this is the most malfeasance explanation of care and is perhaps one reason why there is so much interest in this mechanism, but disparities they also rise from in prison discrimination, where a herridge provider, operating in a time- sensitive -- they also rise from a discrimination, where a busy provider, operating in a time since the environment, makes choices. if, for example, african- american patients are less likely to be compliant, then a physician may assume that her african american patient is less compliant. such reasoning will worsen outcomes for that patient. if he is different from the typical african american patient, it will worsen if for all african-americans if the stereotypes about them is wrong. this is compounded with poor communication between providers and their patients, which may generate enormous psychological

barriers to minority patients seeking care. finally, some researchers have posited differences among patients while others have discounted such conclusions. given the discussion of the mechanism, the question is asked if we have worked on this. answering this a very simple question carefully poses an enormous and formidable empirical challenge. we would need to observe the same provider treating to the impatience with the same economic and social resources, physiology, clinical history, severity, preferences, compliance, and feature prognosis. these variables are routinely observed by providers treating patients but not by social scientists observing providers. the fact that multiple studies all noted that minority patients get less care is often interpreted as pervasive bias in the clinical encounter, but it could just as well be interpreted as one of the basic

shortcomings and all observational studies that focus on the clinical and counter. second, because of patterns of segregation, the same provider is rarely observed treating both white and black patients, so what we have been coming prejudice and the clinical encounter is often a difference in neighborhoods, referral patterns, and the resources of providers that serve in these neighborhoods. this is an unfortunate conclusion, because improving neighborhood schools or change your referral patterns is not the same thing as reforming provider behavior in some hospitals and offices. researchers have made some progress in this challenge by using implicit association tests, where physicians and researchers studying them observe the same information in a laboratory situation. this is an increasing area of academic research, but the findings are still nascent for policy and legislation. we do not know if the decisions made by self elected positions in these laboratory studies are representative of physicians who actually take care of minority patient populations.

my main point today is to elaborate on a new explanation for the disparities in racial care, but they are partially the consequence in differences in where minorities and whites receive their care. if different providers treat whites and blacks, then differences and disparities is not always who you are, or race, but also bring you live. but sources of disparities are injury is to minority health. the first type of variation, which i call within provider variation, is the role of the clinical encounter. the second, which i call between provider variation, has to deal more with the quality of treatment patterns geographically for all patients. if minorities are more likely to be cared for by lower performing providers, some academics center's art in exception to this statement, but the link between being treated at one of these centers and quality is by no means automatic -- some academic centers are instance -- an exception to the statement.

it surely exacerbates this variation. confronted with these realities, we should be extremely cautious in concluding that malfeasance and another is the sole purpose of the medical profession, so what is the evidence in the role of geography in determining disparities with health care? one doctor and his colleagues demonstrated that blacks and whites have different providers, and those that tree providers are was clinically trained and have fewer resources. my collaborators and i have demonstrated that 85% of all black heart attacks are treated in on the 1000 hospitals, where 60 percent of whites receive care in hospitals that treat no african-american patients. within hospitals, we found no disparities in effective care but found the patients were admitted to hospitals that disproportionately serve blacks had a risk of mortality that was 20% higher than that of non minorities are in hospitals. others have noted similar

findings for the performance of neonatal intensive care units and minorities serving hospitals. 40 years after passage of the civil rights act, moderate health care is both de facto, separate, and unequal -- minority health care. segregated hospitals was the original motivation for title six legislation. the new focus on the geography of minority health care should not be taking attention away from the clinical accounting. if we could fully eliminate disparities within the clinically counter -- encounter, the house of blacks would improve but would still lack -- black behind whites. -- the health of blacks would improve but would still lag behind whites. targeting quality improvement with minority providers will dramatically reduce blood-wide disparities in care and would improve the health of both minority and white patients, but the gains would disproportionately go to

minorities. in the context of ambulatory care, my collaborators at dartmouth and i estimate that improving the performance of the 500 largest minorities serving networks would improve minority health care more than a complete elimination of racial disparities within every provider in the united states. indeed, given the greater reliance on ambulatory care, when might want to think of expanding tunnel six of the civil rights legislation to go beyond this care and good to care delivered in office visits and in managed care plan. finally, in closing, let me make one final point. -- expanding title 6 of civil rights legislation to go beyond this year and go to care delivered in office visits. neighbors come economic circumstance, help the secondarily affected by health care but more likely to be influenced by prevention, including the quality of ambulatory care, which can check the progression of diabetes,

hypertension, or chronic disease, and through this, incidences of stroke and heart attack. at the very end of the causal chain is the world of disparities. the quality of medical care matters much more than the disparity in the quality. in the 6.5 year life expectancy gap and the 4.5 racial gap for women, which are larger when one accounts for the conditions, they are unlikely to be affected on the focus -- with the focus. the preoccupation with trina disparities is an end game, and it simply misses the fact that many minority patients simply find themselves confronting the end game sooner than everyone else. thank you. >> thank you. doctor? >> thank you for the invitation to -- whoops -- to address you today. is this better?

thank you, chairman reynolds and co-chairman. my name is sally satel, and i work at a methadone clinic in northeast washington, d.c., and i wanted to give you an overview today of the contours of the health despair to debate. now, what do i mean by debate? there is certainly no controversy over the fact that minorities have poor health status and often poor health care, both variables which dr. chandra spoke of, but the debate has to do with the causes of those differentials, and the causes, of course, lead us to the remedies. so there are two starkly different perspectives on the causes, and dr. chandra spoke to

them already. he characterized them in short as dynamics within the clinical encounter, biased positions, and even body is to help systems. that is one perspective -- by his physicians. and even body is to help systems. -- biased physicians and even biased health systems. there is the 2002 institute of medicine report, and i want to spend a little bit of time on that. that report is called on equal treatment. it got an enormous amount of attention and is largely regarded as an authoritative study. however, i wanted to outline some of the mythological problems with it. that study added emphasis on the clinical and counter and concluded that there was bias

among physicians toward minority patients overtly as well as heavily -- as well as supplely -- subtley. this really catapulted the issue of minority health from a public health issue to a civil rights one. as i said, with the different types of remedies implied, sensitivity training for doctors, affirmative action, and even potential tidal six legal challenges. the main problem with the institute of medicine report is that it sought to prove bias or discrimination, and i just speak from the standpoint solely of methodology. this is almost impossible to prove using retrospective approaches, and using large databases. in a sense, charging bias is a

diagnosis of exclusion. it is the kind of thing that you arrive at, which is not to say it does not exist, it could well, but it is one of the things that you arrived at after ruling on variables that we can measure and identify, other kinds that would lead to differences, and with large databases, this is very hard. one problem with the report, i have chronicled most of them in this health disparities handout that will give you afterwards, one of the problems with this is that of abominable variables, and you referred to this, as well. you're not often going to find the types of variables physicians make their clinical decisions on. for example, if we are going to use an angioplasty, we would want to use certain he can subtleties that are not in large retrospective basis. -- we would want to use ekg

subtleties. there are very relevant to clinical decisions. another feature of the report, much work on health disparities come is procedure counting. how many procedures to one group get versus another, as opposed to looking equally or even with greater emphasis on the outcome. there is research and cardiac procedures as frequently shown that even though there are differentials, the mortality rate -- rates are frequently the same. and i think we want to use our prospective studies and graphic interviews with physicians as to why they made between the decisions that they do, and i really am not familiar with those kinds of studies, but for years, the iom report has set the tone of the debate. i think there is less talk about highest positions today, and i see that as a definite maturing of this issue, but

there is still a absent -- emphasis on one key comment, which is, again, within the health disparities issue, an almost exclusive focus on relative health. a greater concern with the help of groups in relation to each other than whether people are receiving optimal care. the reason why this is one of the problems -- one of the promise of this approach to look and relative health is that you can often miss improvements when all of those are there. you will see no change in the ratio of minority and white improvement, but it could be there. you just will not see it because everyone has improved together. another example of that has to do the classic example of black infant deaths, which between the years 1980 and 2000 decreased by over one-third. now, that is certainly progress, but white infant death increased even greater, so it still looks

as if the ratio is unfavorable to blacks with infant mortality, but that is not really what the whole picture shows. one can be misled by focusing on relative health and on gaps, and one also can get, conversely, a false sense of achievement and the 2005 harvard study found greater improvement in blacks with basic interventions with diabetes and other things. however, the rates for whites and blacks were suboptimal for both, so that was not something to celebrate easter. now, as far as the most relevant determinants of health, you have already heard them from dr. chandra and the other doctor, physician quality, hospitals, the idea that minorities and whites really do not see the same positions, these are drivers and are very powerful. there is the value of looking

for bias, assuming that bias could even be satisfactorily and empirical lead demonstrated. now, perhaps even more profound than the demographics are the early determinants that have been mentioned by my colleagues. the mechanisms are very complex. they are, upstream factors of parental income and neighborhood. the mechanisms are complex, but scholars generally agree that good structured education in the early years enables children to develop self control, problem- solving dispositions, and not least, a sense of the future. now, what does this mean for health in later life? well, obviously, it means more opportunities to obtain decent jobs, jobs with health benefits, more autonomy. that is one of the key lessons in the civil steady which looked at the ingredients of income and found that the sec it to highest gradient still that

disproportionately cardiac mortality than one would expect even though their incomes are very good, but the conclusion was said they did not have the latitude to determine how they worked on the job, and it was a sense of stress, a sense of that without authority, says stress is quite important. also, good education given to the financial security to cushion setbacks. people are better informed, of course, about health matters and have a much positive view of intervention. now, let me move from the more abstract to what i see when i go to my clinic here. the methadone clinic, so, by definition, we are treating people who have a her one addiction, but they also have a lot of other medical problems -- people have a heroin addiction. these are factors that vary by rates, not necessarily because of race, -- these are factors

that vary by race. we see folks that i think everyone is really talking about when we think about the medically disenfranchised. now, there is no question that improved access to care would help these folks, especially black men, who rarely have medicaid unless they are disabled, but even so, there is much more to better health than access. as i mentioned before, there is a continuity of care. a medical home, ample time. the commonwealth fund and one for poll in the late 1990's. i wish they would repeat it. over 1000 people, about eight different ethnic groups, and they asked them about so many questions about how they determine which practitioner they want to go to, and race came in a tight last with something else. people should certainly have the option to choose their physician based on race, but the participants in this survey, and come as a said, it was over

1000, that is the least relevant. the doctor spending time with them was the most, said that is the most important. access, again, as a said, is huge, but determinants often have to do with engagement. will patients engage in the self care that you mentioned. that is so important because so many of chronic illnesses, which contribute tremendously to the health care burden, so i, personally, for this kind of problem, as for my kinds of patients, actually, i am a great fan of local public health clinics. this is very much on the ground. we are talking 5 ft above, not the 1,000 foot in view. we are open nights for the working poor, a location that is convenient, keep people out of the emergency room, staffed with local residents. this goes to the cultural sensibilities that we're all talking about.

these physician assistants and nurses, they help the physicians. they do this support. the to the outreach. they do the follow-up calls, so, in summary -- time for my summary? ok, ok. three points, very quickly. recognize that the elimination of health the differential is not feasible because we cannot eliminate the disparities, the social disparities, many of which take their most profound toll in the terms of habits. health, such an agenda clearly transcends the work of public health and is best left to politicians, voters, and social welfare experts also out there. thank you. >> thank you. doctor? >> thank you very much for this invitation. i am really thrilled that you're having this, and i want to say metaphorically that the fact that i am jammed at the end of the table i find it to be a

great turn of events that the jewish white guy finally feels marginalized. [laughter] i suspect it was just an accident. my name is peter bach, and i am a physician at the memorial sloan-kettering cancer center. this is a hybrid of techniques from economics, epidemiology, and statistics in order to gain an idea about the health care system. for more than one ticket, but my main interest has been health disparities. i am grateful for the invitation to speak to you today about my research and others. i am humbled by this opportunity, and for more than one decade, the national institute of aging, and many others, and specifically credit my colleagues at sloan- kettering and those at the center for health change. we have for together on these problems. the names of their peer are less

relevant. about one decade ago, my colleagues and i wondered if the higher mortality rates with cancer could be due to blacks receiving less affected treatment, specifically in the setting of cancer. we chose a simple procedure. this was for early stage lung cancer. the streets the no. 1 cancer killer, lung cancer, 25% of all cancer deaths, and it is enormously effective. in an analysis printed one decade ago, prior to the report that has been referenced many times, we show that in medicare, blacks with an incurable diagnosis receive surgery 30% less often than whites with the same diagnosis. this was not even due to differences in social and economic status. we also showed that we believe that this treatment that was the explanation for their poor survival outcomes and lung

cancer. the study is personally memorable for me. it was one of the first major analyses in the database which has become a cornerstone of cancer care, and it was also one of the few studies that demonstrate at that time that treatment gaps in terms of disease outcomes. that has been since shown in numerous other settings, but prior to that, treatment that 7 demonstrated. we were unable to determine in our study why treatment rates were lower for blacks. our study was not designed with that question in mind. the data we used was insufficient for these types of questions. we have used a national data covering many years and tens of thousands of patients, but we a little individual level information. other work with disparities is noticeably the opposite. sometimes, covering just a few patients and doctors in a single practice setting from which a lot can be learned about that setting but less about the universe of care. the fallout from the publication was educational for me.

a number of pundits, if he will come a step on top of our findings to use it as a platform to decry the health-care system as racist and, by extension, doctors as racist. "the new york times" carried on the front page in their weekend review a steady that was titled "not just another case of racism in health care." that is framed and in my bedroom. too many people concluded that doctors discriminated against their minority patients, and i noted a few years later in the essay of the "new england journal of medicine to speak that the invocation of racism as the cause of treatment disparities moves the problem, if you will, from one of the health-care system quality to moral failure. our research group saw the potential for another potential explanation that dr. chandra has mentioned. one blade the system rather than the doctor.

a key reason that blacks receive a lower quality of care than whites could be that it went to doctors who were less able to provide the higher quality of care, for whatever reason, than whites. it could be that they were less well-trained or simply less knowledgeable, so a few years later, we published another study in "the new england journal of medicine" and we documented two things that supported our care. first, we demonstrated that the first key precondition existed, blacks and whites were indeed not treated by the same doctors. we looked at medicare patients, and we were able to show that black cure was heavily clustered among a small group of doctors. it took only 20% of the doctors in the west to account for 80% of all care for blacks. whites were different. their care was mostly with other doctors. then we showed that doctors were different. we asked the question, what are the features of my doctor

compared to the typical medicare patient who is white? we found the doctors looked different. and black patience is less likely to of a doctor that is board certified in the prior specialty. we cut that was important, because the session in decades of research to be a key predictor in developing high- quality care. we also found in the primary care doctors to treated blacks if your resources, harder times making referrals for all of their patients. they had trouble selectively admitting patients for work and getting imaging test. more interesting, the financing was different. blacks went to doctors who more often provide free care and cared for medicaid patients, and the net effect was that they and lower revenues per patient and therefore less resources to support their practices and the other care givers in them. they were also more likely hurried, something more research had shown. around the time of this paper and in the years since, these findings have been reproduced numerous times, including by dr.

chandra and his colleagues. lower quality overall, more black patients, fewer whites. recently, colleagues and mine looked at practices that treat these and estimated that the estimated medicaid rates where sizable cashew trees to access problems and lead to shorter patient visits. my colleagues and i have some new unpublished findings that i can give you. we are finding that the predictors in getting care is this economic conditions, and how good the quality of care that your doctor gives his or her other patients, a meeting his or her white patients, we were unable to detect any consistent evidence that doctors are treating their black and white patients differently, per se.

i believe this is along with a plausible hypothesis premier because blacks are accessing up part of a system that is poorly functioning. little appears to be due to doctors singling out minorities for lower quality care. neither my colleagues nor i take the challenges posed by this lightly. in ways, the will be harder to ameliorate, but the payoff will be more durable and robust, so you have asked me here today to talk about health disparities and what our research suggests about our origins. it rests in the social context in which many people -- it lies at the heart of trina disparities. there is an explanation where we have a poorly distributed health care system. these are for the most needy individuals. a mechanism suggests that an approach for these high-risk areas will be the best way to improve care and out come for

patients. thank you again for inviting me. i look forward to your questions. >> ok, i would like to thank all of the participants. this has been a fantastique presentation. at this point, i would like to open up the floor for questions. do not jump in on it once, folks. commissioner? >> i think you all for being here today and for testifying, but i just wanted to ask questions. as far as data collection, i know that for the native american population, one of the issues we of always had, and there are tremendous problems with the statistics, especially in 1998 and 2000 as far as, you

know, moi the population of native americans, what it was in the census -- you know, with a population non-americans. -- with the population of native americans. >> one of the very, very key issues is this issue on data collection. aside from some of the challenges that are mentioned with the census bureau, one of the major challenges we face and the tools we have, many of the service but the department of and human services as well as state organizations and state public health agencies use to capture what is the health status, many times, we miss important subsegments of our population, such as the native american population, as well as