hospitalizations for duals were avoidable. in today's dollars that's about $8 billion. it is very poor care for patients obviously so we have a demonstration going on targeted at beneficiaries in nursing homes. so in closing, i couldn't be a bigger fan of trying to develop evidence-based models of care coordination and looking at prevention. an area that we have not been able to focus on are the preduals. so that the folks who are 45 to 64 before they're coming on to medicare there's a huge opportunity for us to do care coordination or medical homes or care management to improve their utilization and the prevalence and the intensity of the disease when they get on medicare. then the folks who are on medicare, there's a lot we could be doing to prevent their decline on to medicaid. and so far the financing and the incentives between the two programs haven't supported that work. i'm hopeful that once we get through dealing with the 9

million folks who are on the program we can really shift attention to the preduals and doing a much better job of managing i would say the progression of folks to dual eligibility status. so with that, i will close and just say thank you again for being part of the panel. >> thanks very much, melanie. we're going to turn now to joe newhouse. joe it's often asserted that the u.s. is an outlier in spending. how big of an outlier are we if we are? >> thanks, susan. this discussion somewhat reminds me of that movie. not many of you have seen it. its a classic -- it's a classic japanese movie and basically presents four different views of the same reality. and i think that's an example of what we're hearing here. i noticed in the green book that

you were handed when you walked in there's the usual slide which i'm not going to show you on page 7 of the u.s. spends a lot more than everybody else which susan alluded to. what's less well appreciated although ken mentioned it in passing and my first -- on my first slide is that -- how do i advance the slides? the -- what i have done here is look at annual growth rates per person in real health care spending over time. so this is almost 50 years. and the asterisks -- these are the g-7. by germany, italy and japan because there are strange things about the numbers. not just they're different. germany had reunification in this period.

most of the japanese bulge is in the 1960s when their economy was growing 11% a year. but when i look at the other four, the u.s. actually is the highest, but not by a lot. certainly not by anything like the levels. then i'd like to go on to say when we just participated in the medicare trustees 75-year review which you may think is a waste of time, but we were doing our civic duty and in projecting 75 years or even ten for that matter, it's really the growth rates that matter. it's the growth rates that are doing it to federal, state and personal budgets. i just did some calculation over the weekend of the kaiser data

on average family premium against median income. 2009 to 2010. that went from 22% to 28%. so the average family premium -- that's total cost, not out of pocket. so the -- a family premium is now -- now, on the average insurance policy, 28% of median income which i find is somewhat staggering numbers. so the inference i'd draw from that, the rate of cost growth is going to slow down. i can't tell you how, but it is. nothing gross to the sky as the saying in the financial markets and is true here. so the u.s. and other country -- the other countries are also not so different. at least the u.k., france and canada. here are the data by time. i just arbitrarily have broken things into the decades since the '40s. and what you see is you see some variation around that average.

but there's usually something happening in those decades. that depart a lot from the average. medicare, medicaid. the managed care introduction in the '90s. the recession in the last decade. but what's remarkable to me and the reason i put this up is to give you some perspective that this issue of cost growth is a common across countries, and "b" has been going on for a long time. so what i take from that is that while the u.s. certainly spends a lot more than everybody else, that must be something that's u.s. specific. but the growth issue must be something that's common to countries and to decades.

now i take ken's point that what things may have changed over time in terms of what's driving this, but i still think it's important to keep in mind that this is not -- whatever is happening here is not necessarily a result of things that are specific to the u.s. which we naturally tend to get wrapped up in. the second point i wanted to make which also has been alluded to, we have gotten something out of all of this growth in spending. you know, again, the usual line is we spend a lot more than everybody else and we trail in life expectancy. but if i look at the change in spending against the change in life expectancy never mind mo

more -- morbidity, like cataracts and hip replacements, a remarkable change since 1970. looking around the room, most of us are old enough to remember. you know, my students can't tell the difference between 1970 and 1870, but i can certainly do it. so what's quite remarkable is that life expectancy is like a lot of other things subject to diminishing returns. it gets harder and harder to get an increment and we grew seven years which is a major achievement, in my mind, as did other countries. so again, this is going on everywhere. now, everyone can say how much is attributable to medical care?

well, this is a busy slide as it's projected up there. this is a graph from a study of several trials. it's in your handout. and in -- epidemiologic studies, i'm sorry, about what accounted for the change in coronary heart disease. and the darker bar on the left is what these authors attributed to treatment. these are in various countries, not just the u.s. and the lighter bar is how much they attribute to what are called risk factors, but the risk factor part includes the better control of hypertension and better control of cholesterol. as well as the fall in smoking which are the three big things in the light part there. but the dark part is the higher

tech treatments. so it's the decline in coronary heart disease that's almost all of the decline, in the gain in life expectancy in these years. so we actually did get something for all of this. now, that said, i think the -- you know, what joe and ken have put forward about what to do makes a great deal of sense. and what melanie has said makes a great deal of sense. but i would leave you with the notion that this is a rather -- a very pervasive and long-standing issue about cost. something is going to make it slow down because it cannot continue at historical rates, but how that will happen i am not wise enough to know.

>> okay. great. thanks so much, joe. we're going to move now to jim fasules, and this is your chance to explain why all this spending on cardiovascular disease interventions has been entirely worth it. >> i hope so. i want to thank ed and i want to thank mary ellen because you're probably as much responsible for me being here as anybody else. i don't know if that's good until after the comments, okay, mary ellen. i guess we picked cardiovascular disease as the example or the model of chronic disease here. let me just say that i would also going to go into that, go into the data and take the opportunity to say what a disease-specific association can do that's actionable in what we have been talking about. right off the bat, one thing we haven't talked about, cardiovascular disease also has

an effect on the economy. a net loss to productivity of anywhere from 300 to $400 billion a year in lost productivity. also though, dr. newhouse's slides, the last decade we have had a 30% reduction in mortality from cardiovascular disease. probably from the treatment. i'm going to say in the data aspects we have to kind of be careful what we pick. what population we looked at because one slide looked at stenting versus the other case. we're not looking at doing stents compared to the acute cases. and also the stents, whether the drug or the bare metal, if you want to look at outcome as mortality, there's probably not a big difference. if you want to look at outcome as to whether you need another procedure, then there is a difference. so what i'm getting at is the data.

and getting the data also to the physicians is very important. and one of the things we have done over the last 35 years is to develop guidelines that look at the science and try to translate that science into what you should be doing. and more recently actually taking those guidelines and develop what's called appropriate use criteria and we use those in the five recommendations in the choosing wisely which has been mentioned. you can actually set up some continuous quality improvement aspects with the physician. and by giving the physician their data on what they're actually doing and what they see how they're doing against the norms and whether they're using the procedures, the investigations, the imaging appropriately and you can actually get them to improve their care. we found that just giving a dashboard of how the physician

ranks against their fellows in their practice in the community and the region and nationally as far as where they score for instance on nuclear studies for their appropriate use criteria will decrease their use about 15 to 20%. getting down into the range though we think should be about 8% of what we would label as inappropriate use. of course, you can't tell a physician not to do something because it's still -- i think it's been mentioned, medicine is still an art as much as it is a science. but practicing the best science that you can, getting the data to the ifphysicians so they can make a better decision. and with the appropriate use criteria and choosing wisely is the interaction with the patient. getting them to understand and it's been mentioned the shared decision making, where you're discussing ahead of time the options. take coronary disease and an elective situation, you could have cabbage or coronary bypass

or have a stent placed. and giving that a tool for the physician to look to see what's the best this patient fits in and then see what the patient wants as well and help him make that decision. so i think that's how we have been looking at how do we decrease the curve. now, i would say that actually we have discussed the cholesterol and management of coronary artery risk disease -- disease risk factors and mentioned california, we have looked at the data on both coronary artery bypass and the stenting for coronary artery disease and it's gone down all of the last five years in california. if you look at the cms data there's been a 10% reduction from last year -- or actually, the year before to last year. for both cabbages and stenting and actually reduction in imaging. i think that's actually speaks

towards the addressing of the risk factors. i'm glad everyone's looking at obesity and mary ellen knows that i worked on obesity when i was in practice in arkansas. it gets to be quite hard. we were on the commission that set up measuring the bmi on the kids at the school and you could imagine when you had to do it so no one else knew the bmi and sending it home, but we had a 33% prevalence of obesity in the school-aged kids. by doing the measurement, taking the vending machines out of school, et cetera, we at least levelled that. now, we didn't impact the 48% smoking prevalence or i guess tobacco use. it was as much dipping as smoking. one program i'll mention is the durable balloon. when a patient with acute mi,

until they're blowing up the vessel we know that if you do than 90 minutes you save heart muscle and you don't have as much congestive heart failure. we -- now, we went from about a 50%, 60% in the e.r. to 90% of all e.r.s hitting that and hospitals hitting that number. that probably saves two to three days of hospitalization. and also puts the patient back to work after three to five days as opposed to what we mentioned when eisenhower had his heart attack and was admitted to fitzsimmons army hospital, he sat in the hospital with a little morphine and a lot of prayer an they actually have immortalized his bed pan in the hospital. they have a little room for where he was at the hospital there. but where did the savings go?

the cardiologist didn't get paid any more. so here's a program that had savings and we haven't figured out whether the hospital got them through the drj or the insurer got them. whatever we do with the models we have to look at as joe said there has to be some incentives built in. if you're doing the extra stuff or paying for this data, and finally the other thing we have are the registries that we can track the outcomes. maybe when the electronic health record gets to where it should be, we can do it that way. but data -- disease specific outcomes over longitudal time and someone mentioned in the trials they tend to be like most of us up here middle aged, bald, white males. not all of us, but we are the -- obviously, the trial -- the

people in the trial. and we don't reflect the makeup of society, but with a registry you see actually what we're doing and how we're treating. so data, data specific, give the data to the physician so the physician can act on it. and then i would echo what they said, greater physician and patient interaction and shared decision making. thank you. >> great. thank you. so diane roland, at kaiser family foundation, you have done a lot of thinking about the high-cost patients and some of the challenges to delivering improving care and lowering the cost. so tell us about that thinking. >> thank you, susan. i think i can do it with this, right? well, i always come to these forums and we talk about the medicare program and we talk about private insurance. when we talk about health care

costs we don't always bring the medicaid program into perspective. so i wanted to remind us about why medicaid takes care of a lot of people with chronic illness and why it spends a lot of its resources on that population. and of course, this is our most classic slide showing that one in four beneficiaries are disabled and elderly, but they accounts for two-thirds of the spending. a large reason for that of course is that there is a lot of long-term care expenditure on behalf of those populations. but let's focus for one minute on that under 65 disability population for medicaid. and we like to talk these days a lot as melanie mentioned about the dual eligible population. but there are nine million people with disabilities on medicaid who are not elderly and those people who are not dual eligible. only two million are dual

eligibles. so one of the real challenges in the program and ken alluded to it is to get better coordinated care, better services to the disability population for which medicaid has whole responsibility and where it does have a range of services, care coordination, case management are part of the medicaid benefit package and could be much more effectively used on that population with disabilities. and this is a slide that of course we're used to seeing in every part of our health care system. there's only a few people who account for the majority of any spending. and in medicaid the top 5% really account for a substantial share of that spending and the few people there that are children and adults are relatively small, though they are high cost. really there is largely driven by the disability population. when we think about the disabled

on medicaid, we have to remember that they include a substantial share of the population with severe mental illness. and it's often when you look at who is a high cost beneficiary it's the combination of one or another of the disabling conditions combined with mental illness. so a diabetic i think can look -- ken looked at this as someone with diabetes being treated with a mental illness is much higher cost and much less likely to follow the treatment protocols than someone who is just a diabetic. so that is a real challenge and especially as ken even alluded to much of the medicaid care coordination through managed care carves out mental health and manages it totally separately and now much of the drug costs is also carved out in many places. so that you really don't have the integration for that disability population that could help. but we are focused increasing

increasingly -- increasingly today on the share of medicaid spending that goes for the dual eligible population. and as you see from this slide, there are about 15% of total medicaid enrollment are people who have both medicaid and medicare coverage, but -- and they account for over 38% of overall spending. a huge part of that is long-term care services. but i'd like to point out that really if you're trying to manage that low-income disabled population there's another 10% of the medicaid population that are disabled and a few nonmedicare eligible aged that account for another 28% spending. so i would say within the medicaid program we really need to look at how to really better manage and there's an ability in medicaid to do more in the preventive side.

more on preventive services and sending people to the treatment programs like the diabetes management programs there. so the opportunity within medicaid is broader. and finally though as we look at the people who are dual eligibles, they really have a substantial share of chronic illnesses. 55% of three or more. but when i look at that, i also think well, a lot of medicare beneficiaries have similar levels of functional impairment. i mean, it's different, but it's not so substantially different on chronic illness. it really involves really thinking about how you manage the cognitively or mentally impaired because that's where the biggest difference between the dual eligible population and other medicare beneficiaries occurs. so in conclusion, we're really now looking at -- and melanie has alluded to being able to put together medicare and medicaid spending. so in this analysis that we did

with the urban institute we looked at within the dual eligible population are there is only who are higher spenders and some who are lower? and we found interestingly enough that if you look at the top 10% of medicare spenders and the top 10% of medicaid spenders, they're actually somewhat different people, even though they're dual eligibles. only about 100,000 people fall into the category of being a high spender in both medicare and medicaid. and that's largely because as we need to look forward at different ways to look at these populations, the subgroup analysis which melanie alluded to was so critical because here you see that the top spenders in medicare are really top spenders because of their acute care utilization and those are services that medicare can and should be able to better control. whereas the top spenders that end up on medicaid are often in nursing facilities and are also

people who are using long term care services in the community. and they actually spend a little less on acute care because their big spending is on long-term care. in conclusion, as we look at and try to figure out how to better manage both technology which is not on the long-term side, so mountain west of the control would come in looking at the acute care services, need to also think about how to manage different types of chronically ill patients and how the two programs can either work together better or work better within each program for the chronically ill. i think that's where we will see better outcomes and lower costs if we can really zoom in on the disability populations and the differences by subgroup. thanks. >> thanks so much, diane. well, bruce chernof of scan foundation, you have been thinking about this same group that we have been discussing. those with rising prevalence of

chronic disease and often the aged and the costs of those. what else do we need to know about this population? >> well, thank you, susan. i really want to pick up on where diane left off. i think the last couple of slides are incredibly important in our thinking here. as we wait for the slides to come up, we're almost there, folks. that looks good. terrific. so as i begin my presentation i wanted to start in kind of a similar place which is that you have to really look at the folks who are getting care and what their needs are. and that really help us think about targeting and building better programs. i think building off this notion that the medicare and medicaid high spenders are different is a critically important thing for everybody to take away today. and my first slide sort of picks

up on that point which is those with chronic conditions, that's only half the battle. those with chronic conditions and functional limitations together really drive a lot more of the high cost medicare spending. and we'll come to why that is in a second. we spent a lot of time today talk about people in their diseases, being defined by your disease. well, the reality is that function is an incredibly important piece of this. being able to care for yourself in a medical context and in a personal context. and when you start to factor in this notion of function, function in the way you live your life, not the way you live your medical life. i want to pick up on something that joe said at the very beginning and turn it on its own head. patients are people in waiting so trying to get back to the place where most people even those people who are chronically ill, even those people who have cognitive impairments or serious

mental chronic mental illness, they're still people most of the time. they still spend most of their time out and about in the world and that world is not defined by hospital beds, doctor's offices or call triage centers. and so thinking about the role that function plays in disease, and the way that function plays a role in people's lives is incredibly important if we're going to look at how we build effective programs and target appropriately. on this chart, it goes on to say when you add in functional impairments, regardless of the number of chronic conditions, functional impairments is an important cost driver, regardless of the number of chronic conditions that they have. again, sort of echoing off of diane's point which is how do medicare and medicaid hold hands? now, i know why you're here, because it's -- it's this slide, you know, who are the duals, who are the preduals? and what this slide starts to

raise is this whole issue of the preduals, what does a good duals program look like. so when you look at individuals who have chronic illnesses and substantial limitation, half of them are duals, but roughly half are not. so these are folks who run the risk of spending down because their medical problems get them by the tail and becoming duals. i think in an environment where as we watch boomers start their retirement, we see folks with fewer financial resources than they might have had earlier on this becomes a critical problem and a cost driver over the next 10 or 15 years for us. so when we in this scan foundation start to talk with folks about what they really want, we decided that we would do our almost 20 focus groups across the country last year and then a series of polls and what we did was to choose t