of hospital quality scores by patient race, ethnicity, and primary language – will help to improve quality for all patients.
Progressive health care advocates should seize upon health care disparities as a key political issue and an argument for stepped-up quality improvement efforts. Unequal health care is not only wrong, it’s one of many signals that American health care is reeling from systemic problems that hamper the best efforts of hard-working physicians, nurses, administrators, and others to provide the best quality care. Our concern and focus should be on raising the quality of care for everyone, with particular attention to those who are least well served.
Categories: Uncategorized
An impressive share! I have just forwarded this onto a colleague who had been conducting a
little research on this. And he actually ordered me breakfast due to the fact
that I stumbled upon it for him… lol. So allow me to
reword this…. Thank YOU for the meal!! But yeah, thanks for
spending the time to discuss this topic here on your site.
Most of the RAND criteria they are using for quality judgment is only their opinion. Only a few of their recommendations are endorsed wholesale by all the medical organizations and researchers.
> As far as delivering an aspirin…
This is quite simply, insulting.
t
Well, I guess I have to disagree, and maybe even get in the last word or two.
As far as delivering an aspirin to a heart attack patient. What if that patient had a low platelet count, or was on blood thinners? Would it have been a good idea to give him aspirin, or even necessary?
I’m also not so sure that giving a pneumonia vaccine to someone with pneumonia would be very effective, although I am not a doctor.
And just how is it that you characterize medical uncertainty?
Well I believe that many are trying to do just that, and the NEJM report is one example, IMO
> And reports, such as the NEJM report, that tries to
> define quality, by relating it simply to receiving
> the recommended treatment are even more meaningless.
No, not meaningless. If we were in (say) a three-sigma world with respect to healthcare you would have a fair point. But we’re not. We are not even in a two-sigma world. In the world we are in, we can’t reliably deliver the recommended aspirin to heart attack patients when they come into the ED. We can’t seem to deliver discharge instructions — minimal patient education — to people whose health is demonstrably quite at risk. We can’t seem to vaccinate pneumonia patients. The list goes on. We are not trying to discover small, nuanced differences in approaches to treatment or patient preferences.
When you ask about the patient being “too far gone” and so-forth, those questions fall under the ruberic of “risk adjustment” — everyone knows perfectly well that bad outcomes are not necessarily the doctor’s (or hospital’s) fault. Even a patient’s going AMA (Against Medical Advice) can be accounted for just fine. Poor patient compliance short of this is a tougher problem, but there are methods for characterizing that too. Biomedical uncertainty is a fact of life in medicine. Notwithstanding, the uncertainty can be characterized, and we must stop using uncertainty as an excuse not to measure or publish the measurements.
> And hopefully they have learned from me, not to jump
> into something just because it might be recommended.
> There are many things to consider when deciding on a
> treatment plan.
Doctors should recommend what is recommended. One hopes specialists have more knowledge about recommended or even promising alternatives than their patients have, but this is not always so. If there are no (clear) recommendations, then the docs and patients will have to navigate that. Fine. In any case, a recommendation is not all by itself a decision — docs should be forthcoming about the burdens as well as the benefits.
We all know that medicine is complicated. Nobody is trying to reduce quality of care to a single dimension.
t
Tom, I should have clarified physical condition as being athletic condition, but hopefully you knew what I meant.
Thanks Tom, but don’t feel too sorry for me. It’s not as bad as you think, at least not yet? 🙂 While I would much prefer not to be sick, I consider it a blessing in disquise, which I do try to indicate in my blog. It has given me a much different outlook on life, which I think many would benefit from. Also if we met, you would never know I was sick, plus, and although I don’t know you, there is high probability that I am in much better physical condition than you.
Anywaze, enough of that. Let’s go on. If I am using the terms quality and outcome inconsistently, I don’t mean to, but it may be because I have been using quality where I should be using quality of care.
So to clarify somewhat, an outcome can be good or bad, and so can the quality of care. Together they define quality.
If the outcome is good, than it is likely that the quality of care is good, and by definition quality is good.
But if the outcome is bad, it does not mean the quality of care was bad. And it does not mean the overall quality was bad. The quality of care could actually have been the best possible, and the treatment could have been the most appropriate, even if it wasn’t what was recommended. After all that is why we seek out the advice of doctors. But because the condition of the patient i.e severity of the injures, or disease progression, was such the patient had little chance of a good outcome. Plus the patient could have actually been treatable, and the treatment appropriate, but if the quality of care given (I talk about that later) was poor, the outcome could be bad.
At that is what can’t be quantified. At what point is the injury or disease too far gone, and how is that impacted by the quality of care?
Now the quality of care can be measured. I consider the quality of care to include but not limited to, such things as the attentiveness of the staff while under care, cleanliness, the frequency of infections (related to cleanliness) during a hospital star, or even how many instruments are left in patients during an operations.
The problem is trying to lable the quality of health care in the US without taking into account both outcomes and the quality of care are meaningless ventures. And reports, such as the NEJM report, that tries to define quality, by relating it simply to receiving the recommended treatment are even more meaningless.
I am sorry to hear about your cancer. It seems to me you use the terms outcome and quality a little inconsistently. As for what you said about measurement, all I can do is refer you back to the fourth paragraph of your first post in this thread. Even if you meant to say you don’t think the quality of treatment can be fairly measured, I would still object.
Patient satisfaction is an important part of quality; important enough to completely overwhelm a particular clinical outcome. It is true that some patients will find the benefit/burden ratio of the recommended treatment unattractive. Some dialysis patients stop dialysis: you evidently did not want bone marrow replacement. I am glad your doctors were willing to find an alternative path for you. And I hope your clinical outcome is as good as the quality you have experienced.
t
Boy Tom, you love using those big words that I have to keep looking up. Interloqutor! If you were talking about me, I didn’t really say outcomes couldn’t be measured, I said they couldn’t be quantified with any accuracy without taking into account how severely injured a person is, or how far along the diseased has progressed in any given person. That is what cannot be measured or quantified.
I should add that my opinions are based on my own experience. When I was diagnosed with NHL, over 4 years ago, my doctors were ready to give me the recommended treatment. That is they were ready to infuse deadly chemical into my body, give me total body irridiation, and then perform a stem cell transplant. For my type of lymphoma, that was the treatment that was considered the best available, i.e. recommended, since my lymphoma was considered extremely aggressive. (You can read more about my experiences on my blog. You have to start from the beginning though. I have been doing a lot of rambling lately.)
Had I taken their advice, would I have been better off? I seriously doubt it. But they would have gotten high scores, while I would have endured 3 to 4 months of pure hell, with long term effects that would be too numerous to mention here, not to mention the risk associated with a transplant.
The point I am trying to make here again is that we are all different. Even with the different subtypes of lymphoma, there are subtypes within each one of those, and no one knows for sure what is the best treatment or even when to treat. There is still so much more to learn.
Fortunately, the internet exists, and I was able to learn a great deal from a great many experts around the world. And fortunately my doctors listened to me. They were very receptive to what I told them, and to this day, I feel basically as good as before I was diagnosed, and am probably in better shape, despite my abnormal blood counts. Not everybody is that lucky or that informed.
And hopefully they have learned from me, not to jump into something just because it might be recommended. There are many things to consider when deciding on a treatment plan.
> Tom you said that if nothing was known about
> outcomes data, that going to a doctor would be
> worthless.
I did not say that. What I said was:
Contrast the meaning of what I said with what I think you misread me to mean:
My interloqutor said he does not believe outcomes can be measured. He thereby attacked the very notion of outcomes research and, by extension, treatment recommendations. He is wrong, and you apparently agree he is wrong.
If a medical practitioner says to me:
I will likely say something to him I will not write here and run for the door. Something worse than this has happened to a dear friend of mine. I suppose Marc would say “well, since we can’t possibly know that the recommended treatment will work at all for me, let’s go right ahead.” Or would he?
I had not brought this up, but I do not criticize doctors because everything isn’t standardized yet. I do criticize doctors who, in order to protect their professional autonomy, criticize the standardization project itself; or who say that since measurements are not perfect, their individual performance should not be measured at all.
I do not think that many doctors should be researchers, and some researchers shouldn’t be either. I think treatment recommendations are essential: good as doctors are, medical knowledge is too vast for any one doctor to hold adequately in his head and keep up with.
If we get into a non-trivial case where there are no recommended treatments, I think the doctor should choose one from a menu of treatments under investigation and follow it, reporting what happens. If we are going to have big EHR and DSS projects, this is where I want them to go: everybody is basically in a clinical trial all the time. Eventually we should find the treatments that work best even for the unusual cases. Pie in the Sky, I know; many problems to solve. But Steve Beller will like it .
t
I read the study again, and some of the “quality” criteria in there is a joke.
One of the criteria that doctors were penalized on was whether or not they asked about seat belt usage.
I’m sorry but thats not the doctors job. Alot of the stuff in their “quality” criteria list is very shady to say the least
Marc and Tom are both right to some extent, but neither is right exclusively.
Health care outcomes CAN BE measured and standardized, provided that sufficient inclusion and exclusion criteria exist.
Take for example the thrombolytic drug TPA. TPA has been proven beyond any doubt to have a positive effect in treating acute stroke, but ONLY for those patients that meet a laundry list of inclusion and exclusion criteria. Whether TPA is effective in the aggregate for people with stroke who dont meet those criteria is just guessing. Some docs say yes, others say no. Thats why ER docs, who have no training in neurosurgery or neurology, are allowed to script out TPA in the ER. The ER docs have an agreement with teh neurologists about when its appropriate to use TPA. The ER docs have a very specific protocol they have to follow, very much a “cookbook” approach to TPA.
Marc on the other hand is correct that much of medicine is not standardized. To standardize treatments and measure outcomes requires years of research. Furthermore its very difficult to set up these studies because factors vary tremendously between patients.
Medicine is both art and science right now. Eventually as medical research continues, it will shift more towards science and away from art. The art of medicine comes in with uncertainty.
I dont think its fair to criticize doctors because everything isnt standardized yet. This isnt making widgets. Standardizign widget production is infinitely easier than standardizing patient treatments. Patients are far too different from one another to approach it in that fashion. At the same time, doctors SHOULD be pushing for evidence based medicine. We should not be satisfied to just throw our hands up in the air and declare that nothing can be certain so we should just leave it up to individual doctors and patients to decided everything.
Evidence based medicine is what we need to push for, and we are doing that. But its going to take a LONG TIME to push EBM thru to every treatment modality and diagnosis. So in the meantime, we have to make do with what we have in terms of medicine being art as well as science.
Tom you said that if nothing was known about outcomes data, that going to a doctor would be worthless. I disagree. A layperson who understands nothing about biology or medicine is less likely to be “right” than a doctor who has training in both. Take for example a patient who is complaining of musty body odor, convulsions, and mental retardation. Lets assume that there is no evidence one way or the other as to which way to treat this condition. (This disease is PKU).
If you go to a layperson with these symptoms, they are going to have no clue what you are talking about. They would have absolutely zero clue as to where to start.
If you go to a doctor, although they might not have any outcomes data to prove that their treatment really works, they are more likley to be right than a layperson would, because at least they have some idea of where to look and understand something about basic biochemistry.
So just becomes outcomes data is not available for some things does not mean the doctor is worthless. He might not have proof of efficacy, but he’s more likely to steer you in the right direction than anybody else would.
Well, I might consider my statement polemical, but certainly not a canard!
Have you ever heard the statement doctors practice medicine.
Certainly the recommended treatment may increase the odds of a good outcome, but only in the average person with an average severity of disease or injury. Oh if we were all only average, how simple things would be.
And medicine is anything but irrational, but neither is it necessarily “rational” for lack of a better term. There is no one absolute right answer, and that is the reason we go to doctors. It is their training and expertise that we depend on to determine what the best treatment is in in any particular situation, not just in the average situation.
If we used the guidelines set up in the NEJM study, we wouldn’t need doctors. We could just have a technician read from a manual, much like tech support for your computer, and administer the proper treatment.
OK, for you quality means I walk out of the hospital, or doctors office, in better shape than I went in. Fair enough.
The authors of the NEJM report take as a given that following recommendations increases the odds of a good outcome on your definition. You are making the Cookbook Medicine criticism, which is thoroughly discussed elsewhere. I think it is a strawman argument, and a polemical canard.
If you don’t think outcomes can be fairly measured, then all medicine is a spin of the wheel and it is irrational to seek the advice of a doctor because there is no basis for your belief that you might walk out in better shape than you when you went in.
t
The way I see it, there is a difference between the quality of care one receives, and whether one receives what others, unfamiliar with a case, have determined to be the recommended care.
Everyone of us is different, and it is the doctors job to determine what the appropriate treatment should be in any particular situation, and it may not be the recommended treatment.
I have read the NEJM report, and was not impressed. It did not indicated the outcomes of those who were treated. Did those who receive the recommended treatment have better health outcomes than those that received some other treatment?
I have not read any of the other reports indicating disparities in treatment among those of different ethnic or socioeconomic backgrounds, so I can’t speak to them, but I suspect they don’t address outcomes either, because I don’t believe outcomes can be fairly measured.
Since we are all different, and present in different stages of disease progression, or in the extent of injuries, trying to measure outcomes without taking into account the likelihood of anything being effective, is worthless. Doing so would only result in many doctors and hospital not attempting to treat the severely sick or injured to insure good quality scores.
As I have said before on other blogs talking about health care quality, I don’t care whether I receive the “recommended” treatment or not, I just care whether I walk out of the hospital, or doctors office, in better shape than I went in.
Thanks Tom. I fixed the goofed up link
I have read the putative challenge by The Opportunity Agenda people. Asch studied the narrow question of disparities in treatment quality among those who successfully “accessed” medical services. The Opportnity Agenda prefers he study treatment quality (not) delivered to people who have not accessed medical services. It seems to me Smedly is upset because Asch did not study the question he wanted him to study. Because of this, the oh-so-polite advocates intimate that Dr. Asch is not “serious”, and that the NEJM has tarnished its own reputation by publishing his “unsound” paper. As for their specific challenges: The mismatch between the study population and the overall population is not important in any way I can think of to the question being asked. The selection bias problem is probably real, but this problem is well-understood in survey research. It looks to me like most of the data came from medical records anyway. To contradict 600 previous studies does not prove that this particular study is wrong, even if this study and the 600 others truly studied the same question. I do not know why this canard was thrown out at all. Well, yes I do…
I find the comments of the experts unhelpful at best.
t
I think the link underneath ” patients receive a little more than” text has an extra ‘http//’ in it which produces truly suprising web navigation results…
t
Hi, this is Mike Connery, I work with Brian at The Opportunity Agenda. I would also like to call attention to the fact that there has been something of a small uprising within the health equity community about the Asch study, and a coalition has formed to challenge the findings of Asch.
A group of experts, a number of advocacy organizations, and 5 congressmen have all cosigned a letter addressed to the New England Journal of Medicine challenging the findings. You can read the letter and view the signatories here:
http://www.opportunityagenda.org/nejmletter
Matthew, thanks for printing this piece. Brian, thanks for writing it. Your comments on this study echo those made by Dr. Richard Allen Williams, founder of the Association of Black Cardiologists, about this study. If you haven’t seen his comments, please go to my blog, Envisioning 2.0 (http://fardj.prblogs.org) and look under the “Race and Medicine” category.
Thanks for your great addition to this conversation.
Thanks for your comment, Eric. There are several important steps that must be taken to improve equity and accountability in health care, particulary where public funds are involved. The first, as I suggest in the opinion piece, is to establish mechanisms for data collection and public reporting of hospital quality scores by patients’ race, ethnicity, income or education, and primarly language. Several organizations are already involved in efforts to create these kinds of disparities “report cards,” and of course hospital quality data are widely available (but are typically not disaggregated by patient demographic factors). The move toward electronic medical records will help this effort if the collection of patient demographic data is standardized. Second, broader acceptance of the use of evidence-based guidelines will help improve the quality of care for all patients, particularly where these practices are supported by the use of information technology, such as bedside electronic decision supports, as currently used in Veterans Administration Hospitals. Third, we should expand the use of trained lay health navigators to help patients better understand how to utilize health care systems and ensure that their needs are addressed. These are among a few of the strategies that will help to level the health care playing field.
As you point out, Eric, addresing health care disparities shouldn’t be a partisan issue. In fact, all three of the prescriptions above have been promoted by Democrats and Republicans alike (notably, Senator Frist introduced a bill with many of these elements, as did the Congressional Black, Hispanic, and Asian/Pacific Islander Caucuses). More importantly, these policies will likely help to improve care for all patients. What is needed now is the political will to implement them.
Ok, Brian.
You have my attention.
But you have no solution other than to say
“Progressive health care advocates should seize upon health care disparities as a key political issue and an argument for stepped-up quality improvement efforts”
What would you do first? and then second? and then third?
Your conclusion of “Our concern and focus should be on raising the quality of care for everyone, with particular attention to those who are least well served” is not one that has party-specific resonance.