obviously, that's part of technology. equipment is part of technology. not just the equipment that is in the hospital up against a wall, some big thing. but also the little things. the stethoscopes work a lot better these days to pick on the thing that i mentioned that marcus welby had. but also it's medical technique and i think that's part of technology, too. it's knowing how to do something, even if the basic materials are the same as they were 20 years ago. but you now know how to do it. that is apadvance in technology, and all of that adds to both the supply of services that are available to treat disease and diagnose disease and the demand for such diagnoses and treatment. one of the things you'll see in the literature, which i've never particularly found useful. i'm an old labor economist at

heart. but technology, you see these studies that try to parse out how much of growth is accountable by various kinds of factors, including technology. and technology is that one thing that can't be directly measured, although i argue that the other factors that people point to aren't that measurable either. but everybody admits that technology is not directly measurable. so technology is usually treated as a residual. so it's not technology, it's the i don't know factor. so if you see somebody say technology is responsible for 60% of cost growth over some period of time, maybe. maybe not. it's just not at all clear. and technology alone, as susan said, technology alone is not the culprit. it's everything. but in the end, if there wasn't

money to buy it and if people didn't want it and doctors weren't prepared to do it, then that piece of technology wouldn't be used. so it's more complicated than most studies are capable of dealing with. so when you have better technology, you generally have better care. not uniformly, but over the vast span of time, we see this to be the case. and my example is cataract surgery. there's some evidence that a crude form of cataract surgery, that was literally somebody putting a stick in your eye, was practiced some time in the babylonian era. but more concrete evidence, there's some evidence in the 16th century which is essentially sticking a fancy stick in your eye. that didn't work too well. probably wasn't used very often,

and of course, there wasn't anaesthesia at that time. so you had to really want to do it. in the '60s, hundreds of years later, vast improvement in inpatient operation. we had learned something about infection, so we were far more capable of dealing with infection. that was a technological improvement. but it was risky. this was the kind of thing that you used a sharp knife, probably sharper than you're likely to find on the streets of washington any evening. probably a little smaller. but it's the same basic principle and essentially because this was so risky, few patients ever got it. it was reserved for those patients who couldn't see out of that eye. often only had one operation. and the idea was to extract the lens and after the operation, which was highly ricky, they

sewed things up and the patient was held for at least two weeks with sandbags so they wouldn't move. very uncomfortable. not so many patients were willing to try it and those who were, were absolutely at the end of their ropes. that wasn't that long ago. so then we move to today, and i'm not sure when today started, probably some time in the last 15, 20 years, we have a much more sophisticated procedure. doctors are looking through microscopes to make sure they're cutting in exactly the right place. they're using more sophisticated procedure to take care of the lens. they replace the lens with something that makes you see better than you ever saw in your life. only the unfortunate few don't make it to medicare before this happens. but if you are a medicare patient, they pick you up, give

you lunch, zap out an eye. the fact is, better technology is generally consistent with more successful results and if you have something that works better, you generate greater demand and although the price may be lower, you generate more spending. okay, so overuse, underuse, misuse. these are the terms that everybody hears and a great example has to do with treating coronary disease. this is from a paper by john skipper. they classified using another study from somebody else, they classified different kinds of treatments according to their cost effectiveness and their cost.

you can see that the way they did the classification, and i'm sure anybody could have ample reasons to argue one way or the other on any specific intervention. but i think the overall sort of pattern here is interesting, that effective low-cost treatments were, according to them, accountable for more than half of the mortality decline, due to coronary disease between 1980 and 2000. of course, they didn't have the guts to say that anything was not cost effective, but less cost effective and probably there wasn't anything that wasn't effective in a sense, and the word "cost" i think is probably an issue here. you can see that, according to their categorization, the more aggressive treatments, stints, cabbages, cardiac rehab, they're much more expensive than aspirin, for example.

they account for maybe 19% of the mortality decline. you want to be a little careful about this kind of a display, because you have to ask yourself something that they didn't ask in the paper, which is, what was the condition of the patient? somebody who really needed a cabbage, you could shove a lot of aspirin in their mouth on their way to the morgue. so it's not at all clear -- this is in fact a kind of a residual study. it's not very reliable either. but it does say something about our use of services and it does imply something about the economic incentives associated with the complicated things. aspirin, who makes money off of that? not even the drug companies. cabbages, who makes money off of that? you know whoer.

but would you go back to marcus welby's day? where they knew about infection but couldn't do much about it. of course not. the fact is, we want all of those things. and i'm still sore at rick, because he hasn't given me the drug that will help me. but i'll give you until 2:00. but can we find out about evidence? here's a study from elliott fisher, looking at regional variations and medical spending. this is in the -- this is an index -- these are medicare patients at the end of their life. i didn't read the article close enough to know how close to the end of their life they were, but pretty close. and look at the distribution of tests and procedures that were done on people close to the end of their life.

and low and behold, what do you see? very few minor procedures, but a lot of tests and imaging. that's where the money is, for the very sick people and that's where the money is for the not very sick people. there's plenty of money in those fancy machines, but where is the real throughput? it's in the seemingly ordinary interventions that we're all used to and expect. that's where the money is. okay. well, everybody says let's do some research and figure out what we should be doing and don't do the things we shouldn't be doing. the only problem is that there are a lot of things that we do. and there are very few studies looking at what we should be doing and i don't care whether you look at the stimulus funding and you look at the billions of

dollars going into other places. the research can't move fast enough, you can't spend enough, you'll never get ahead of it, because things that we accept for granted, we do without question. and occasionally when we question it, such as the on and off again discussion about the blood tests for prostate disease, we get a lot of resistance, because that's not the way we do it. it's the part that's already heavily regulated that tends to be drugs and diseases first, everything else last. and that doesn't strike me as being the exactly intelligent way to allocate resources.

even worse, this wasn't supposed to be a diatribe against the effectiveness of research, but i've always been skeptical. these are highly retyped studies, so it isn't at all clear that they necessarily tell you what would really happen to the average patient in the average setting and that's a problem, because that's what you want to know the answer to. is it going to work most of the time or is it not going to work? because things that work under ideal conditions might not work under normal conditions. and let's not forget about the patient, because the patient might not be adherent, either. and professional judgment changes all the time. why is that? partly because there are changes in the way you do things. partly it's our experience grows. we see more patients. and our viewsn't what to do

changes all the time. so effectiveness of research is interesting. it's going to make a lot of people a lot of money. i don't think it's going to have substantial impact on how we spend it. can we spend our money better? i think there are things we can do. part of the problem is that hardy anybody in this country pays for what they get. yes, they do pay for it, of course, 100% for it but they don't know it. when they go to their doctor, the doctor can't say how much it's going to cost. and this brings us back to p and q. but that's where we have to focus our attention. if you don't know the price, you don't know much of anything. and as i said previously, knowing clinical effectiveness doesn't get you halfway to knowing that's something i want. because what you want to know is value. so there are lots of things we

can do. the medicare program has tried lots of things but they haven't been very successful because it is not a health program, it is a political program and political programs cannot make decisions. they try, but they can't. i was tied the up with the centers of excellence project. it worked perfectly and it was shelved. so what else could you do? well, conservatives talk an awful lot about financial incentives, and i think this is a case where we need to apply financial incentives to the whole system. so i won't get into it, but premium support could take us a

long, long way. private insurance i think is the more likely place where you're going to see action along these lines. along any of these lines, because although they do ultimately report to congress, they don't report directly to congress. so there's a possibility of some progress in making hard decisions and try to make them stick. i've got to say i haven't seen much evidence of that, but as economic conditions tighten, as business conditions tighten, as the resistance of employers to higher premiums, necessitating to keep premiums not so high, necessitating higher co-payments and deductibles, we're going see that turn around and we are seeing that turn around with some insurers. one theory that shandra advances is that why don't we attach differential co-payments to measures of effectiveness. that sounds like a great idea

until you ask yourself, can you really trust those measures of effectiveness? my answer was, no, not really. but sure, the idea of value based insurance design is a sound one. providers. professional consensus drives a lot of things, so it was great to see this group of specialty societies put out under the banner choosing wisely. a list of things that they strongly urge nar practitioners to reconsider whether they really need to do those things. new business structures that provide real financial incentives. and what about consumers? in the end, what is a consumer? it's a patient. i'm a consumer.

if i could find the right thing, i would buy it and pay for it with my own money. so what i would like to know is not just what is it going to cost me, a question that can barely be answered today. but also how is it likely to affect me? and that's really hard to get answan answer to. and finally, what about expectations? i left that here for consumers. in tend, if we don't change our views about what we demand as an absolute minimum, and rick, i want to assure you, until we get realistic about those things, we will not get control over costs. thank you. >> thank you very much, joe. so as joe said, technology adds to supply and demand and one of the things we have an unending

supply are patients with chronic disease. so ken, over to you to talk about that. >> first of all, thanks, ed, and others for inviting me. a pleasure to be here on this panel. great to see everybody that you often don't see all the time. so welcome back to work. i'm having a tough time making the transition myself. so i'm delaying this a little bit. i'm going to talk a little bit about another angle of this, but it's not unrelated to what joe has stalked about as you'll see in a minute. it always fascinated me that in health care, some of the most fundamental big questions get the least amount of attention and study. about six or seven years ago, i looked at some of the literature on what do we know from the published data that's driving the growth in health care spending? one of the last few pieces that i saw was a piece that joe did.

i guess it was in '92, '93. looking at the time period between 1940 and 1990. and if you think about it, that was a very time period than the most recent experience that we've had. the number of uninsured over that time went from 90% to 15%. so clearly the amount of induced spending as our whole system changed over time was roughly related to changes in the structure of insurance and demand and as joe just talked about, innovation. we brought medicare and medicaid on. we had a whole bunch of really important new innovations that fundamentally changed how we treat patients. neonatal intensive care units and treatment of low birth weight babies, treatment of cardiovascular disease. so it's a very different time

period. so what i'm going to spend my discussion on is looking at the time period between the early 1990s and today and i think the point i'll make is even during that time period, the year-to-year changes and what's driving the growth in spending is somewhat different. but i want to look at some of the long-term drivers here that are more recent. i've sort of taken -- you can decompose this in a lot of ways. i have tried to sort the data two three ways. looking at the change in spending linked to the change in prevalence of treated disease. looking at the change in spending linked to how much we spend to treat a case. and obviously the interactions between the two of those. and if you look at a long-term component here, say late '80s to today, about 60% of the growth is linked to rising prevalence of treated disease.

we'll go into what accounts for that. some of that is going to be good. some of it is going to be bad that we can potentially go in and do something about. so to give you a sense of the magnitude of the changes here, and you can go down by medical condition and see the prevalence increases, and for each of these conditions, the factors driving the growth are somewhat different. so if you look at the treatment of cholesterol, mental disorders, those have increased very dramatically. much of that is technology related. we have new approaches for treating patients with cholesterol. we have new medical innovations to treat people with mental disorders that we didn't have 30, 40 years ago. diabetes, i'll come back and talk about that. that's almost all an incidents

increase. largely more cases of patients who are diabetics. this is a spectacular problem in the medicare program. it's one of the key drivers of rising cost in medicare. so you can go down the list and see there are e nor mounormous s in the treatment of disease. so if you ask what is driving this growth? some of it is going to be things that we should be happy about. some are going to be things we shouldn't be happy about. so the first one, as i mentioned diabetes, our detection rates haven't changed much in 20, 30 years. we're detecting 72% from total cases of diabetes. that's gone up from the upper 60s, but we're not doing a whole

lot better. so the treated prevalence numbers are incidents increase, not detection increases. a second one is obviously somewhat debatable and controversial. but there's no question that over time, we change the definition of disease. so we've had a change in clinical thresholds for treating different kinds of conditions. i think most of the studies that i've seen that have looked at that, think that's a good thing, that a more aggressive treatment of cardiovascular risk factors has been a leading cause of declining rates of cardiovascular mortality over the last 20 years. new medical technologies, obviously treatment of mental disorders, and joe went through that. provide more tools for us to treat patients that we didn't have 30 or 40 years ago. we're living longer.

so you're going to see some increases in disease prevalence linked to longevity. and changing definitions of disease. if you look at the definition of something as simple as diabetes, that's changed over time. one of the things that's very different in this time period that we're looking at is increases in obesity. if you look at the 1960s, '70s, '80s, that whole time period the share of adults considered obese was about 17%. department change for about 30 years. so clearly not a contributor over the time period that joe was looking at to recently. if you look at the long-term

trends here, it's doubled since the mid '80s. and if you look at some of the calculations that are just linking -- or looking at how much of the growth in spending is due solely to obesity, holding treatment intensity and technology constant, depending on the time period, it accounts for about 7% to 10% in the rise in spending. so it's -- of all the things that we can quantify to joe's point, it is an important contributor. if you look at medicare spending, about a third of the growth in medicare spending over the last decade is linked to sort of cardiovascular risk conditions that are in part lifestyle related. diabetes, hypertension,mental disorders, et cetera.

the things that is interesting about those conditions in the medicare program, those are largely conditions that are ambulatory treated with appropriate medication. unless you botch it up, with the exception of kidney disease, you have nothing to do with in patient hospitals here. and the ultimate irony of that is that traditional fee for service medicare is the only major payer, unless you're home bound, that has no care coordination. we've done some tabulations that kind of try to sort these out in terms of spending increases over time. and as i mentioned, the share of

spending increase linked to disease prevalence versus cost per case does differ over the time period that you're looking at. the most recent couple of years, we're in an economic slowdown, so we have slower rates of utilization. over the last couple of years, spending per treated case, and more important explanation of rising health care and spending, but if you look at these long-term trends, it is disease prevalence increases that are driving it. if you try to drill down a little bit more, if we can look at treatment intensity, meaning how much are we spending to treat a particular case of diabetes or heart disease, over time, and how much of it is due to increases in obesity, again, holding technology constant, as i mentioned about 7% to 9% is due to obesity alone. and if you look at the 1987 to

2001 time period, about a quarter of the growth in spending is due to increases in the intensity of treatment of how we're engaging and working with patients. some of that is due to changes in technology, clinical recommendations for how aggressive we should be treating patients. they're bottled up into the same bucket here. but treatment intensity is a major component of this, as well. if you look at both obesity and treatment intensity, anywhere from 20% to 30% of the growth is linked to both of those two combined. so what are some of the challenges here that we face? we know that any given year that obese adults spend about 40% more in health care and depending on -- there's a whole range of different estimates of how much is in the base of spending linked to obesity and i think the last piece i saw was

around 20%, 21%, is due to obesity alone in the base of health care spending. obviously, we spend a lot of time focusing on the health care piece. but if you look at the total cost of chronic health care conditions, for every buck that we spend on the medical side, we're losing about $4 in productivity. so there's a bigger component to this that's really important. let me go back to my medicare challenge here. the medicare program, if you look at age 65 and look at lifetime health care expenditures starting at age 65, and if you compare the lifetime spending of an obese adult to a normal weight adult, and there's a couple of good models that have done these estimates. anywhere from 20% to 40% more spending over the course of a lifetime for an obese adult. so the point is, this is a very

different story than smoking. smoking is a mortality discussion, this is a morbidity discussion. so it seems to me if you think about medicare, there's two opportunities that i think that are important. one is to find ways to change the incoming health profile of people coming into the program. and let's just go back to the coordination piece. i'm sure melanie will talk about, this and they're doing a great job of this, but if you think about the opportunities here to put in place evidence based care coordination enter negligenc