This raises the question, could we identify these individuals and prevent in some way, either their entry into these expensive episodes or lower the utilization given that you might have an entry.

One issue we addressed in our report-and I won't go into itis sort of whether you could just look at them on the basis of their demography and say these are likely to be the high cost folks. The answer is pretty much no. Although they are a bit older, they don't stand out in any other particular way.

If you look at their health, however, a key feature is the presence of chronic conditions, particularly multiple chronic conditions, where compared to the typical population, 75 percent have one or more chronic conditions versus about 40 percent in the rest of the population. About half of them have two or more for sure.

So that does stand out. So that becomes one of a series of illustrative strategies that we used in the report to see if we could identify high-cost Medicare beneficiaries. That is the final slide, where we took three that we thought of as stylized strategies that one might undertake to pick out who is going to be expensive in the future. Take a person who has multiple chronic conditions and then see how they turn out. Look at someone who has had a hospital admission and then track them. Or look at someone who is simply very expensive in the beginning year and see if they continue to be expensive in the years thereafter.

What the slide shows is a comparison of those groups versus a random sample of Medicare beneficiaries. We look at them in initial year, 1997; identify them using one of these strategies; and then see if we could predict that they would be more costly in the years to come on the basis of that identification.

Indeed, to some extent, this appears to be the case. It is suggestive that this kind of strategy might be successful in identifying high-cost beneficiaries.

Compared to the control group, each has greater spending_certainly in the base year, but also in subsequent years. For those who get admitted to the hospital or who are expensive, you see a bigger drop off. For those who have the chronic conditions, their spending drops off less. It tends to stay elevated in the years thereafter.

Now, the final question, of course, is whether this would allow the Medicare program to somehow control their costs in the future, and there it raises the hope that something like a disease management program might be successful in reducing overall costs. We can come back to this in the discussion later, but I think that the things that I would note at this point are that disease management means different things to different people. There is a variety of different elements of either education or patient monitoring and, thus, practice, or care coordination, or case management. So exactly what goes into disease management is not always the same. It is worth investigating that.

Asking whether it works is really a question of first comprehensively measuring costs over the entire future of a patient's experience and comparing that to a comparable patient without the disease management. That is a high scientific standard. None of the work that we have examined to date meets exactly that standard

and at each point stepping down the standards, you have to ask whether we have got the evidence we need.

Then finally, even if this strategy works, the important issue for this committee is a tradeoff in costs. It may be the case that some sort of preventive disease management program will work for Medicare beneficiaries-in the sense that it will lower costs other than what they would have been-but it will be costly to identify the people who enter into such a program out of large population of seniors. The question is whether it is cost effective in both senses. You may spend so much finding the folks that will ultimately benefit from disease management that you overwhelm any cost saving you would get from putting them in the program.

Those are the two elements of the decision, and that is the difficult design issue that would face someone trying to put this into place in the Medicare population as a whole.

So we are pleased to be here. That is the high speed overview of the report. I will be happy to answer your questions and pursue it any way you like. Thank you.

Senator KOHL. Well, thank you. I am curious with respect to your opinion on the following thought: are there people who have some chronic conditions who use the system—and we are talking about them now-and to a great extent those are the ones whothe 25 percent who cost us 85 to 90 percent of Medicare, but others who are seniors who have similar conditions who just do not check in that often, use the system that much, manage to deal with these problems in a way that doesn't require them to be so involved with Medicare?

Mr. HOLTZ-EAKIN. There are certainly those who would have one of our list of seven chronic conditions. Diabetes stands out. Among the high cost beneficiaries are those with diabetes. However, if you look in the low-cost population, there are lots of folks with diabetes as well, three times as many, in fact. So it is not the case that if you are diabetic, you are automatically high cost, and it is not the case that if you have one of our chronic conditions, you alwaysyou inevitably-end up there. They are in both populations. This goes to the last point I made, which is that you have to be able to find the diabetic who will benefit from some sort of intervention to lower costs.

Senator KOHL. But is it true that there may be two similar people who are seniors who have conditions that are not entirely dissimilar?

Mr. HOLTZ-EAKIN. Oh, yes.

Senator KOHL. One will access the system an awful lot and prove costly in a dollar and sense way. The other one will access the system an awful lot less and be less costly, just because they are a different kind of individuals.

Mr. HOLTZ-EAKIN. Certainly, and we could probably go into the data that we used for this report and find people with chronic conditions and show you the averages on both sides of that observation.

Senator KOHL. All right. Thank you. Senator Smith.

The CHAIRMAN. Doug, I am interested in whether or not you all have factored in the impact of Part D, and what it might do to Part A expenditures?

Mr. HOLTZ-EAKIN. It is not the first time this has come up, which is not surprising. We certainly have tried to look very closely at the degree to which additional therapies in the form of pharmaceuticals might lower costs elsewhere. But it is hard to get that out of the data for a variety of reasons.

No. 1, the Part D really covered the costs of pharmaceuticals. People were taking the drugs they needed anyway in many cases, so you haven't really changed their therapy in any deep way. So you wouldn't expect a change in the costs. So that is sort of the major reason.

The CHAIRMAN. OK. I understood in your testimony that where there is simply private coverage and Medicare is not involved, these same populations are still using those kind of resources? Mr. HOLTZ-EAKIN. Yes.

The CHAIRMAN. So probably not the savings we might hope for? Mr. HOLTZ-EAKIN. No.

The CHAIRMAN. OK. Do you believe there is any benefit to comparing data from Medicare managed care plans that employ chronic disease management programs with the data you have compiled for the fee-for-service programs? Are the Mr. Holtz-Eakin. It is hard to imagine that it wouldn't be valuable to compare them as long as you were careful about the comparisons. You know the key issue is what constitutes the same kind of group going in, and given that the people who chose to go into the managed care versus the fee-for-service do so voluntarily, they are, by definition, not identical. They have chosen differently, and so you have to somehow get a handle on that before you start doing comparisons across the groups.

Senator KOHL. Senator Wyden.

Senator WYDEN. Thank you, Mr. Chairman, and I want to thank Dr. Holtz-Eakin for excellent testimony.

I am curious what CBO has in terms of numbers as it relates to spending on health care in the last 6 months of an individual's life. You know there are constantly studies, you know, thrown around on this point, and I am wondering, you know, what, if anything, CBO uses as statistical documentation on that point?

Mr. HOLTZ-EAKIN. We rely on the Medicare claims data, so it would be among those folks. For the numbers I have for this hearing, we can try to see if there is more detail in the last 6 months or for the last year. Twenty-five percent of Medicare spending is in the last year of life ballpark. So it is a fairly substantial sum.

It is, of course, one of those backward looking computations in that you don't know when the last year of life will be necessarily. But looking back, those are the facts.

Senator WYDEN. That will be an area I want to follow up with you on as well for the Citizens' Health Care Working Group because those issues, of course, were tough before the Terry Schiavo case. They are now infinitely harder and my hope is that we can find some common ground. Senator Smith and I have introduced bipartisan legislation, the Conquering Pain Act, to try to create some options for folks, but we will be anxious to work with you on that.

I wanted to also explore with you a topic you and I have talked about. Senator Sununu and I have been concerned about the fact

that public programs, programs like Medicaid, the Public Health Service, the VA, are paying for prescription costs, you know, advertising. In effect, those programs end up getting shellacked, you know, twice. There are tax breaks for the pharmaceutical folks to advertise on TV. Nobody is quarreling with that, trying to take it away. But after that expenditure is made with taxpayer money, then more money gets spent for in effect like Medicaid to pay for all those purple pills, you know, dancing across everybody's television set. So we are trying to address this issue and obviously advertising increases utilization of prescription drugs and, of course, the program.

Let me ask it this way: The official sources on drug advertising seems to be that the country spends between $3 billion and $5 billion a year on prescription drug advertising. According to the bipartisan experts, after the Medicare drug benefit kicks in, Medicaid is expected to be about 10 percent of the prescription drug market. That seems to be a kind of consensus recommendation.

So Senator Sununu and I are interested and working on the language of this and would very much like your counsel so as to focus on utilization and focus on market share. It is our sense that if we do that, the government could save about $300 million to $500 million a year on Medicaid, in effect over a billion dollars over a 5year period.

Do you feel that that is essentially a reasonable kind of analysis? Mr. HOLTZ-EAKIN. Yes, given that the language was tight enough, that it could find a way to actually recoup the costs, and that we can, you know, get a sense that the numbers are on the mark. They certainly seem reasonable. Yes.

Senator WYDEN. Well, I appreciate that, and I would like to work with you on the language because I know that the way it is framed so as to focus on utilization and market share is really, really key, and if we could follow up with your technical folks. They have been very helpful to us already. This is a bipartisan bill, and I just point it out because we have Chairman Smith here, and he has done excellent work on the Medicaid program. He is trying to get $10 billion worth of savings without hurting people on Medicaid, and I would just like to make it clear for the record that Dr. Holtz-Eakin has said we could get more than a 10 percent of the savings in the target that Chairman Smith is looking at by the advertising provisions along the lines of what Senator Sununu and I have been talking about. So we will be anxious to follow up with you, and we got to figure out how to save $10 billion on Medicaid, and we all want to do it without hurting people. We just on the record a way to in the ballpark to get 10 percent of the money. That is what we ought to be trying to do is sharpen our pencils.

Chairman Kohl, I thank you for this, and Dr. Holtz-Eakin for all his analysis.

Mr. HOLTZ-EAKIN. Thank you.

Senator KOHL. Thank you, Senator Wyden. We also have with us this morning Senator Blanche Lincoln from Arkansas. Senator Lincoln.

Senator LINCOLN. Thank you. A special thanks to Senator Smith and Senator Kohl. They have been tremendous leaders in the Aging Committee, helping us focus on the important issues that

face this country, both financially as well as for all us emotionally because one of these days we are all going to be old. We are all aging, and we are grateful to both of you.

Mr. Holtz-Eakin, we should have you as an honorary member of the committee. We have heard from you a great deal, and we certainly appreciate all the work that you at CBO have done in helping us realize that we can do a better job in administering these programs, particularly for these high-cost beneficiaries.

I would urge you to take a look at legislation I have been working on as well, S. 40, and would appreciate getting any help with scoring it. I would love to work with CBO on a way to ensure that a new Medicare benefit for geriatric assessment and chronic care management of individuals with multiple chronic conditions would save money to the program. I know in my own personal experience with my father who went through a long period with Alzheimer's, Disease with other diagnoses, I saw how important it was to have coordination of all the medical professionals, in treating his multiple chronic diseases. Fortunately for us in Arkansas, we have the Don Reynolds Center on Aging, which focuses on patients with multiple chronic conditions and management of chronic illnesses, which makes all the difference in the world. My constituents see a difference when they go from visiting six or seven different health care providers to a care team that manages all of these chronic diseases together.

You said in your report that reducing spending among the highcost beneficiaries would ultimately rest on the ability to devise and implement effective intervention strategies, clinical or otherwise, to change beneficiary use of medical services. I think that by giving an individual a geriatric assessment, which assesses a person's medical condition, functional and cognitive capacity, primary caregiver needs, and environmental and psycho-social needs would go really a long way toward reducing some of the unnecessary and expensive medical services.

I just wanted to see what you thought about that in terms of the research that you have done. Would that assessment be beneficial and could it be helpful to us in saving financial resources?

Mr. HOLTZ-EAKIN. It is on the list of appealing strategies that comes up all the time, and in that regard it always falls to me to throw a little cold water on some of the hopes. The first is that in many cases you could not see lower costs, but it would still be worth it. You know, you are paying more and people have better health for longer periods and function better in their lives. That is not a cost saving issue, but it is still a good step.

Then the second caveat I am compelled to offer is that there isn't any systematic evidence to date that we can, in any broad way, get a lot of savings out of the Medicare population from this. That doesn't mean that it isn't true. It means that, to the extent that researchers have gone and looked at to the best of their ability groups with and without these kinds of checkups or other services, you can't find a compelling scientific case that the costs are lower for the group where you have undertaken the new treatments. There are lots of reasons why that might be the case, and I would be happy to work with you on that.