Charles E. Phelps: How we and our government created, exacerbated, and extended the health care mess

November 10, 2011 1:31:05
Kaltura Video

Professor Phelps discusses the current healthcare situation, the Affordable Care Act and ways to change public policy to improve future outcomes and costs. November, 2011.


>> Good afternoon.  Welcome to the Ford School and to this lecture by my old friend Chuck Phelps whom I will introduce in the moment.  I want to acknowledge the support of the Gilbert Omenn and Martha Darling Health Policy Fund and of the organization of the Health Policy Student Association.  Some of its members are probably here.  My name is Paul Courant I'm a professor in Ford School, as well as, a part-time university librarian.  And it's really my pleasure to introduce, introduce Charles Phelps who is I found in a variety of contents.  We were provost at the same time.  There's a bunch of secret hand shakes we both have.  He has been a mathematically rounded economist with interesting policy for a long time.  That's a fairly small group people although there are several of them have a [inaudible] full faculty as well as the University of Rochester faculty and I know of others.  I first knew him as one of the founding leaders of an organization from the Association of Public Policy and Management which is still active and important and he was the secretary of that group.  And he was the first person I knew who carried around a portable keyboard that wasn't attached to a typewriter.  It didn't--it wasn't much of a computer right now if I figure out exactly what it was.  But he was there producing ASCII files long before people knew what ASCII files were as secretary and [inaudible] Association for Public Policy Analysis and Management.  Has been an extremely important entity in the development of analytical policy, policy analysis, and management.  And Chuck was one of the leaders in that area.  He started an enormous amount of work in the--health economics and related areas, health policy, medical decision analysis, cost-effectiveness to various interventions.  He has written a textbook on the subject which is in its 4th division.  I can now tell you from a librarian perspective when there's textbook in the 4th edition, somebody got to send a bunch of them in college and then some.  And he's been as I said a leader both in health economics, in health policy, in the organization of public policy and universities in the leadership of his own university, at the University of Rochester whence he's now retired.  And without further ado I wanna ask Chuck to come up.  Our own worst enemies: How we and our government created, extended and exacerbated the health care mess.  Chuck will at appropriate times he'll be in charge of his own show.
>> Alright.
>> That we a lot an opportunity for interaction.
>> Thank you.  I'm pleased to be here.
[ Applause ]
>> Except Paul and I have known each other in number of circles but in provost, in the world of libraries.  And of course provost emeritus, in my title emeritus means I got out of town before they run me out of town after I stopped being provost.  Provost in case you don't know the origin is from the Latin word for jailer.
[ Laughter ]
>> So I'll just leave you discern your own conclusions about that for my good friend Paul and myself.  What I want do today is have a conversation with you.  I know there's some time for Q and A that's sort of some time for talk and some time for Q and A and I'd rather mix them up.  So if you have a burning question, I'll try and see if we can deal with it on the fly.  And if I think if I'm gonna come to the point of answering it little later on, I'll ask to defer.  And then if I don't get around to it by the time we actually come off the real time for questions you can come back and ask me again.  Well, that's my rules for the engagement and I guess I have to enforce them.  So the first thing I'd say is about shameless commotion much of the material that I'm talking about today comes from a book.  It was published last year by the Hoover Institution Press out of Stanford.  8 questions you should ask about our health care system even if the answers make you sick.  And so if you like what you're hearing today, you can find some more of it in this book.  And if you don't like it you can get the detail and refute it if you wish.  So let me give a quick overview before I'd like to take this today.  There's been for, if you go back into it at least to the days of Franklin Delano Roosevelt's presidency before World War 2, long standing discussion in the United States about whether or not we should have universal insurance, health insurance coverage and if so what should it look like?  And that issue arose during at least during the--prominently during the administrations of Roosevelt, Truman, Nixon, Carter, Clinton and finally Obama and only the last of those actually resulted in legislation.  One can save Medicare during Lyndon Johnson administration for the elderly and Medicaid for the low income populations but that was not universal so that's something that would be kind of half way step.  But all of these discussions in some sense are sitting there with some sense of disquiet about how our health care system is working.  And disquiet has been increasing through time.  I'm gonna show you some data on this in a minute but I would say the disquiet centers primarily on the following issues.  First is that we are spending a lot of money on health care in this country either in dollars or as a percent of our gross domestic product per capita income if you will more than any other country by far, and we're doing so in an increasing rate and that alarmed some people and say, look to the future.  The second problem is that if you compare at least on things that we can measure easily across countries, we're not getting very much bang for the buck for our medical spending.  And the two most prominent of these, the easiest to measure, maybe we're looking for the keys under lamppost.  But if you look at life expectancy and infant mortality which are every easy to measure and compare across nations, we're not getting nearly as much for the buck as other nations even though we're spending a lot more than other nations.  And then as we look to the future, I think it's almost certain that the money we're spending on health care is going to rise through time.  And that's gonna come through 3--3 prominent reasons, all of which are moving almost glacially and hence unstoppably.  The most important of these is the population is aging.  I'll show you some picture of this in a minute.  If you're a young it should scare you a lot.  If you're old you can say, thank you to the young for helping with this problem.  The second is that despite the current economic trials of our country and the rest of the world, over time our economy will grow and health care grows on average at a faster rate in the economy in general.  The income elasticity is about 1 and half to 2, if you wanna throw a few numbers out when you look across nations so growing income will add to the spending.  This might be a really good idea.  By the way, I'm not saying it's necessarily bad for those who worry about it.  And then technology is going add new technology comes in.  New technologies don't save money.  They improve our well-being but increasingly in ways that cost more money than their low-hanging fruit that we had in our health care system earlier.  So as new technologies come and our insurance systems decide to cover them, we're going to further spend more money and you couple those 3 things and we're going to spend increasingly large amounts of money.  We spend about 16 percent of our gross domestic product on health care right now and I can easily see before I'm incapable of assessing numbers anymore that could rise to 20 and perhaps 25 percent within the rest of my lifetime, hoping I live long enough to see that but hoping it doesn't happen necessarily.  And finally I will show you some numbers that lead me to conclude that about what we're spending, we're wasting a lot of money.  And we're wasting in ways that are pretty easy to demonstrate at least on broad geographic terms.  There are huge differences across regions and how much money we spend on health care and when we look carefully, no measurable improvement in health outcomes across the wide variety of measures in the areas that are their highest spending rates.  And so that can be one of life's great mysteries is how we got into this world of these hugely desperate patterns of medical care use and in many cases where there seems to be no measurable benefit arising from it.  So that's kind of an outline of where I'd like to move this today.  So let's look first at the low bang for the buck.  This slide got a little weird in the projection.  The graph here is per capita health care spending on horizontal axis ranging from this case from a thousand, about 7,000, and life expectancy on the vertical axis life expectancy at birth, and it's kind of a sloppy fit.  There're a couple of countries down here.  Number 11 is Hungary and number 21 is the Slovak Republic.
>> They have quite low life expectancy.  And nevertheless there's a trend line that's kind of the sloppy fit, but nevertheless, the trend line it says more medical spending is associated with more life expectancy and then the United States is over here.  That arrow was supposed to be right about that.  There's the United States, way below the trend line which would be up there at the medical spending we have.  Now life expectancy is of course not the only measure of health benefit.  There're a lot of gains we can get for our health care system beyond life expectancy.  But does one of them, we can compare across international boundaries.  And the second which is this level of infant mortality, it's really perinatal mortality which is a definition of infant deaths both before and after near the time of birth and these are data from the OACD looking at in mortality rates.  And here the fit is somewhat tighter.  Again, country number 12, it turns out to be fairly low in this Iceland.  I think that's a little bit of an anomaly of data reporting in Iceland but a pretty descent trend line fit on that, that the more we're spending per capita on Medical care the better our infant mortality rate is, except for the United States, way up there off the trend line.  Now, these are the only kinds of data you can actually made comparisons of outcomes across nations with any sensibility.  There had been some more refine studies that look, for example between Upstate New York and Ontario province right across the border in the across lake, Hunter, very similar climates similar population and education and earnings.  And for example, United States does about twice the amount of coronary bypass surgery as the same citizens age-adjusted and sex-adjusted citizenry, I guess in Canada.  And there's no difference on mortality outcomes.  But when you go and then look carefully underneath that in a more finely tuned comparison, you see that the citizens in the United States with the higher rate of coronary bypass surgery have higher capability and activities of daily living like being to walk upstairs and walk at the store.  They have more freedom from chest pain.  So in the lot of more refined measures you're actually seeing health benefits arising out of this and they don't show up these international comparisons.  But nevertheless, these are what we have, and as I said we might be looking for the key under the lamppost, but these kinds of numbers are disturbing.  I've also, I don't have a slide to show this, but had people wonder, "Gee, if all of this affect on life expectancy comes with infant mortality, what happen if we could clean out the infant morality affect and then see what happens to life expectancy?"  So, I've looked what the life expectancy is for a person aged 25 instead of at birth and the picture looks essentially just like this one.  So cleaning out the infant mortality story doesn't change the fact that we're still having low health outcomes for health care spending in this country.  So that's problem number 1, problem number 2 is we're getting older.  Now these folks should make the young folks of you on the room real nervous.  This is a standard demography picture, it's called the population pyramid, and when they look just right you start out each horizontal line as an age bracket and the size it shows how many people there are in that age group.  The one's on the right are females, the one's on the left are males and the traditional pyramid which is what it's called that, it just slopes up so it's shape like a pyramid and peaks with the single of narrow thing at the top.  And this one in 1950, you can already see the beginnings of changes in the population about this young age group and particularly, you can start to see the baby boom.  And then as we move to the year 2000 that baby bloom is now up in the age bracket, and so like a lot of us, the population pyramid in the United States as a midriff bulge, a topic to which I will return shortly.
[ Laughter ]
>> Take this out to 2020.  These are extrapolations, but demographers can do these extrapolations with very high degree of accuracy.  And you can see that, that pyramid is almost become a cylinder.  Let say, there's almost just many people in the elderly age group that start at the younger.  And remember in our system of Medicare financing, the people in this age group were paying the Medicare taxes that are financing the healthcare for this group that works a lot better in this population than it does in this one.  And we take it out to 2050 when I won't be around.  But you can see that this is now it had become column but it's become in fact an inverted top hat.  And you also notice that the line on the right side is a lot bigger.  Women live a lot longer than men, about 6 or 7 years longer life expectancy in our society.  Some of it is probably hormonal protection.  I think most of it is due to the fact that men smoke a lot more than women earlier in their lives, another topic to which I will return.  These numbers now, of course what's scary about these numbers that elderly persons, because our bodies are wearing out, we're just using medical resources a lot faster than younger people.  So if the population ages like the 2020 number which is not real far away, 8 years away, those kind of population distribution just guarantee your gonna be spending more money on healthcare, no matter what else happens.  And the 2051 is actually real scary.  Now, my son will be up in that age population, 2050 but I, I will not.  Okay, sorry, wrong direction.  Okay, now problem number 3 is income will continue to grow and so medical spending this turns out to be an extraordinarily strong predictive power.  If you know how much income per capita a society has you can tell with the high degree accuracy what their per capita medical spending is going to be and that statement is true almost independent of how health care is organized in their society.  So we have again, you can see the exact nations, but in the middle cluster here, we have nations like Germany and Japan that have a health care system does not completely dissimilar from ours.  But universal insurance you have Australia which is quite marketer oriented, you have Canada just to the north of us which has social insurance--health care system not the insurance but health care system is  much like ours.  You have Great Britain which has the British National Health Service.  And you have Sweden which has health care provided by counties, and all of these are on this extremely type fitting regressional line.  And it doesn't matter how the system are organized, you get to the same health care spending because of the income, it just a very strong relationship and it occurs in every data set I've looked at across many years.  So what that tells me is as our income grows, in fact of all societies, as our income grows, we're gonna be moving up in spending line.  Now the other problem of course is there's the United States is way up off that regressional line.  It's a very high outlier.  If you think that there's a curvature to it, you can bend a little bit of a log fit that seems like it goes up there a little closer but the US is a big, a above the line outlier on any counts.  So not only is our income growing but we're way up there on spending.  In several reports, I've seen say, a nontrivial chunk of that is due to the way we compensate positions relative to other nations, but that's a question which I'm not gonna try and delve into today.  Problem number 4 is technological change.  Now, this I'm sorry is a very hairy slide but I'll just give you an easy way through it.  Okay?  Let's look at actual dollars spend in 1960 and 2010 in the US economy on healthcare.  These are billions--billions 'cause we're talking about the federal government or the total economy, 23 billion dollars spent in 1960.  That was the total amount of money spent of medical care in the US.  In 2010, that was 2, 435--2.43 trillion dollars.  So the ratio of that 23 billion to that one, the ratio is a 106.  The amount of dollars being spent increase by a factor of 106 over that 50 year of period.  I just wanna look at these numbers across the bottom so, you know, what they mean.  Just simply taking adjustment for inflation using the consumer price index and we totaled these in cost of 2005 dollars, that number drops to merely 14.  Still the CPI adjusted 14 fold increases.  We can then say, oops the population is growing, let adjust for that so we can do is on a price adjusted and in per capital basis, the number is now 8.2.  Now comes the last step which makes every health economists nervous.  Let's adjust it by the medical price index which is a component of the CPI.  That's a real nervous time because the component from medical care in the CPI doesn't take quality adjustments into account, for example a major component of this, a semi-private hospital room.  Well, that's a very different [inaudible] now than it was in 1960.  So if you make that adjustment, this falls to 2. 8.  But some of that is technical change, technological change.  So something between about a factor of 3 and about a factor of 8 is the effective new technologies introduced into our health care system.  And I can't tell you if that a big gap.  I can't really give you a lot more precession in that somewhere between 3 and 8 fold increased in real per capital spending that's technologically driven and that technology is continuing to come down the pipe.
>> Now, just to reflect back and then--many of you in the room are not old enough to remember some of these.  But the changes in medical capability over this 50 year period have just, they're almost beyond imagination.  In 1960, if you had a serious bodily injury they could take a plain film X-ray.  Now if you've ever look at an X-ray compare to an MRI or CT image, the X-ray looks like a crude children's drawing for most medical problems compared to what you can see in modern imaging.  And then PET scans actually cannot only measure structures but physiology.  They can actually find out where cancer is growing with the PET scan by injecting with a little radioactive sugar and then watching it glow on the images.  Heart attacks, another technical change.  When I was young and growing up in Denver, Dwight Eisenhower, the person had a heart attack while he was on Denver.  The best medical treatment around at that time including for the president of United States was 8 weeks of bedrest.  We now know actually that's pretty bad for the patient.  But what we have now instead of that 8 weeks of bedrest is we now have coronary bypass graphs, implantable defibrillators, stents, a whole bunch of drugs that reduce the risk of the heart attack itself and eventually heart transplants.  All those extend life and in fact about the third of the gains we had in life expectancy over this period of time have been do to improvement and the survival from heart disease and they all cost a lot of money.  Middle illness treatment is preceded similarly.  The treatment in 1960 was if somebody was schizophrenic or other serious medical disorders, you put him into a mental institution forever.  And now we have a whole bunch of non-trivially expensive pharmaceutical products that actually reasonably well.  Not perfectly but reasonably well-control psychiatric illnesses.  Knee replacement, something near and dear to my heart, I tore an anterior cruciate ligament in an intramural basketball game in college and it has been grinding a way slowly ever since then.  And I know instead of just wearing a leg brace around and taking a lot of Advil, I have a metaled knee.  Okay.  It's hugely improved the quality of my life.  Before that I would not have been able to stand here and lecture to you, I would have been setting in a chair here.  The only downside is that every time I take an airplane trip, I get a free massage from the transportation security agency on the way through the gate.  And it's a whole bunch of implant devices like that, some of which you shouldn't bother with, but certainly knees and hips have worked well. Shoulders kind of a risk, don't bother.  I've experiencing almost of these questions in my own make up.  Strokes, if you have strokes in 1960, you got some physical therapy and thank you, good luck.  Now there's an anticoagulant, they could be administered if you get them soon enough from the emergency room after your stroke that will vastly reduced the neurologic damage.  They have robotic surgery to clean up the messed up arteries and blood circulation to your brain to improve your recovery.  And diabetes, it was a once a day shot with insulin which left these wild swings in the insulin control.  Now you have pumps that continuously monitor the blood sugar of the individual and feed in insulin slowly and steadily.  The vaccines, just one more, we have almost eliminated polio, haemophilus, influenza B, cervical cancers to some extent, and most of the common killers of the 19th and in the first half of the 20th century with vaccines and antibiotics.  There are now 25 active vaccines in used around the world.  The next will cost a billion dollars to develop and each one thereafter that or more.  So problem number 5 and this is the one that puzzles me most.  When we start looking regional patterns of medical care use in the United States, we get very strange maps appearing.  These maps come out of the Dartmouth Atlas and Jack Weinberg and his college at Dartmouth Medical School.  This shows in the color coding down here the darker the coloration for a region the more per capita spending.  The national average is about 9,000 dollars in this year.  This is in 2008, the most recent year they have up.  And you can see these clusters of high spending, Texas, Florida and then this band along Northeast and this band along there.  Now, I wanna just take a little more careful peek a couple of these things.  Zoom in on Florida.  This is very illuminating in a number of ways.  First of all I remember the national average.  This is a couple of years older now so the average in this was 8700 dollars.  Miami is at the top of the heap at 17,274 dollars per Medicare recipient, these are all age and sex-adjusted by the way so this isn't just the Medicare people in Miami or older and these are all age, sex-adjusted.  Just across the state, at 8300, a little under the national average in Fort Myers, Tampa is a little above, Orlando is a little above, Tallahassee is still lower up at the top of state.  Almost 10,000 dollars a year difference in treating Medicare patients per age, sex-adjusted per person in Miami than there is in Tallahassee.  Now, the other thing that goes on is when you go in and look carefully at these and this was what the reference to Fisher Wennberg and Stukel from 2003.  You don't find any differences at all in mortality or activities of daily living, functional measures or satisfaction with care or patient's self reported health status.  There are no differences across these various regions in any of those measures from Medicare recipients.  So, this really raises the question up of if these differences are going on and we don't seem to be able to measure benefit in very many ways, if any in, in this work at the ones they've looked at, no way.  How do we get into that situation where we're apparently spending health care dollars without any significant benefit?  Yes, sir?
>> Is cost of living in general and Miami [inaudible].
>> Cost of living is about 15 percent higher in Miami than, but certainly not than in Fort Lauderdale.  The other question I'm often asked is this, maybe it's malpractice?  And that's one of the reasons why I really like this slide because this is all and exactly the same legal environment.  It's the same malpractice law in Florida and every city.  So, it's really hard to pin this on the malpractice law because it's all the same here.  So, one of life's great puzzles is to how this happens.  If we look, I'm gonna turn up the microscope a bunch of ways instead of total cost of treating Medicare patients.  This looks at hospital discharges for thousand enrollees.  The US average is about 336 and then just wandering down interstate 94, Kalamazoo where I spoke about this yesterday, is about 340 and Arbor is about 360, Detroit is about 405.  Remember these are all age sex-adjusted measures.  So, what's causing this?  Maybe it's the Appalachian trail?  Probably not but it's a pretty good fit.  It's better fit than some of the regressions I showed you earlier today.  But that's--now I wanna keep turning up the microscope here just rapidly and show you what's going.  So this is hospital discharges for thousand enrollees.  So, now, let's look at in-patient care in the last six months of life.  And you see the same pattern essentially but now not just for hospital care but for a more specific type of hospital care and that is in the last six months of life.  Of course we never know what's the last of a person's life are until they passed away, but nevertheless, there's a trend here that's showing up and when we look in at the percent of all debts within the hospital where we do know what some of the alternatives are for example hospice care or nursing home alternatives, the map looks extremely similar.  You still see the Appalachian trail effect.  And in fact we can turn up the power of the microscope one notch further and look at the use of intensive care in the terminal hospitalization and again you see essentially the same pattern.  There's a little more of this out in the, in the south west of the United States and southern California.  And if you look down on this floored area, every time we turn up the microscope we see the same sort of patterns emerging and this tells me a couple of things.  Obviously, one of them is that it's not waste fraud and abuse.  There's certainly a lot of that going on in the Medicare, there's fraudulent, billing for things that never happened, not in intensive care units folks.  That's not where the waste, fraud and abuse is going on and these patterns still keep emerging.  So, I'm convinced from looking at this, there are very important differences in the way communities come to believe that the health care system should be used.  There are agreements between patients and their doctors.  I have no idea how those implicit agreements are set but we confine them and they keep repeating themselves year after year after year and it's not due to age, demography or illness patterns, yes?
>> [Inaudible] is there any comparison between utilization and Medicare patients and whether or not there are additional Medicare or they're on Medicare each moment?
>> Yeah, most of these data, these have been done over a series of years and they look very similar from a year or even before Medicare advantage part C came into effect.  The Medicare part C enrollment is getting large enough.  Now you begin to parse that out.  It hasn't been done yet by the folks at Dartmouth.  I don't--it would be interesting to see what the answer is.  Of course I'm now on Medicare since I retired from the active faculty of the University or Rochester, I had to go to Medicare, and before that I was on a high deductible HSA savings--consumer-directed health plan.  They wont let me have one of those in Medicare and it just drive me crazy 'cause I thought it was the best thing for me and certainly not for everybody but my choices are limited in that.  So, there's the policy problem staring us in the face.  I mean certainly there's a lot of variability in medical care spending.  We cannot say, and I wanna emphasize, we cannot say with any reason, plausibility that Miami is too high.  All we know is it's bigger that anyone else.  But they might be dead on just right and we might all be spending not enough.  Now of course we do have this notion that we're not seeing any health benefit from that extra spending compared to other places in Florida.  And that sort of analysis is giving us a stronger suspicion that in fact the high end is too much.  But that's really the problem, yes?  We have a question in the back, yes sir?
>> I was just wondering, what about the distribution of subspecialist, what, what does that look like and the distribution of hospital beds or--
>> Yeah, the folks at the Dartmouth Atlas have, the question is, does this, is this following patterns of medical specialists in hospital resources.  And the answer is the folks of Dartmouth have done some work on that and you could probably if you want to prowl around at the Dartmouth Atlas, it's a pretty good website, and you can find more detail answers to that.  The problem that economist see when they look what that is, we don't know whether the resources have cost utilization or the fact that people want the resources and they flowed there to respond to that and it's very hard to disentangle that particularly in these types of data.  So, I suspect that there is some correlation between those.  I'm not willing to assign causality at the fact that if you put some resources in there that people will come and use them.  They have to want to use them after all.  Now they're getting--we do know that there're some important features about that.  Florida is rampant for example with physicians owning various types of like imaging facilities.  We know from very careful study in Florida that physicians who have ownership's taken imaging facilities, orthopedic surgeons recommend imaging studies at four times the rate of orthopods who don't own imaging facilities.  So there's certainly some of that going on and I would be foolish to say otherwise, I can't tell you how much of it.  And then one also must wonder, why is there so much of that going on in Miami?  Well, if other parts of the same state which have the same tax law and the same malpractice law and everything.  There're a lot of reasons we think we know what's going on some of it but I don't know anybody that can say with confidence how much of this is due to various parts.  And I would, if I said that you should run me out of town because I have no logical basis for asserting how much is due to what.  And thank you, that's a very good question.  This leaves me to the puzzle if, I'm prepared to say, we seem to have a lot of medical treatment going on because these variations without apparent benefit, so how do we avoid that?  And I'm gonna [inaudible] some.  I'm always thinking about incentives so I wanna puzzle around about the incentives that lead to that.  One of them is setting aside the Medicare population.  We have too much health insurance in this country for those who are insured.  Our tax system subsidizes health insurance.  The average subsidy is about 35 percent so we end up with people buying health plans that don't have any cost consciousness build into them.  We end up with people ensuring, if there are dentists in the room please forgive me, there's no particular reason to buy dental insurance, there's just not much, very much financial risk associated with that.  If you cave your face in, in a vehicle accident your medical insurance will pay for it anyway, the reconstructive surgery.  So, we're really talking about insuring of dental insurance, things that are very predictable on average, not quite as predictable as grocery bills but pretty predictable.  And we ensure that because our tax system subsidizes it and that has derivative effects on how much medical care we use.  And that's true also almost every type of, of health care.  How do we guide decisions to the most appropriate intensity of care?  In fact, it'd be nice to know where is that, the most appropriate intensity and we don't know that.  How we pay for care?  Why do we pay for care with little or no benefit and these are not easy problems to fix and I don't know the answers but they--I think there are some with the core of the issues confronting our society as we look forward.  But that is not the fundamental problem.  The fundamental problem is ourselves.  The first set of problems are related to the government and government policy.  The second problem is to steal the line from pogo.  We have met the enemy and they are us.  This is one of the most eye-opening, and to me eye-opening and transforming research studies I've ever seen in my life.  Let me walk you through this.  The first of this was done using 1990 data by McGinnis and Foege and then replicated using 2000 data a decade later and somebody I'm sure is furiously working on doing this with 2010 data now or will be very shortly.  But the numbers are pretty striking.  So let's talk about what this means.  This is the actual causes of death in United States.  Now death certificates, there's a national death registry in the US, and the death certificate list the cause of death that the physician has written down.  So the physician says lung cancer, that's the cause of the death on the death certificate.  What McGinnis and Foege did was to go through all of those deaths and look at the deaths for lung cancer and then sort out how many deaths are attributable to tobacco use.  And then those which are not and the deaths that are above and beyond that which will be predicted for nonsmokers are excess deaths that are attributable to tobacco.  And then they add that up across lung cancer and a bunch of other cancers and emphysema and chronic obstructive pulmonary disease and heart disease and the many other things that smoking does to cause, medical care cause and medical expenses and death and they added up all the extra deaths associated with tobacco.  There were about 2 million deaths in the United States. This 1 million was about 50 percent of all the deaths in the United States, 400,000 of that w million were attributed to tobacco.  Next on the shopping list in 1990 data was 300,000 to poor diet and physical inactivity, alcohol consumption was third, microbial agents is fourth.  That's the bugs that we run, toxic agents in our homes and in our work places.  Motor vehicular accidents, firearms, sexual behavior, elicit drug use, 1 percent of all the deaths in the country, 30,000 deaths in that year.  That number has gone down a lot since then because of HIV treatment.  So, and then elicit drug use have 20,000.  So let's look at the 20,000 and the 300,000 and 100,000.  This is a completely side comment but folks were fighting the wrong drug war.  Let's leave it at that.  This set of excess deaths from this things adds up to about half of all the deaths in the United States.  Incomparable by the way in medical care expense, I'm quite sure, although I have not seen detail on that.  Just a second, I'll get to a question.  Almost all of these are lifestyle choices we make.  They're our own behavior.  Perhaps the only counter examples to that will be microbial agents and toxic agents just some extent, but I'd say almost all of these are behavioral choices that we make.  Now the question, yes sir.
>> I'm just wondering how they're determining the baseline for the national history of illnesses. You're saying that they determined that baseline and then whatever is involved in that, that's what going in and they're treating [inaudible].
>> Yeah, so--
>> So how do they getting up baseline?
>> Yeah, the way to do that is epidemiology studies in every one of these cases.  Thank you, a good question.  Epidemiologists calculate the relative risk of smokers and nonsmokers, and then you look at the cohort, it's just tied in.  You can tell from good survey data what fraction of them are smoking and you look at the excess risk from smoking in that population and you can backout excess deaths out of that. So it's basically all hinge on careful epidemiology studies of the relative risk of smokers and nonsmokers.  That's a quick summary answer.
>> Do we have those for separate cause?  In smoking [inaudible] study, I think that we had that for other thing?
>> Yeah, yeah, there's good data on enormous variety of the--In tobacco consumption for example.  I'd even seen really good relative risk behavior that says the pets of tobacco smokers have a higher risk of death than nonsmokers.  Second hand smoke prevailing its way in so yeah, I'm confident that McGinnis and Foege, their article explains this in more detail and then the subsequent work, they've got pretty solid up the evidence on all of these relative risk numbers.  They maybe wrong a little bit but even if they're wrong a little bit, the story is quite strong, I'm convinced.  I have question here and then with you, yes?
>> You may be getting to this but the causative deaths don't necessarily translate into higher health cost.  Those people who are dead are no longer on the healthcare system.
>> Yes.  Well, that's particularly--particularly true of tobacco.  Actually, there's a lot of medical expense, I'll show you a slide on this in just a moment, attributed to this behaviors.  But in particularly if you add up across all that public pools, tobacco users/smokers die earlier and hence although they'll spend more in Medicare, they don't consume as much as social security resources.  So, there's actually a careful book written about this by Will Manning and some colleagues called "Do Smokers Pay Their Way, that goes into that in quite some detail and the answer is almost when you count the fact that they're not drawing social security resources nearly at the rate of nonsmokers so.
>> I just wondered if that's been compared to European countries.
>> Yeah, actually, not just European countries but anybody who has traveled knows there's a lot more smoking going on in Europe and Japan and there is in the United States.  So one of the puzzles which I have pondered a lot is, how come if they're smoking so much more they have such a better mortality rate and I can't prove this.  My suspicion and it's just a guess is that they make up far more in these other dimensions probably mostly on obesity and lifestyle.  They walk a lot more.  They're not as heavy and in general and body mass index calculations across time and space.  And so, my guess is that they're making it up on this other behaviors.  But it's a good puzzle and it's a great dissertation student for somebody here in the Ford School.  Track down data that would actually look at the rates of these various behaviors and other societies and unravel how it is that they can smoke so much.  And partly by the way in Japan, the standard of the highest longevity anywhere in the world is Japanese women.  They don't smoke, okay.  So you could get some comparison just by looking at men and women in Japan on that effect.  Yes?
>> I don't see any comparison with diseases who doesn't have cancer, [inaudible] so it's hard for us to--
>> Okay, so the rest of that like the cancers that are for people that didn't smoke, they're down on the next 50 percent.  Did I get your question correctly?
>> Yeah, like breast cancer.
>> Yeah, oh, so right.  If you have breast cancer and it's not and if it doesn't have an extra relative risk compared to these things, it's down here somewhere below the top 10.  They're all there in that list.  That's the other half, the other 50 percent that are on this list of the top 10 and probably, you know.  Some of the cancers even though it's non-attributable to these lifestyle choices are in that next 10.  But I really want to focus on the effect of lifestyle on the top 10.  Yes?
>> It would be interesting to compare is how physicians Europe fill out the death forms because in United States it's very common to kind of be a little sloppy but also include the lifestyle issues when you fill out death certificate.  And I don't know how they do it in other countries.
>> Yeah.
>> But they may not quite--
>> I don't think physicians are given a choice much in filling out death certificate that the cause of death was smoking.  You know, okay, okay.  They have just treated those directly.  And honestly I'd have to go back and look at the details at the McGinnis and Foege work to know how they have done that.  Again, many of these numbers aren't right.  Maybe they're not right by half but I'm really convinced there's a kernel of truth there that says our own personal behavior is causing a lot of our medical spending in deaths.  And there's some good news.  This is a graph of cigarette consumption per adult 18 years and over since 1900, it peaked at almost 4,000 cigarettes per 18 year old.  So here's the history quiz.  No fair if you're an adult in 1965.  What happened in 1965?  There's gonna be on the final exam for the policy students.  What happened in 1965 folks?  Surgeon General support, yeah.  Surgeon General of the United States put out a large scientific analysis that said smoking is really bad for you.  It was the first time there was official government statement that said that.  I was in high school in Denver in the 1950s.  We knew that.  The junior high school students--junior high schools [inaudible] called cigarettes cancer sticks in the 1950s.  This wasn't like it was news.  It was just, this was the official statement ever for the first time and then following that came all of the societal changes.  Taxes on cigarettes, warning labels, no smoking in public places, no smoking on airlines, changes in social mores, some with the development of ways to unaddict from nicotine with patches and things like that, but the result has been this dramatic decline in smoking, so are now down to well under half of what the peak was.  And so, we're making great headway on smoking.  Not perfect but still great headway.  The smoking patterns are very tightly linked with educational attainment.  This is a snapshot of the current time.  I'll show you another one later.  People who have not completed high school, there's a final educational attainment, about a third of them smoke cigarettes now.  Those who have high school degree, it's a little under a quarter.  If they have some college probably most of them have gone through to junior college and got an AA degree.  It's a little under that about 22 percent and college graduates are more down to about seven percent.  And physicians and the health care system, it's about 2 percent.  A good friend of mine is a radiologist, keep smoking and that's why he does he's beyond me because he keeps looking at x-rays of lungs of people to smoke but he still smokes.  Nicotine is extraordinarily addictive and there's no better addiction mechanism, no addictions science I have talked into.  There's no better way to addict somebody than to have small doses delivered frequently, cigarettes, okay.  So, it's powerful addictive.  I'm gonna back to this education thing a little bit later but there's this powerful gradient of smoking and education.  Somebody over here asked me, doesn't it cost less?  Smokers consume about 40 percent more year than nonsmokers in health care cost, age-adjusted.  None of this appears in their health insurance, premiums in general particularly to employer groups which is very much unlike life, home owners, and auto insurance where your experience created in a lot of behavioral things.  You know, if you smoke and try to buy a life insurance, you're gonna pay a lot more for it 'cause they know the effects.  They know them equally well and they're health insurance business, they just can't do it in player group policies.  Auto insurance charges more for under 25 year old males, testosterone and gasoline are really dangerous mix.  And even home owners insurance charges more for smokers because they fall asleep in bed and set their homes on fire more.  So we have all this experience rating on lifestyle things and other areas but not in healthcare and that's another one of these puzzles.  It's not just the health care cost, there's a lot of lost productivity in terms of work loss, people are either ineffective or can't appear for work.  When asked what I've seen which I wouldn't put a whole lot of faith on the exact number, but it says the true cost is a 150 dollars per pack.  Most of that is productivity loss.  And I have mentioned this irony since smokers die younger that health care cost in the elderly and particularly that social security cost are saved.  Most of the smokers end up going through an expensive end-of-life illness.  Often an intensive care with their various illnesses but they do exit the Social Security systems sooner.  So what's to worry about?  We're getting all this improvement in smoking.  There's little flattening out in the declined teenage initiations, there's a little concern for people that are really carefully looking at, but as I showed you in the chart of McGinnis and Foege work and the one a decade later.  When that's done again with 2010 data, smoking is gonna lose its unenviable position at the top of the charts.  The reason is the United States is about to drown in the sea of fat.  Let's just focus here on the upper three lines here.  These are children of various ages, so let me just focus on the top three lines here.  Overweight but not obese is steady flat.  How is this measured?  This is body mass index.  The body mass index is calculated most conveniently in metric.  It's your weight in kilograms divided by your height in meters squared.  If you don't know your body mass index, run, walk home after we finish this session today and Google BMI index putting your weight and height and you will find out what your body mass index is.  I'll show in a minute how important that is in terms of--at least in terms of mortality outcomes.  Overweight but not obese has been flat pretty much among the adults 20-74 over this period from 1960 , the most recent times available on this.  That's a BMI of 25 to 30.  BMI of 30 to 35 is the screen line.  It runs charging along into about 1975 and then [noise].  And then the total of those has the same behavior.  This is adding up everybody who's both in this line and this line so that kinks [phonetic] at essentially the same time.  And there's this sudden change in people with the body mass index of over 30.
>> Their prevalence in our society about 1975.  And I had a lot of reasons so I can understand why people are getting heavier but why it occurred suddenly and that kink behavior.  And 1975 is one of life's great mysteries to me.  Perhaps we can have a discussion about that in a little bit later.  Now, this is going to come happy, surprised to some of you.  By the way my body mass index is 28, just probably 29 when I get back from this week long trip but traveling is not good for my body mass index.  The upper set of these is the mortality rate by body mass index for males and you can see it hits the bottoms.  That's the healthiest area around body mass index of 23, 24, 25, 26, not a whole lot of difference there.  As you climb back up into very thin people, the mortality rates increase.  I think some of that is due to when people get really sick, they lose weight so I don't think it's causative.  On this side of the graph, I don't know if many ills, illnesses have caused you to gain a lot of weight.  So, I think probably this side is causative quite directly.  And the difference between this set, and by the way this is a summary of about 900,000 lives across, a bunch of prospective studies.  So, the statistical power of these members is just enormously high.  The bottom set of blocks is, is women.  Women have lower mortality rates and live about 7 years longer in our society than men.  Good part because they didn't smoke so much.  But fat is, is definitely not good for you, probably and even better measure is grease line circumference.  We have good data on all these studies on body mass index so that's what I can show you today.  Obesity cause accounts for about 10 percent of the health care cost.  This is again one of these excess cost measures, about 40 percent more.  They're very similar to tobacco smokers than people with normal weight.  And this problem is going to become larger, pardon the expression through time because our population is growing increasingly obese.  And smoking just makes you feel terrible.  This is a study by a really good health economist David Cutler and colleague published in New England Journal of Medicine a couple years ago.  They took survey data from a whole bunch of people around the Unites States and basically to answer a question.  Rate on a scale of 0 to 1, 1 being perfect health, how do you feel?  And this summarizes the answers from this very large population study.  So, let's just take college age students who are at normal body mass index and don't smoke and their answer is about 90 percent are perfect.  Now, if you go among the non smokers across body mass index you see there's a, somewhat of a decay in that reported, how well I feel?  And when you drop down to those current smokers and then run across to people who are over 35 body mass index and smoking, that number is 74 compared to 90, smoking makes you feel terrible.  That's true also if you look at people aged, let's say, aged 55 to 64.  82 that's not as good as 90 because you know, time wounds all heals, it wears us out a little bit.  But if you make the same comparison of that age group from nonsmoking healthy weight to smoking and higher weight, it drops down to 0.67 from 0.82.  You see the same kind of proportional drops in the quality of life that people report when they're in these unhealthy states.  So, people in these, smokers and people who are overweight are bearing some of the cause personally in this immediate health status and make no mistake about it, yeah that question, yes?
>> Well, I was just gonna say it will be interesting if this, I was thinking about mortality.  I mean there are some studies that show that mortality among obese people is not higher than not obese people except for those people who developed diabetes or hypertension or heart disease.  Except for the morbidly obese like you're not--.
[ Inaudible Remark ]
>> Right, but I think that, well, I think, I mean obviously people that are obese are more likely to have diabetes.  But it would be interesting to see what happens with the, like the Asian population announced that there's a rapidly increasing percentage of people with diabetes without high BMIs.  And so it would be interesting to compare mortality with the diabetes and the lower BMIs.
>> Right, we have, I have seen the studies some time ago.  I couldn't quote the author for you yet.  Right now they looked at first generation Japanese immigrants in Hawaii and it turns out when, you know, Japanese citizens immigrate to Hawaii and adopt western diets, they acquire western mortality characteristics quite rapidly so there's probably something there.  I'm not saying that it's just the fat but there's also, well there's the diabetes that comes from fat and other things are beyond, is kind of beyond my point.  I've good data on the fat mortality story and smoking and that makes my point I think.  So now comes the question, why are we having this increasing rate of obesity in the United States?  And a sub-question is why it has started suddenly in 1975?  I have some reasons for it that suggests why I should be growing but not why there's that kink in 1975.  We have more sedimentary jobs now than we used to and so we're not shedding calories during our work which means we have to go outside to the track or to the gym and spend valuable hours getting rid of those calories and our time is increasingly more valuable as the society becomes more educated and our opportunity costs rise.  And so basically the cost of getting rid, we're not, getting rid of this as many calories in around work and the cost of getting rid of them is rising.  The other half of this is that food, because of the efficiencies in the agriculture sector calories are becoming cheaper through time.  And if there's one thing economists know is when things become more expensive people do less of it and when they become less expensive, people do more of it.  And so calorie cost shrink and customer getting rid of calories rises and you have an obesity problem emerging right straight out of the economic forces here and probably happening faster than our genetic selection is going to alter our basic metabolic rate to deal with this.  So, that's just kind of normal laws of supply and demand if you will.  One simple example is, I'll show you some more detail on this in just a moment.  But the amount of food that people get in the home has changed versus restaurants or order in or buying package foods in stores and preparing at home has changed dramatically.  So, food in the home has gone from 1930 at around 85 percent of all the meals were prepared in the home and it's about 55 percent now in a very steady decline.  And the converse of the flip of that is meals eaten outside the home and those lines are gonna cross pretty soon.  And then you couple that with what's happening to prepared foods in the grocery store when you buy portion foods and go out to restaurants.  And there's a study from the National Heart, Lung and Blood Institute website, there's a sample.  Every, every food they've looked at has increased dramatically in the calorie content of what was described as a portion over the last 20 years.  You can read these yourselves, bagels are much bigger, chocolate chip cookies are four times as large or have more sugar or butter or both, french fries quantities have almost tripled, that wonderfully healthy chicken caesar salad has got twice as many calories as it did 20 years ago, theater popcorn is more that doubled, my daughter worked in the movie theater in Rochester, she was actually rewarded by the number of people that would, at her request, would you like to super size that?  Her pay increased when she had more super sizes rung up on the cash register.  Another, I love eating turkey sandwiches for lunch, they're very healthy, they have two and a half times the calories as they did 20 years ago and on and on and on.  Soft drink you know, it used to be a 10 or 12 ounce soft drink, it's now 20 ounce soft drink and you buy the big cup.  And beer is increasing in proportion, not by the way, they're now looking at 20 ounce beers, that's the primary container instead of 12 and you know, this is, and all they have to do is look at television to realize that the fast food chains are competing on the size of their portions as much as on prize.  My hamburger is bigger than yours; my French fries portion is bigger than theirs, that's the vote of competition.  So, that's the problems.  What have we done about them?  Well, obviously the most important thing that has come along in the last, let's say 20 or 30 years since, probably since the formation of Medicare in public policy is the Patient Protection and Affordable Care Act or PPACA.  If you are a good fan of Rush Limbaugh, you probably call it Obama Care.  So, by summary of what the Affordable Care Act did in dealing with these questions comes right out of the book of, Episcopal Book of Common Prayer left some things undone, we should have done, we have done some things we shouldn't have done and there is no health in this.  So, let's take a careful look at the key features of the Affordable Care Act and then ask, what is this gonna do to deal with the set of issues that I've just laid out?  Probably the most prominent and controversial and under legal duress probably in the Supreme Court within a week or two is that mandate that every individual in the United States obtain health insurance coverage.
>> Most of us have through our employment groups, I suppose almost everybody in this room does either by as employee of the UM or as a student here, I've always head up to my employers of random university of Rochester, that's the common mode in about 80 percent of the people under 65 in the US health insurance comes through their employment group.  So, hence the laws or mandates that all employers provide health insurance for their employees and families of at least bronze label.  There are like the gold medals in the Olympics.  The act defines bronze plans as those that will cover at least 60 percent of the actual real value of the plan.  Silver is 70, gold 80 and platinum is 90 that's an extremely generous health insurance plan.  And they mandate you to at least have a bronze plan.  It establishes regional insurance exchanges.  The states who are in sub regions they can cooperate on these.  They will help people who don't have employer insurance find health insurance at lower costs by setting up these if you will, shopping marts.  And it also provides subsidies for low income families directly through the exchanges and also there're tax credits through the income tax system.  Quite importantly it bans the use of preexisting conditions by ensures and underwriting.  Now this is really, this is a very interesting issue for two reasons.  One, it ties in to the mandate itself and then the other it just a separate question of whether it's a good idea.  So, let me give you the legal argument which a constitutional scholar friend of mine and I have developed that talks about the individual mandate for a moment.  All of the legal challenges that the individual mandate say that it is an unwanted and excessive extension of the Commerce Clause of the constitution.  The commerce clause says the congress may regulate activity of commerce in the United States.  I think it's widely believed probably 8 of the 9 members of the Supreme Court, maybe by all 9 that that includes insurance and hence can be regulated by the federal government, the McCarran Act in 1946 made some specific rules about that but it's clear that that's illegal activity.  So, the, the mandate that says you've gotta have insurance has been dealt with this idea, that it's basically saying in the common phrase going back to George Bush's statement, the big George Bush Senior.  Now, that I'm president they can't make me eat broccoli, you remember that statement?  He didn't like broccoli apparently.  So, this has been picked up in the political discussion about this mandate and says, if they can make you buy health insurance, they can even make you buy broccoli.  So, you see the general logic being attached to this and that is, it's enforcing consumption decision on people to buy something that they don't want to buy.  The alternative is to come back down to this ban on preexisting conditions.  Whether you think that's a good idea or not and I can make a good argument, I think that it is.  It's not a perfectly good idea but it's a good one is, as soon as you put that rule in then you have to require people to have insurance for the insurance market to function because, otherwise everybody that has two brain cells functioning will wait until they get sick and then they'll go buy their insurance because the insurance company won't be able to charge more for their preexisting condition.  So, it now falls into the separate power of the constitution which is the necessary and proper clause that says, the commerce may enact laws that are necessary and proper to carry out other legally allowed activities including regulating insurance.  So, if you think that the congress can regulate insurance market by saying you cant use preexisting conditions anymore, you kind of immediately, automatically, it's an easy legal path in to get to the point that says, okay we gotta require that everyone have insurance.  Not through the extension of the Commerce Clause, they can make you buy broccoli but rather because the insurance market will stop functioning if you don't have that mandate.  So, that's a complicated legal argument that's probably gonna hit the Supreme Court this week because there are now 6 circuit cords that have come to different rulings about whether this individual mandate is legal.  And the Supreme Court just has to take this all in my view.  So, back to the question about what else is going on under the PPACA.  They got a big incentive pushed to get electronic medical records in the physician offices basically a big bribes to cover a lot of the cost.  There's a lot of hope.  I would say some of the possibly grounded in fact but not a lot that this will reduce medical care cost.  There are certainly some things within the hospital realm where we know the electronic medical records can help at least in improving health outcomes if not saving costs.  You avoid a lot of drug prescription errors and drug interaction errors.  There's an amazing amount of deaths that has reported at Institute of Medicine a couple of years ago on the amount of death in the hospital due to pharmacy errors where they just gave the wrong dose and it killed the patient.  Hospitals are dangerous places to be.  I mean look those people die in hospitals.  They're really dangerous place to be, a little logical fallacy for you to tease out there.  So the question how much these electronic medical records are going to help remains open but that's one of the pushes.  It also really made a push towards a new type of medical care organization called the Accountable Care Organization.  These ACOs are gonna be paid not only in terms of how much health care they deliver but on the quality of care that they're delivering, measured on some fairly esoteric ways and increasingly astringent through time.  So, they're not just simply processed measures like, how many of the kids are getting their vaccinations of schedule, but much more interesting and sophisticated things.  They're not mandatory.  No provider has to join them, no individual has to enroll in them and we don't know how well they'll work.  Probably the closest thing that we see to them right now is the, kind of the full fledged HMO like Kaiser or one of those things but they're really somewhat of a different [inaudible] and there's hope there but no evidence yet about how much of the growing health care cost those are going to deal with.  They've expanded medicate eligibility to bring more people into the Medicaid net for low income people.  So now every state will have to have medicaid eligibility at least 233 percent of the federal poverty line.  That's gonna expand the number of people that have Medicaid as their source of coverage which is partly designed to offset the burden of those who don't have insurance coverage through their employment place.  There's some emphasis on prevention, for example insurance plans, a whole set of preventive medical interventions that you might envision those approved by the US preventive services task force, get it shunted around so there's no co-payments or deductibles.  They are all free by law in all these plans.  If prevention really works to save money that will help most preventive activities don't.  And by the way those are all medical interventions.  None of them deal with this lifestyle thing, none of them.  They're all vaccinations and screening tests and things like that.  They're putting a tax on so called Cadillac Plans.  That's probably a nasty word to speak about a car in adverse terms in this region of the country but what they mean is very expensive health insurance plans.  They're gonna have a tax on the excess above the cut off.  All the people in Miami, they're buying private health insurance, they're gonna hit that 'cause their health insurance premiums are high not because they bought a really generous plan but because health care costs are really high because of the style of medicine they practice in Miami.  Citizens of Miami are all gonna be paying this Cadillac tax pretty soon.  There are other ways to go about that.  I would strongly prefer instead of that a much more useful elimination of the tax subsidy to employer paid health insurance of which would just the way everybody's health insurance work.  And then take that money and put it back into the tax system by reducing marginal tax rates by comparable amounts.  So, you can grow the income base by 10 percent by putting all of those employer paid premiums in.  You can turn around and reduce marginal tax rates by--everything by 10 percent of what they were.  You'll get at least just as much tax revenue, it's a mathematical certainty 'cause you increase the base by 10 percent and you lower the tax rates by 10 percent and then you're guaranteed to get an increase in economic performance because there's just no question at all that lower marginal tax rates stimulate the economy.  There's some argument to how much but I can just guarantee you with a moral certainty that if they did that it would increase the economic activity in this country, it would remove a distortion that subsidizes and gets us having too much health insurance for those of us that have through work groups, and they would be at least a budget neutral almost certainly gained in tax revenue.  So, that's my, one of my, if I were the Czar.  And I would also, the PPAC also has beginning a link, part A hospital payments to quality of care outcomes.  In some ways, it's kind of the Accountable Care Organizations.  Those are the key features that a lot more but that's really where they're going.  Now, comes what I say are the unresolved core issues staring us in the face still.  To me the, the biggest issue confronting us in the future is we still have this enormous disconnect between how much medical interventions worked to improve our health?  On any scale you wanna measure mortality, morbidity, saving health care cost, feeling good, anything.
>> There's a big disconnect between how--how much those procedures work for us and how much we pay to get them out of pocket which is what's driving our decisions.  And we need to fix that and hence with that, the attitude is about how we use healthcare in this country.  The conversation right now with your doctor, I just--I can just bet with high degree of certainty, you've gone into your doctor and the doctor says, "Well, you know, I think you might have this, let's get this lab test or let's get this in MRI image and see what's going on" and the patient thinks about this more, says, "What will this do?"  He says, "Well, it will give me a little better information."  "How much will it cost?"  "Your insurance will pay for it."  "Okay, let's do it."  And if the--instead that the doctor said, "It cost a thousand dollars to produce that MRI image, you're gonna have to pay 250."  I think the patients will begin to ask a little more, do I really need this or can you tell me what's going on?  Can we maybe try a drug trial balloon and see if that solves the problem?"  Every single medical decision that we make in United States or anybody with any kind of health insurance has that subterranean discussion going on, and it ultimately is why we're spending a lot more money than elsewhere.  And it differs in part also with this cultural difference across regions that I just have no idea of why they happened but I can show that they do.  Yes.
>> [Inaudible] there is a real sacred cow about the end of life care, and that's where our patients don't have much to say about it because, you know, they're in intensive care.  And--
>> Right.  And by the way, a [inaudible] Medicare spending is in this end of life care over the last six months.
>> Right.  And it doesn't change the outcome.  And I will say that one of my patients who's a union administrative secretariat told me when she was in a meeting with the administrators at the hospital and they said, "You know, the best patient that we can have is somebody that comes, you know, comes from up north and gets life line again.  It goes into the intensive care unit, gets every test known to man and then dies.
[ Laughter ]
>> That's right.
>> Because that's what's reimburse.  And I just--it's really, I think it really needs to be addressed how much the reimbursement.  You know, if you build it, they will.  If you build it, the test done and cover it, they will do it.
>> Yeah.  Look, a nontrivial part of this is how we pay our healthcare providers.  And right now, their stuck on a treadmill, doctors and hospitals alike and dentists and pharmacists.  They're stuck on a treadmill where volume is important.  And there are different ways to pay healthcare providers that get you off of that treadmill.  The most extreme form of those is per capita payment for a year.  You get a bunch of doctors in hospitals together and you say, "You take care of the set up patients per year and here is all the money you get to do that."  They're called HMOs in the purest form.  That phrase-health maintenance organization has blurred a lot and its meeting to time.  But the pure form HMO over the physicians that are on salary, they're not volume driven.  It's well known even in randomized trials at the RAND Corporation where I did the, you know, participated in the very large RAND health insurance experiment.  One of the arms in that was absolutely free care, everything paid for.  Another one was an HMO where everything was absolutely free to the patient.  The HMO patients cost about 20 percent less and no different health outcomes.  The affordable care organizations that they've--are trying to spawn have many of those similar incentives.  Now, it's not a perfect world.  The fee for service system, every economist that has looked at this says, the fee for service system where you get paid on volume has all sorts of reasons to believe that it's generating excessive use of medical care.  The doctor gives you advice that brings money to the doctor's pocketbook and particularly if it's covered by insurance, that's an easy decision to make.  And patients go to doctors because they're willing to accept their advice.  If you didn't willing to accept what your doctor recommended you, you probably aren't gonna go there very often anyway.  So, doctor recommends when it's to their financial advantage to make a recommendation, they do it more.  How much is it?  Interesting question but they do it a little bit more in some cases and a lot more in others.  On the complete flip side of that where the doctors are paid on a flat annual salary, two things happen.  One is they don't work as hard and that's quite provable.  Doctor on a flat annual salary in a large organization where they can't monitor their care is about twice--half as productive as seeing patients as a doctor and a full fee for a service arrangement.  And just in terms of patient volume.  That's not the whole story and maybe it's better quality of care by spending more time with the patients but there is a change in that.  But the one I like most of all is a randomized controlled trial that occurred in a medical school at Washington University in Saint Louis in their Pediatrics Department some years ago.  I'm just a huge fan of randomized controlled trial.  That's the way medicine really tries to figure out whether a new drug or procedure works.  You can also do them in social science.  The RAND health insurance experiment that I alluded to was a randomized trial.  There's also a randomized trial about physician compensation.  It was really cute.  It took a bunch of pediatric residence and put them in to the well care clinic.  All these residents are seeing patient.  They're all on the similar training.  They randomize some of them to receive within the resident normally do as annual salary, and they randomize the other half to get a fee for service arrangement that they predicted would bring them about the same amount of money for the year.  And then they randomize the patients as to which one saw which type of doctor and they turn the system loose and watched.  And when they finished, the doctors on the fee-for-service system were generating about one more patient visit in the pediatrics clinic per year than those on the salary and almost [inaudible] well care environment.  There was almost no difference at all in--in really acute illness and injury stuff.  And the part I really like is my--my wife is a pediatrician so I--I know this reason.  The American Academy of Pediatrics has a--and also the CDC now, they have a schedule of recommended vaccinations for kids.  Pediatricians know the schedules as well as they know their middle name, okay?  This is not a great mystery to them.  The doctors on the fee-for-service system overshot the American Academy of Pediatrics recommendations a little bit.  They actually--in other preventive, they did more than it was recommended.  The doctors on the salary system were a little lazy and they undershot not by huge amounts but what happen was exactly what the economic theory predicts here.  And that is one of these gives an incentive to over recommend medicine and the other give incentives to under recommend because it's still doctor's advantage in both situations to go that direction.  And by the way, these were all recommendations to patients.  The patients didn't follow through on all the doctor's recommendations to come back for another visit and they were a little more suspicious about the doctors on the fee-for-service system.  That is they turn down a few more other visits and didn't bring their kids back.  So these all just absolutely fits.  The more powerful one on this effect is studies in Florida, physicians, orthopedic surgeons who own imaging facilities use imaging four times the rate--four times the rate of those who do not own imaging facilities.  That may account for some of that Miami thing.  Maybe there's more ownership of these things in Miami then elsewhere in Florida.  They use physical therapy 40 percent more often when they have an ownership stake in physical therapy.  There's just a whole bunch of really strong evidence of this, affects physician's recommendations and hence what medical treatments emerge.  But the magnitude to that affect just differs hugely across setting and I--I just can't extrapolate any of this that I know about more generally than the situations they're in.  Yes sir.
>> [Inaudible ] very good numbers in there.  What--what effect does CYA with respect to legal have to do with the studies?
>> Yeah.  So I guess the--probably the most powerful thing I can show you on that is go back to those Florida numbers.  You got twice as much money is being spent in Miami as elsewhere.  It's the same legal structure.  It can't be because they're worried about malpractice more in Florida than--may be there're more malpractice lawyers in Miami or than elsewhere but it's the same legal structure.  It's really hard to pin and protect, you know, defensive medicine.  It's really hard to pin this behavior on defensive medicine when you see such discord outcomes in the same legal environment.  And the only policy lever we have to fiddle with that is to change the legal structure and there it's all the same.  We've also had a whole bunch of things through time where states have changed their malpractice law so that they put a cap on pain and suffering and you can tease out the kind of what the effects of those changes in the law are.  Data have blurry in small effects on what the apparent amount of defensive medicine is.  It's a very tough topic to unravel.  The best economist in the country has worked this for a long time.  Ms. Patricia Danzon at Wharton School, I have talked about this extensively to Patricia.  She has written a book about it.  I don't think she or any one else has a really clear picture of the extent of this.  It's only 8 percent of physicians cost structures, their malpractice insurance premiums, but that of course doesn't tell you about the defensive medicine side where there's change in their behavior and response.  It's a complicated problem and I wish I could give you a clearer answer sir but I just can't.
>> I had a question at the back also.  Yes?  Okay, so let me deal with the first one.  We actually have some pretty strong evidence on the magnitude of this.  When Medicare started paying hospitals on the lump sum per admission, it took place I think in 1983 and then a phase in over several years, there was this big concern that patients are gonna be discharged and the language was sicker and quicker, which I think was the issue to which you are alluding.  First of all, Medicare by the way won't pay them if they bounce back in right away.  There is a time log there.  The patients gotta come back in, is it 30 or 90 days now?  I can't remember what--do you remember what the number is?  I think it's 90 days delay.  If they come back in within 90 days, the hospital doesn't get paid the second time.
[ Inaudible Remark ]
>> Right, now, but the--the second point is that a least measures the mortality in readmission rates didn't change in any noticeable way except for some various select diagnosis were you're dealing with small samples.  A very careful study is done in this around the nation.  The readmission rates and mortality rates didn't change any when Medicare instituted this payment scheme that gave the incentives to discharge quicker.  Now, length of stay have been dropping years before in a dropped--continues to drop since then in ways that are really quite remarkable.  I think much of this is due to technology, sometimes due to putting more people on the system to make sure that there's out of hospital support.  We can all point to some horror stories about somebody that was discharged too early.  The systematic data just don't show any meaningful effect when Medicare moved to the prospective payment system in terms of certain at least in terms of mortality and readmission rates.
[ Inaudible Remark ]
>> Yeah.  Well, if they are discharging them, the sooner we can get them out the hospital.  They would have gone up if they were making mistakes about that to get them out of the hospital and save treatment cost.  And they didn't.  Yes.
>> I became aware of this when I was working in Columbus on the state level.  We were trying to get a home health care instituted to more people and, you know, the nursing home lobby is so strong.  Can you comment on the amount of Medicaid dollars that end up putting patients in nursing homes?
[ Inaudible Remark ]
>> I can't--I can't comment knowledgeably.  Let me--
>> I think it's extremely.
>> Let me put that down to one of the many things I never learn while I was provost.
>> No, I don't mean the exact dollar.
>> Yeah.  I just, I'm not very knowledgeable about that part of the world.  I probably stick my foot in my mouth if I try to comment on that.  Yes sir.
>> Alright, [inaudible] there's this study that there's no real difference between performance on fee-for-service or flat salary.  There is no real difference in other of these factors we described.  The main difference seems to be--this other guy's question didn't quite answer the amount of money that goes into the insurance system and wouldn't the system overall be much better off.  As a single payer, it did not have is [inaudible] 35 percent plus overhead for them to shove the papers around [inaudible] service.  It doesn't seem to be any significant difference in requirements between the variety of the service.
>> Okay, so let me talk about single payer a minute 'cause I know there's probably a lot of single payer fans on this room.  The big comparison is like with the Canadian health care system where much smaller fraction of the total healthcare business is spent on administrative cost.  And part of that, I think probably an important part of that--first of all, I have to [inaudible] that administrative cost is actually doing something for us.  And what is--it's not just processing claims.  What they're doing for us is doing comparison shopping to find lower price doctors.  And this is particularly true of managed care organizations preferred provider organizations.  They go out, they collect a bunch of doctors in hospitals and they bargain like crazy with them to get the price down, that they pay and hence you pay through your insurance plan and then they tell the insurance company, "Oh by the way, I have 200,000 patients in the Ann Arbor area.  You can--I can move them to you if you'll give me this good price.  And so, a lot of what they're doing is basically doing price bargaining force that we can't do on our own and that's valuable service.  Now, if we have an administered price with a uniform single payer, it's gotta be a federal government to do that.  You're running then into a question that's more political preferences of people in the United States versus elsewhere about how much freedom of choice they want.  The US, this is really out of my territory but the US is in some sense, I think much more dedicated to having free choice than many other nations particular European nations and hence we're paying the cost of that free choice with this administrative cost.  It's not perfect either way.  The big concern I have about single payer comes in technology introduction.  Now I've talked about all these new technology that had come in the door over the last half century.  Think about what happens if there's one single gateway to get that technology into out health care system.  Let's just say its Medicare.  They make these choices for Medicare patients.  We have, in addition to that in the United States we have hundreds and hundreds of manage care organizations and insurance plans.  They're trying to decide whether they wanna cover some new medical technology or some new surgical technique or not, and they can actually do a lot of experiments for about how much these things are benefiting patients.  And so some of them don't cover and some do things that are called experimental techniques and we learn from those.  If you only have a single gateway into that system you've lost those data.  You've lost that information, and one of the things that makes me particularly nervous about single payer models is the fact that there are not multiple pathways to get evidence in about whether the things or not.  There's one decision maker in a single payer system it just has to be that way.  You can have the single payer system. And have 20 people deciding whether the autologous bone marrow transplants are gonna be covered or not.  So that's nervous time for me on that.  There're pluses and minuses in both system.  That's one of the minuses that stick in my mind.  Okay, I'm getting the signal that says we're wrapping up soon.  I'll try one yes sir?
>> You started on a theme about prevention and you're wrapping up so let me ask this question right now.  There are some employers who benefited or catalyzed employees who don't take care of themselves.  They don't smoke if they do smoke.  If they don't exercise or if they do exercise, it shows up in pay check.  I like your comment about and also how far do we go there?  Do we go termination, taking care of yourself or you're terminated.
>> Well, certainly first of all, there's not nearly as much of it going on in the field of employee health or employer paid health insurance as there is in other types of insurance like, you know, life insurance and stuff, were much less experience rating in it.  Most of that is because the employers are very heavily limited by other federal laws, employee laws, ERISA and HIPAA about what kind of information they can collect about employees.  Now you can rate employees and change their performance, their salary or their hourly on performance outcomes.  But it's very hard to do that on behaviors themselves because of various laws.  So I guess I would rather see more of that going on so that people with bearing the full economic consequences of their health rates.  I'm saying that people shouldn't smoke.  I'm just saying I'd rather not pay for it being sharing the same health insurance policy with them and the same with obesity, and of course I would probably be paying above average in obesity.  My waste line is bugger than it should be and I'll be the first to admit and thank you for all you thin people for helping to subsidize my medicare premium just because I don't pay anymore for Medicare either.  I think one of the things that I'd like to see happen in this country is to see a tighter length both but particular in health insurance premiums both private plans and in Medicare, a tighter link between behavioral choice outcomes.  That is not whether you're trying to quit smoking or not but do you smoke or not?  Now that of course brings me full circle to kind of a philosophical and moral problem, you're asking, how can economists talk about more problem?  Some of these things including the propensity to be addicted to nicotine and the propensity to gain weight have genetic underlying, and hence if there's a genetic predisposition here it makes me nervous saying, I wanna go all the way to have full experience rating on this things 'cause part of these, there's a risk I cant control if genetically linked.  And I just, I don't know how important those are, but it comes back to the question about keeping employers or keeping insurance companies from using previously existing conditions.  These genetic illnesses are one of the big things on that and that's actually one of the reasons why I think there's a pretty good case for getting rid of the ability to use the previously existing conditions.  Because in fact what that, what this does now with this new law has opened up the market for ensuring against genetic time bombs that I can't do otherwise.  So there's several reasons why that's a good idea.  The other is job luck people that are getting their insurance to employers can't change jobs easily when they know they're gonna lose their insurance coverage for 6 months even under the HIPAA laws, that's gonna go away in 2014 under that Affordable Care Act.  And there's a comparable entrepreneurship act.  People are afraid to go out and start up small businesses 'cause they'll lose their health insurance coverage.  So I think there's a pretty decent case to be made for the economic sanity of eliminating the privilege, the existing conditions.  It's very popular in surveys of the US populations.  But its also comes the background of the question, how much you wanna tax people on their health insurance premium for smoking when in fact you know that's smoking is at least some how related to genetic predisposition?  And then one of the grand ironies of life, the very same gene that makes you susceptible in nicotine addiction also increases the risk that you'll get lung cancer if you smoke.  So let me leave it with that and say, thank very much for your attention.
[ Applause ]