-- interventions, over the next decade, medicare is going to spend billions. so why don't we build something into the program that is effective in doing this? and the frustration is that we have interventions that are effective. so we have a program that's -- eric holman has developed some time ago out of the university of colorado in denver that has several trials that shows we can cut readmission rates by 50%. that should just be a major component built in to fee for service medicare. so i think that there's opportunities here, but we've got to focus on two of the problems that we can do something about. one is preventing and averting disease in the first place, and

would love to have a discussion about some of those opportunities through things like the diabetes prevention program. and second is to really build into programs like medicare evidence-based components of care coordination that we basically have decades worth of randomized trials to show that they work. things like transitional care. medication therapy management. health coaching. having an integrated care coordination model into the primary care practice. so we know the elements that are effective. i think we just need to find ways to integrate them and build them into medicare. so with that, i'll keep this short and look forward to the discussion. >> great. thanks to both of you. so we now have some time for clarifying questions if people want to ask questions specifically of ken or joe to draw them out on points they made. we don't want to get into a lot

of deep analysis of what they said at this point or a debate. but just clarifying questions. if you do have a question, please introduce yourself by name and affiliation and switch on your mike. if i look at your slide that was labeled high cost of advancing technology, where you label x-ray machines, et cetera, and i look at current technology, i see two items mentioned there, surgery robot and treated stint. both of which recent studies have shown do not materially produce better outcomes for patients. and in fact, with respect to the treated stints, i think johnson&son's unit stopped making them, because they were not just effective, but problematic for patients. that underscores that we have a

lot of technology out there that either does not produce better outcomes but costs a lot more. you in fact can be harmful. of course, more broadly the institute of medicine tells us half of all the medical interventions we engage in, there's no evidence they work at all. so i was curious as to why you didn't put even more emphasis on some aspects of technology assessment as being potentially more actionable solution. >> well, since everybody believes that, i wanted to make the somewhat counterpoint that what we're going to focus on is the treated stint. people want to believe that somehow putting a machine between the surgeon and their body is necessarily an

improvement. they would really like an improvement because they think they could be seriously harmed or killed. but the fact is that we do have a tendency, especially on stints, for example, we do have a tendency to look at those. for one reason, it's easier to examine the effect of a stint, because it is a purpose. it doesn't have multiple purposes. it is less dependent on the still of the physician to place it, for example. a singular product is more immunable to testing. we have a tendency to examine the things that are easy to examine, and not the things that are really hard to figure out. and the hardest things to figure out are the things that are standard practice.

so yeah, there's some hope for technology attestment. there isn't much hope going from good technology assessment to sound medicare policy, but there could be plenty of hope going from good technology assessment to good professional standards. >> ken, quick question for you. you have used the phrase "treated prevalence" here a lot, suggesting you're distinguishing between just prevalence and treated prevalence. can you desegregate those two pieces? we're giving statins to a lot of people for high cholesterol and there's a lot of debate whether that is even the correct set of interventions. so how much of this is treatment independent of actual prevalence

versus treated prevalence? >> boy, now i'm confused. it's not even happy hour. >> are you sensitive about these -- >> the phrase for me "treated prevalence" is to distinguish the fact that we are only engaging a fraction of patients that have different conditions. if you go back to the diabetes example, 28% of people that live with diabetes have not been diagnosed and don't have a medical intervention. at some point, hopefully they will. but at any point in time they're not. so that's the distinction. the other part of it is, you know, is an important issue. i tried to distinguish in this discussion that there are

components of prevalence that we could intervene and do something about, and we want to. so issues around obesity and lifestyle and diet and exercise, things that we can reserve the curve on. there are other components of this, where if you go back to my slide and look at the treatment of cholesterol and hypertension that are a medical call that says, you know, if we are more aggressive at treating those diseases, plus, we have the new technologies to do it, that it does produce better value, that we're producing cardiovascular mortality, improving the quality of life, so on. those are also part of the discussion here is that we have changed and made a medical

decisions and treatment decisions to say if we're more aggressive at treatling certain types of cardiovascular risk factors that they do pay off. >> yes, brad? >> brad stuart, northern california, where we're building systems of care coordination for centers. my question to you, i'm a primary care doc for a third of a century and research, as well. all of our discussion, and i'm coming from a provider place, is focused on the providers. what about preference of patients, particularly seniors in this near end of life population, where our data is showing they would prefer not to be patients, they would like to be comfortable and stable and safe at home. i think we have systems to begin to do that. but my question around the data

is, we have a lot of studies now on effectiveness -- in other words, is the treatment necessary and valuable? what kind of data do we have on preference, where we begin to know whether these treatments are actually wanted or unwanted, which to me is much less controversial than trying to decide what's necessary. you often can't know what's necessary until after you do it and it hasn't worked. it's not controversial to know that people really don't want this stuff. and as it turns out, many of them don't. >> that's a great question, and as i think about these different models of primary care and care coordination, and to your point about giving people options and decisions about the type of care, how aggressive they want care to be towards the end of their life, lord knows i don't want to bring up death panels.

but that's a legitimate discussion that needs to be built into medicare. it needs to be part and parcel how families and patients and health care providers talk about options. you know, i think that's a classic example of working with patients to give them options and give them information and have the time to actually from a physician's stand point, to talk about that, is important. and we're -- i've seen more and more -- interesting care models come into place. hospice, you know, again is another important component of that. but having the time to have that discussion in a fee-for-service system is a real problem. it's just not built into the coding, it's not built into the time the amount of physicians get to counsel patients on important decisions like that.

and it should built in to how we think about doing care coordination and primary care with patients to give them options. and then have the options out there available. just to give you another example, in you think about to me sort of the incompleteness of the medicare wellness benefit, we built in a welcome to medicare physical. we were going to do a health risk appraisal. we're going to give you a personalized plan, but we don't do anything about it. but if you think about it, there are programs out there like the diabetes prevention program that have been put in place, that we've shown in randomized trials that they generate a 5% to 7% weight loss. well, that should be an option, built into the medicare program that would give people a choice of, if i really want to make a difference in terms of changing life tile or improving my blood

sugar levels, that should be a component of what medicare covers. so on both extremes we don't give people a lot of options because of the way medicare works. >> rick smith? >> ken, you have referenced several initiatives around the cluster of chronic conditions that you identified as significant cost drivers. can you speak a little bit to what happens to utilization and outcomes as these interventions take place and how does care change and what does that ultimately add up to? >> yeah, that's a good question. you can look at the prevalence data, and the more question is, what are we getting from it?

is it worth it? i referenced david's work on this and others, that on balance, the more aggressive the treatment is worth it. we're getting improvements in longevity and the quality of life. that those are investments that not only are clinically driven but generating better outcomes. then there's a series of prevalence increases that are bad, that we should try to do something to reduce. so there are different issues and how we think about them i think are very different.

>> tom miller, and i saw a hand up there. we'll come back to mary ellen. >> ken, we've gotten quite good and clever about coming up with new names to call chronic conditions. got a code, we can find the technology and a bill to throw at it. you had a list of the ones that have had greater treated prevalence. when you do your time series, what have we had any reduction in? i know we made great savings on smallpox, i'm sure. but beyond that, is it only an athded key to the keyboard? >> that's a good question. we've broken these into -- i guess we have 260 that we've looked at. i would say most of them are fairly constant. i mean, obviously the big ones like heart disease and cancer we're getting improvements in. the one that has been the

biggest decline, which actually adds to, you know, adds to a lot of the cost is trauma. so the prevalence of trauma cases has gone down fairly substantially. so that's a big reduction. but most of these have seen fairly substantial increases over time, and a lot of it, as i mentioned, has been obesity related and a lot of them are related to these cardiovascular risk factors. i think kind of the interesting thing is that if you look at the spending growth in the united states, going back to the '40s or '50s, not that it's a whole lot different, 2% above gdp, but the factors generating those inkress over time have changed and obviously you're not out of whack internationally either. but we have seen some differences in what's generating that delta. if you look at a medicare

patient in 1965 versus today, they're very different. the typical patients driving spending in medicare today is an overeight 70-year-old with high cholesterol and depressed. those are all conditions that require nothing that medicare does. >> let me add something, though. this is a pitch for technology. one of the reasons we have more treated prevalence on nearly everything, it's easier to treat. and there's the push for so-called convention, which means earlier diagnosis. so it's hard to know where all these fit in. but the march of medical

progress has contributed considerably. >> so mary ellen? >> should we be thinking more about targeting hot spots or targeting industries? a lot of what we generally talk about is broad policy changes in medicare and other areas. but certainly the obesity and other risk factors seem to be located in certain parts of the u.s. i would argue. should we start to focus on those areas and maybe not have across the board sort of improvements but in order to get -- given limited resources, to think about that a little bit more? >> that's a good question. i'll give you my pitch on the diabetes prevention program. we have a population of 80 million people nationally that are prediabetic. they have a program that we know through ten years of clinical trial followups that has a cumulative reduction in the

incidents of diabetes of 34%. so it works in a point of time and over time, we can reduce it. united health group and the ymcas have put these in place in 25 states. we can reduce weight by 5% to 7%. we can reduce the insurance dent of diabetes by 58%. and for older populations, 71%. that program could be scaled nationally in the next 12 to 18 months for $80 million. why in the world don't we do that out of the public health fund? that's just something that is an investment that we should do. we should build that one simple program nationally. have it available to small employers can use it. medicare patients could be referred to it. and exchanges, plans in the

exchange could prefer patients to this. that's a simple example of something we should be doing because we know it works. i guess to my point, we have a whole variety of interventions that would target these problems i talked about. we've had years of data to show that they work. transitional care models. diabetes prevention program. we need to flip the switch here and get into implementation mode. we're not going to pilot project ourselves into a solution here. i mean, we need to sort of take things we know that work, target them to at-risk populations, and we can make an enormous difference. >> that was my point, about trying to get those services to the people at the highest risk in certain parts of the country. >> i live in the obesity triangle. so if you just take the cdc data

on obesity rates over time, looking at changes and diabetes prevalence, they're do right no would make a difference. you know, we just need to flip the switch and focus on implementation and you can tell what i'm saying is i have complete pilot fatigue. sure, we can need more information, we need to pilot some other different projects but we have so much data on programs that we already know that work that we should just -- you know, we just implement and build into how we do businesses in the exchanges. if you think on the exchange side, something we're not talking about is that in the definition of the central health benefits, we have inpatient, outpatient, the usual services but also a component of certification for plans to be in exchanges, prevention and care

coordination. well, what do we mean by that? what are we certifying and asking plans to do on the care coordination? there are simple things that would make a lot of sense, like transitional care models and evidence-based lifestyle programs like the dpp. >> all right. dan callahan and then move on to the reactor panel. >> i'd like to raise one issue that has not been touched on. that is how do -- i'll take ken's example of the 70-year-old with all of the things wrong. we talk about coordinating care. the question is how do you assess care with multiorgan failure or multidiseases at the same time? we are good at it doing it with individual cases, but if you get a team of if i sphysicians toge how do they assess the overall work and their interaction? >> i'll put my md hat on for a minute. ke

ken, i think this is a great example of having team-based care and if you look at some of the health systems that do a pretty good job of that, whether it's him, marsh field clinic or guisinger, where you're building teams to deal with multiple problems and you're really treating patients as holistically. it's probably the best way to go with this. if you think about, you know, let's take medicare, good example. even medicaid when we do care coordination, a lot of the care coordination still segments off care coordination into different buckets. you'll have behavioral care contracted out, acute care. you know, dealing with the patient that has all of those problems. so even coordinated care sometimes and medicaid is not coordinated at all. it's really still fractured. so to the extent, you know, that you can continue to drive this

towards payment reforms that move us toward team-based care, that really engage patients for the whole range of medical problems is i think our best bet. it's not certainly fee for service medicare. that's not how that program works at all. >> thank you. we're now going to move on to a series of a short presentations from our reactor panel. these are nonnuclear reactors, but they are nonetheless very energized. and we'll get through their presentations i know energetically so we have time for a break and then a beefy amount of time for discussion. we'll start with melanie bella who is leading those efforts at care coordination for the dual eligible population. melanie? >> good morning. thank you. ed and others for inviting me to be here today. i was trying to figure out why i was invite and then it hit me, i

think dual eligibles are poster children for those who use the high cost. i think we'll get into the prevalence of chronic disease, but the fragmentation between the two programs just exacerbates the use of technology and the high costs that are driving the system. so i just want to spend a couple of minutes kind of on that theme what is actionable and talk about a couple of actionable things we are trying to do at cms to try a get handle of the opportunity to improve quality and cost for this population. and the first is all about data. until we understand better this population, the subsets of the population, what's driving their care needs, what the utilization patterns are, by very discrete sub populations we won't be as effective as we can be in developing new models to improve transition and to improve the

use of long-term care services. and so there's a few things we're doing in that regard. one is i'm pleased to say that cms now has an integrated data set, so medicaid and medicare at the person level, it's not 2011 unfortunately. but we're getting there. over time. it's really going to help drive our analysis as well as those of others in the room and other interested stakeholders. the second is we'll soon be releasing state profiles so it will be a state by state look at the demographics, the utilization, the costs again for -- from an integrated data set perspective of the individuals who are duly eligible in the state. not meant to compare state to state because we're not controlling for the differences in the medicaid program. but again it will be a useful tool to get out there. another thing we're doing that i think is very relevant to this discussion is looking at the simplest way to call it is a pathway analysis. so it's very different if you

start on medicaid and age into medicare versus if you start on medicare if you have a functional decline that makes you medicaid eligible. and the models are whether should be would be -- whose care is improved by care transitions. all of those things are highly dependent on what drove a person to be in the program, what their care needs are. who they trust to get information about those care needs, all very different. depending on which "m" you start with. so we're doing a lot of work in that area. then lastly i would say -- those of you who are researchers will appreciate, we're making enhancements to the chronic warehouse, so we have conditions for diagnoses for serious mental illness, for alcohol use, for intellectual and developmental disabilities. if we again are going to truly understand the prevalence of chronic disease for this population by sub populations, we have to continue to add to those -- to the diagnoses that are in the ccw, particularly those that are going to reflect things that medicare maybe

hasn't looked at so much as medicaid has in the past. so that's a critical part of the effort. as part of our efforts to work with states in this arena, we have been focused on making sure states have access to medicare data. for care coordination purposes. so we have tried to streamline a process that while abiding by all privacy and confidentiality rules allows states access to these data. we have 22 states that have either received or in the process of receiving parts "a" and "b" and 20 states who have received or in the process of receiving part "d." it's critical that states are requesting the data to show us that they have an understanding of their population. and they can tailor their interventions and their demonstrations to the needs of those populations and they're very heterogenous career pathways. so the next thing i would mention quickly, again, along the theme of actionable are

demonstrations. and i appreciate ken's comments about -- i think pilot fatigue. we geerting all sorts of feedback on the demonstrations. one is they're too big, we're moving too fast and we have others telling us, boy, it's about time, can't you go a little quicker? so i would say certainly i understand the pilot fatigue, i think for this population we have not yet tested -- there is no evidence base for truly integrated, improved coordinated care, particularly that bridges the behavioral health, the long-term health and the acute and primary. however, there are the nailer and komen models and we expect to see those things where they're relevant for those populations and the demonstrations. we have a state-based demonstration opportunity right now and that involves two models. one is a cap thaitive model. we have 26 states interested in pursuing one or more models at this point. some states are targeting a 2013

implementation date and others a 2014. we expect to see care models and care teams and care plans that are tailored to the different needs of the populations and we have not done it as good a job in the past as we need to. the needs is of someone the prototypical 80-year-old patient is very different than those who are under 65 or someone who is an institution. understanding all the varieties is a great opportunity for us to test in this -- in these new situations. and then i'm very excited the other demonstration that we're doing is for dual eligible beneficiaries who are in nursing facilities. there are about a million of them at any given time. potentially avoidable hospitalizations is a critical opportunity for improvement here. the churn between hospital and nursing home largely driven by the misaligned financing of the two programs is it's actionable, improvement in both quality and cost. cms