February 3, 2003
The Right Prescription for Medicare Drug Coverage
Dr. Robert Goldberg
Today's discussion is about what the right prescription for a Medicare drug benefit would be. It's timing is impeccable because the administration is going to release its budget numbers, and it's estimated that as part of the President's proposed reform package, he is going to spend, over the next 10 years, $400 billion on providing prescription drugs to senior citizens.
We also have to remind ourselves that over the next decade, and beyond, Medicare itself, apart from Social Security, will claim a larger share of the tax burden. Indeed, there are some estimates, according to the actuaries that look at the finances of Medicare that up to 30 percent of payroll will be taxed just to pay for the existing package of benefits. That's not including prescription drugs.
So the issue of how to control the costs of the Medicare plan is always entwined with the issue of prescription drug coverage. Unfortunately, cost containment always seems to focus on the new "new" thing, on the drugs themselves in a vacuum, without respect to what impact the administrative mechanisms used to control those drug costs have on total health care spending or the total well—being of the patient.
The old adage "an ounce of prevention is worth a pound of cure" is sometimes lost in the shuffle in this debate. Prescription drugs are only 10 percent of our total health care spending. That's how much they were in 1965. At that time, of course, there was very little that physicians could do for patients suffering from cancer, suffering from heart disease, suffering from arthritis and ulcers. Now with that 10 percent we can control AIDS, we can begin to control Alzheimer's and a variety of other fatal diseases. We are reducing the death rate of many chronic illnesses. According to my colleague at Columbia University, Frank Lichtenberg, for every dollar that we spend in new drug consumption, we save $6 in other health care costs.
So, as we approach the discussion of the right way to provide prescription drug coverage, we have to look at whether or not we should be focusing on drugs alone or the total picture. Our first panel is uniquely qualified to discuss that particular question.
Susan Horn has done path—breaking work on the relationship between access to drugs, drug restrictions and total health care spending and well—being. Susan is Senior Scientist at the Institute for Clinical Outcomes Research. She was at the University of Utah and at Johns Hopkins doing path—breaking work on controlling the outcomes for severity of illness.
John Graham of the Fraser Institute, who runs the Pharmaceutical Policy Research Institute there, is going to provide his latest report on what impact reference pricing had on total health care costs, pharmaceutical costs, and patients in British Columbia. And Dr. Ralph Hawkins, who practiced medicine in British Columbia and is now practicing at the University of Tennessee, will talk both about what happened in Canada and the latest National Institutes of Health (NIH) study, which argued that everyone with hypertension should receive the cheapest drug in the class. The latter is especially important since there's some movement afoot in Congress to have the NIH begin to become the drug prescriber of choice through these big studies.
The second panel, which will include myself, Grace Marie Turner and Ron Pollack of Families USA, will talk about whether the private sector or the public sector is the right context for delivering a prescription drug benefit.
Dr. Susan Horn
By background, I'm a health services researcher. My original career was to measure how sick patients were, control for that, and then look at their treatments and outcomes to discover what works best for specific types of patients. It's in that context that we have done several studies relating to prescription drugs, because in almost every clinical area medications are used. We've examined what happened when different types of medication were used, and that is what I'll be sharing with you today.
The study designs for every one of the studies whose results I'm going to share are based on a lot of controlling information. We were, of course, focusing on outcomes, but we adjusted multiple ways for how sick the patients were, and we also looked at various medication uses and particular management strategies of medication.
The first study was published in 1996. It's based on data from the earlier part of the 1990s called, "The Managed Care Outcomes Project," where we were looking at the consequences of a whole variety of cost—containment strategies used by managed care. The main study question was when one looks across multiple managed care organizations, at a year's worth of actual data on the care of thousands of typical patients treated by their regular doctors, how was the amount of health care services used associated with their cost containment efforts? We all hear about, and probably even experience, the cost containment practices of managed care, but to what extent do they actually work?
The key cost containment practice for today's discussion is the limitation on reimbursable drugs. The concept is called a formulary, where people within the managed care organization decide which drugs can be prescribed without going through prior authorization; in other words, you can prescribe the drug, the patient can get it, and you don't have to do anything more. Prior authorization means before you can prescribe a particular drug, you have to go through an explanation process and get permission to have that drug covered.
We decided to look at five very common ambulatory diseases. In fact, based on our national databases, about 70 percent of ambulatory visits will be associated with one of these diseases. There were ear infection patients, who, of course, were mostly younger people; asthma patients, who were both younger and older; and then arthritis, hypertension and stomach ulcers patients, which were mostly older. We had six HMOs in different parts of the United States work with us on this project; three eastern, three western, half not—for—profit, and half for—profit. Each HMO site followed 500 patients with each of those five study diseases for a total sample of 2,500.
In total, we had complete data on 13,000 patients that had over 99,000 office visits over the year and used over 240,000 prescriptions, quite a large study. So this gives us a pretty good glimpse about what happens in real life to patients over the course of a year when certain practices are instituted.
We controlled for many different variables in this analysis, all simultaneously, so we could see the impact of each when everything else is considered. In terms of patient variables, we used our severity—of—illness measure to make sure that we had controlled for how sick patients were. Other measures, such as age and gender, number of months in the study, and number of different physicians seen were all controlled for as well.
Also there were variables that I've called HMO site variables, that we didn't consider cost—containment practices. And finally, there were all the different cost containment practices that we examined, such as how strict were HMOs' second—opinion requirements, how strict were their gatekeeper requirements, their case management, their drug co—payments, their visit co—payments, their restrictions on their formularies, and finally the extent of use of generic drugs.
Most of the cost—containment practices turned out pretty much as expected with one major exception, and that was the limitation on drugs. Our thought process was that if you had fewer drugs on a formulary, you could get better volume discounts; hence, lower costs per pill. We assumed that should save money on the drug costs, but should have no impact on the office visits, and hospitalizations, and emergency room visits because the drugs were really all very similar to each other. So why not use the less expensive ones?
Yet when we looked in the actual practice of care, every one of our utilization variables was going in the opposite direction from what we had thought it would. For example, if the Food and Drug Administration approved 90 different drugs to treat asthma, what percentage of them could a physician prescribe in that managed care organization without prior authorization? And as that became a smaller and smaller percentage, did that have any effect on the outcomes?
What the data kept showing is that the smaller the number of available drugs compared to the number the Food and Drug Administration had already approved, the greater the number of patient visits to physicians, emergency room visits, hospitalizations, estimated costs of prescriptions over the year, no matter what kinds of discounts we used, and also greater numbers of prescriptions over the year.
We first thought that maybe although the patients were using more prescriptions the more restrictive the HMO was, it might have no impact on the cost since they were using less expensive prescriptions. But, across the various diseases, the total cost of prescriptions rose over the course of the year along with the number of prescriptions. Whether looking at asthma or arthritis we still see lower costs when the HMOs don't restrict, and the more they restrict, the higher their total costs. Not only was the pattern consistent across diseases, it also applied to hospital visits as well as prescriptions. As fewer drugs were available without prior authorization, visits by the patients increased.
We also went on, after looking at the whole population, to see whether there was a possibility that elderly patients, who of course are the focus of Medicare, could be even more susceptible to problems related to limiting access to drugs, given that there are often physiologic differences in elderly patients, they can be taking multiple prescriptions, they absorb drugs differently, and so on. We found in that study that this varied depending on the type of drug we looked at.
So we concluded that sometimes the elderly were at greater risk than younger people and other times they were at equal risk, but both groups had this problem of limiting access being associated with higher utilization, not the same or lower utilization.
We also took that same 13,000—patient database and looked at limiting access to mental health services and psychiatric drugs. Although that wasn't one of our initial five study diseases, we were able to analyze this as a subset because we had measured comprehensively all of the severity—of—illness factors about patients and all of their utilization over the whole year. I was a bit surprised to learn about a third of our study population, whether they were elderly or non—elderly, had either a psychiatric diagnosis or were taking a psychotherapeutic medication, in addition to the medical illness that made them part of our initial study.
So, in our managed care population, about a third of the population has a psychiatric co—morbidity. Not very many of those people, even when they were diagnosed with depression or on antidepressant medication, saw a mental health provider. That's because, in many managed care organizations, that policy is also somewhat restrictive. They saw a lot of primary care providers, they saw other medical specialties, particularly the elderly and other surgical specialties, but the lowest utilization in terms of a specialty was in the mental health area.
In terms of the types of drugs that were used, because they were the newest at the time, the SSRIs were usually not on formularies, or only one would be on a formulary. Their utilization was considerably lower compared to the older tricyclics or benzodiazepines.
The person who was leading this analysis of the effect of this was Dr. Steve Bartels, who is a psychiatrist and a professor at Dartmouth. And he said to me, "Susan, I've been seeing what's going on in managed care. They limit access to mental health services and psychiatric drugs, and they think they're saving money. Could we possibly look at this issue within your population, now that we see these kinds of restrictions, and see what happens to the people who are not getting the optimal treatment with regard to drugs or mental health services."
And what we found is that patients spent about 50 percent more on non-psychiatric medications when they weren't being given the services they need to deal with their psychiatric problem. Similarly, they had about 50 percent more non—psychiatric visits. They had more emergency room visits and hospitalizations. Their medications cost about 50 percent more. In particular, we found that people who were using antidepressant drugs and were switching, where people assume they'll be started on older drugs and kept on them unless they don't respond well as opposed to using only new medications, had almost the same higher rate of visits as those on no psychiatric medications at all.
So no matter how we looked at this, we kept finding you need to give the services that will get people better as quickly as possible, and often requires both services like mental health services, as well as newer medicines, because newer medicines often work better, with the result that the total cost of care will be reduced.
That was the finding of our managed care outcomes project. Every time we've looked at it, we keep finding if we limit access, don't give the freedom to prescribe the best drug to get the patient better as quickly as possible, we end up with increased resources, rather than decreased resources.
Some people have tried this out. One study was done at the Lovelace Health System in New Mexico, where they tested the theory that high utilizers are likely to have psychiatric comorbidities by giving those patients the Zung self—reporting questionnaire to screen for depression. They found 37 percent of those patients with high utilization screened positive for depression, and they changed their whole practice to promote the use of the newer drugs, SSRIs, rather than restricting them.
Before their depression diagnosis, the combined 12—month medical costs of the 2,000 patients in that 37% was $13 million. After they were diagnosed and treated, their total costs were only $11 million the next year. In other words, when they gave treatments that got patients better, even though their psychiatric costs went up and their psychiatric drug costs and their psychiatric visits went up, their total costs went down because people were not coming back trying to get other drugs and other visits to control the fact that they weren't feeling well.
Some have argued that even if the newer drugs did work better, they could still limit access to only some of them. But a separate study by a different group of researchers shows that no single SSRI worked for all patients and all forms of depression, and if you limit prescriptions to one SSRI, about 25 percent of the patients will not respond to that agent. Those patients who had to then switch SSRIs were in treatment 50 percent longer, and cost approximately 50 percent more. So a care provider was actually much better off having several antidepressant agents allowed so that there are more options to continue treatment.
Several years later in 1999, an even larger study than our original one, and one specifically related to the elderly only, was published by researchers at the School of Pharmacy at the University of Southern California. They looked at almost 23,000 HMO Medicare enrollees who were treated at three multispecialty clinics. Essentially, this was a natural experiment when an HMO got all of these enrollees, went to the physical location of where the people lived and asked physicians in that community to contract with the HMO to treat them.
There were three big groups that the HMO contracted with, and they had a contract that limited access to medications. They had a drug capitation. Two of the medical specialty groups agreed to that contract. The third refused to restrict access to drugs, but because they couldn't find another medical specialty group in the community where these enrollees were, the HMO agreed to contract with that third group without the limitation on drugs. However, they were terrified that the costs were going to be astronomical in that third group.
When the data were analyzed a year later (and because the HMO didn't believe they had done the analysis correctly, they asked the University of Southern California School of Pharmacy researchers to redo it) it turned out that, controlling for multiple different confounding variables, patients with drug capitation ended up having 14 percent higher total health care costs over the year and 29 percent higher pharmaceutical costs.
The conclusions were that curtailing access to medications via cost control mechanisms can adversely affect other health care costs and increase health care utilization, and that any individual capitation of health care components may not be the best way to approach controlling costs and assuring quality of care.
So we now have multiple organizations and studies that are pointing out that, unfortunately, there is no simple way to cut costs. Curtailing access to medications via any of these cost—control mechanisms ends up having adverse effects on health care utilization; more office visits, more emergency room visits and more use of medications overall.
In fact, one researcher recently has said the rising costs of drugs may actually be part of the solution, not part of the problem. But they can only be part of the solution to control spending if reimbursement mechanisms are based on sound clinical research and prescriptions are targeted specifically to those patients who could reap health benefits from them greater than the increased cost.
In closing, I want to discuss two different studies that show that while open access is better than simplistic limitations, there is a third option that can effectively reduce costs. The first is a study done in long—term care, particularly related to the elderly, where we followed the care of 2,500 residents.
There is a national guideline on what to do with patients with dementia who make up a large proportion of patients who are in long—term care today, and particularly those with agitation who are the most difficult group to deal with. The guideline says: Use the fewest number of medications possible. We don't want chemical restraints on these people, so you must document behaviors before you give them anything. Then, if you're going to give them something, limit the use of benzodiazepines. If you're going to give them an antipsychotic, use newer ones. If you're going to give them an antidepressant, use an SSRI, and definitely avoid combination therapy because everybody knows polypharmacy is bad.
Well, we had data from these residents, and we thought let's see what happens when we look in the actual practice of care. What we found was 800 of those 2,500 residents actually had signs of agitation, along with their dementia throughout the year. About a third of them got no medications at all in actual care. When we went to some of those long—term care sites and asked why, they said, "We don't have the nurse staffing to document the behaviors, so the easiest thing for us is to just not give them anything."
Now, because the guideline doesn't really give any general guidance as to what to do for which kind of patient, about a third of those patients who got anything got antipsychotics, another third got antidepressants, and the other third got antianxiety agents. And it turned out of those who got something, 42 percent got combination therapy. So this was a wonderful database to see how bad that treatment decision was.
And what we found was, to everybody's surprise, combination therapy, and particularly the combination of an SSRI plus an antianxiety agent, was associated with significantly less hospitalization and emergency room use, significantly less urinary incontinence, significantly less pressure ulcer development, and significantly less physical restraint use. And giving no medication was absolutely the worst treatment.
We would have never known this unless we had data from the actual practice of care because there's no mechanism to study combination therapy in any kind of a randomized trial, and they couldn't have done a randomized trial on this kind of a population. So what we're learning is that we can discover what's best for specific types of patients, and thereby reduce costs as well. But we need data to do that, and the idea of just putting together a guideline or restricting access has not worked.
The final study I want to share with you is an antibiotic study where they switched from the process of using a formulary, which they used starting in 1988, to using a patient—level algorithm, rather than a disease— or drug—level one, in 1994. The algorithm would recommend, based on the patient signs and symptoms and characteristics, what was the best drug to use for an antibiotic for that particular patient. Many more people got antibiotics in 1994 in this setting — double the proportion of broad spectrum, more expensive, newer antibiotics, and a sicker population — but the total cost of antibiotics, in terms of acquisition cost, went down by more than $350,000.
All costs didn't decrease in this situation because they only had the algorithm for antibiotics, which went from 25 percent of the total pharmacy budget to 13 percent. Because the algorithm provided not only what drug should be used, but a complete picture of what you should do in delivering that drug, from dose to duration, the actual cost per treated patient went down by $70. Patients who needed prophylactic antibiotics got them almost 100 percent of the time but, again because of the algorithm, fewer doses were given, adverse events went down by 30 percent, and mortality went down by almost a full percentage point. We can do a better job than nothing, but we need patient—level algorithms to make those decisions, allowing the data to tell us for specific patient characteristics, what treatments are associated with better outcomes.
If we do that, as we have done in about 30 studies over the last few years, we have found a 30— to 50—percent decrease in the cost of treating that particular patient group. There is an enormous potential to save money. Unfortunately, it's not as simple as saying you can use this but you can't use that to treat a particular disease. It's more complicated than that. But it's doable.
John R. Graham
I'm going to assume ignorance of the Canadian health care system, so I'll just briefly introduce the Canadian system, specifically as it relates to prescription drugs, and talk about a couple of prescription drug issues that are of concern across the border.
One is the international price difference and the question of reimporting, which is an acute issue now because GlaxoSmithKline has decided to no longer supply Canadian wholesalers and pharmacies through mail order or ship to the United States.
Then, I'm going to explain reference pricing, which is something similar to a preferred drug list, which is being discussed in many, many states, and at the federal level, and it's one of the things at issue when talking about a Medicare drug benefit in this country.
Canada is a federal country of 10 provinces. Provinces are similar to states in your country, and provincial governments are monopoly health insurers. Neither doctors nor patients can effectively opt out of the government monopoly. Legally, you can, but you can't really; like you can't bill a different rate than on the fee schedule that's centrally bargained by the medical association and the provincial government. That's one feature of the Canadian health care system that you're probably broadly familiar with.
Hospitals are completely funded by governments. Capital costs are funded by governments, so while the hospitals are nominally privately owned, they don't raise their own capital. It comes from a block budget from the government. This is important to know, because there is a cardiologist in Canada named Dr. Peter Devereaux who's done a couple of articles comparing American private for—profit and American not-for—profit hospitals and tried to use those to promote further government intrusion in the Canadian health care system. Because of the functionally public nature of all Canadian hospitals, it's completely inappropriate to compare one to a private American hospital.
Private clinics are heavily restricted. Some clinics that do day surgery are allowed to operate, like abortion clinics or those doing cataract surgery, but generally speaking you cannot just open up a private clinic and do whatever kind of surgery you want. That's pretty much forbidden. So that is the Canadian health care system writ large.
Now as we look at prescription drugs, it's a little different. Just as America doesn't have a Medicare drug benefit, in our country prescription drugs are not completely under government control. Each province has its own prescription drug benefit plan, and they differ in significant ways. Nevertheless, provincial governments pay 42 percent of the entire prescription drug bill. So we have a much greater level of government intervention in the prescription drug market than you do in the U.S., where government expenditure is about 15 percent of total prescription drug costs. Federal government spending in Canada is very small. Then there's the Social Insurance and Workmen's Compensation Board, what I just call Social Insurance. Social Insurance is a specific program in the province of Quebec that's fully funded through compelling everyone to pay a premium, but it's a discrete premium. It's not funded through general taxation.
Private insurance is about one—third of the prescription drug market and about one—fifth of the market is paid out of pocket. So there is still a significant private market, and that's going to become important as we talk about reference pricing.
I'm going to quickly go over the Canada versus U.S. prices. It's one of the things I've written quite a lot about for the Fraser Institute, Canada's leading market-oriented think tank. Why is the price of prescription drugs so much cheaper in Canada than in the United States? The first answer is that the price of everything you buy in Canada is cheaper than it is in the United States. It's sort of a natural law of the international order.
Basically any good you look at will have a huge price difference. Take cars for example. Automakers, when they ship their cars to Canada, get the dealers to sign contractual undertakings that they will not ship those cars back to the United States because they sell for a much lower price in Canada than the United States. And during the last dozen years or so, that general price difference has gotten a lot bigger primarily because of the Canadian dollar, which has gone in the tank but without generating inflation to adjust for that. We're not a very productive country, and the average Canadian worker is now earning about a third less than the average American worker. So, basically, we couldn't tolerate the kind of prices Americans pay for things. It's not a conspiracy by the drug companies because they don't control the foreign exchange rate, and the macroeconomic cause explains about 95 percent of the price difference. That's sort of an explanation for why there is a price difference.
There's another thing as well, and it goes to the argument that the government is a good buyer as well as selector of prescription drugs, because it can exercise what economists call monopsony power, basically getting a good deal by strong arming the drug companies. We looked at every state in the United States for the year 2000, using a variety of different data sources. We found that the average price of a prescription was $45 and change, and some other things that one would expect, that states that had higher incomes had higher prices for example.
But the important thing was that as you had a higher share of Medicaid paying the prescription drug budget in a state, the higher the prices were. For example, Vermont has a huge share of their prescription drug expenditures paid for by their Medicaid budget, while Arizona has very little government payment, and as a result prices are far higher in Vermont than Arizona. To me, this suggests the classical economic insight that if people are paying out of their own money, they're a little sharper as to how they do it. To paraphrase Friedman, "There's three ways to spend money. You spend your own money on yourself, you spend your own money on somebody else or you spend other people's money on other people." Government does the last and it's not actually an effective negotiator or monopsony buyer.
The other major issue is reimporting, and it's highly relevant now given the actions of Glaxo in the last couple of weeks. Glaxo obviously observed that there was a situation arising in Canada, specifically in the province of Manitoba, where there were large businesses growing that were doing this so—called reimporting or gray—market trade in prescription drugs. So they sent a warning that they would no longer supply these wholesalers and retailers, and this caused a reaction from Representative Sanders and some other American politicians who have focused on the Canada—U.S. price difference and have encouraged their constituents to go North.
Glaxo's response was so predictable. It takes about 10 seconds to figure out why reimporting doesn't work. You've got a $7 billion market in Canada and a $145 billion market in the U.S. If the U.S. market is trying to get the same prices as the Canadian market, the solution, for the manufacturer, is quite obvious. They either raise the price in the smaller market or simply stop supplying it altogether, which is what Glaxo has made some noise about doing.
Now, I don't think Glaxo has done this because it's really a problem yet. The estimates of cross—border trade are about $1 billion in $145 billion market. So I can't believe that in this year Glaxo is going to see a catastrophic effect on its profits because of reimportation, but they are obviously thinking that this is something they have to take a lead in getting a grip on for the future. So reimporting is definitely not the answer for the problem of high drug prices you're facing in the United States.
Now, getting to the meat of this presentation, I'll talk mostly about British Columbia, but also a bit at the end about Quebec and Manitoba. In the mid—'90s, all three provinces faced this problem of managing their public drug benefit, and they took three different approaches. British Columbia had the reference drug program, which is similar to a preferred drug list. Reference pricing occurs when a government reduces its citizens' freedom to buy medicines of their choice for whatever condition they have by taxing them and allocating the proceeds to drugs selected by a government—appointed committee. There were five different categories of drugs, which I'll discuss in more detail later, that the government focused on. What they do is say that out of a group of drugs that are all in the same therapeutic class, the government will pay for the cheapest one. If you want another one, there's two ways you can get the more expensive one: your doctor signs a special authority or you pay the cost difference out of pocket.
In Quebec, on the other hand, rather than prefer certain drugs, they simply made a general increase in copayments. It was interesting because they also made the program universal. They had not had a universal program, and there were a lot of people that fell through the cracks who didn't have employer—sponsored private insurance and weren't on the already—existing government program, which was focused on seniors and the poor. So they made it universal, and then once they realized they were going to have a big budgetary problem as a result, they increased all copayments. And finally, Manitoba changed their deductible to a means—tested deductible, where they just looked at your income and said you're capped at a certain portion of your income.
The most significant of these for our purposes is the reference drug program, because I know in Oregon, Michigan, Maine, Vermont, and in a lot of states, they view the British Columbia model as a success. The B.C. program started between the years 1995 and '97, and it continues to this day, and it reimburses the lowest cost drug in each of its five classes of drugs: ACE inhibitors, dyhydropyridines, nitrates, cardiovasculars, and calcium channel blockers.
The philosophy of reference pricing is that there's a category of drugs, such as ACE inhibitors, and they're all the same, we can just pay for the cheapest one. And there is a fundamental incoherence to this whole philosophy. To illustrate this, let me point out three quotes from a paper on reference pricing published a year or two ago by a group of managers of Pharmacare.
The first one is, "If there is no evidence that a higher price buys better effectiveness or fewer toxicities, then the extra costs should not be covered." Well, that makes sense theoretically, even to me. If there's no evidence, then it shouldn't be covered. But later on, in the same article, they write, "a key feature of reference drug program in B.C. is its flexibility to allow full funding of non-reference drugs if a physician reports that the patient has a specific clinical need." And, finally, "NSAIDs and nitrates were judged equivalent in therapeutic effect, differing mainly in their adverse effect profiles."
Now there's no way all of those three sentences can be true. The first statement is belied by the second and third statements, and the third statement is absurd in and of itself. I'm not a doctor, but how can you say two drugs are equivalent in therapeutic effect, but have different adverse effects? It doesn't make any sense to me as an educated layman that you could say all of those three sentences in the same article. So there's an incoherence in the whole underlying concept of reference pricing.
But being an economist, what I do is I crunch numbers; let's look at those. Reference pricing was initiated in British Columbia in 1995, and implemented in all of the five classes by'97. The goal was to contain costs, which had become difficult to afford under B.C.'s previous program. Under that program, patients were only required to pay the small dispensing fee for the drug, while the government picked up the full cost of the drug itself.
Now if you look at the public plans, per capita prescription drug expenditures in the 10—year period preceding the introduction of reference pricing in B.C., they grew 101 percent in B.C. and 167 percent in the rest of Canada. So B.C. was actually doing better than the national average in terms of growth before reference pricing. And that was true of its private market as well.
But from 1995 to 2001, after reference pricing started, B.C. did a lot worse. There was an 87 percent increase in the B.C. Pharmacare budget versus 57 percent growth in Pharmacare budgets in the other provinces and territories, and on the private side, 63 percent growth in British Columbia and 45 percent growth in other provinces and territories. So reference pricing obviously did not solve the budgetary problem. In fact, while it's not conclusive, it appears from this data that it back—fired relative to other provinces which did not use this cost containment strategy.
The one exception to this was with nitrates, which was the only class where there was really unequivocal budgetary savings. But this is tainted because of a separate factor that came into play. Very early on in the reference pricing experiment, 3M launched a new nitroglycerin patch, Minitran, throughout the U.S. and Canada at a very low price to build market share. It quickly became a market leader. So you can't really say much about reference pricing for nitroglycerin because the savings in nitrates should largely be attributed to the competitive function of 3M entering the market with an inexpensive alternative product.
Meanwhile, the other drug classes were cost containment failures. With calcium channel blockers, Pharmacare itself saved money, but only because out—of-pocket costs increased dramatically, and there was an overall reduction the number of calcium channel blockers prescribed. The private buyers paid a great deal more money, and the increase is even more dramatic if you normalize spending for prescription volume. In other words, reference pricing in British Columbia didn't demonstrate that all calcium channel blockers are the same. Rather, we can see that a lot of people said, well, we're going to pay out of our own pocket in order to stay on the drug we were on originally.
This is demonstrated even more dramatically in the case of ACE inhibitors. Of all of the people who were on ACE inhibitors that were to become restricted, 46 percent of them chose to pay the difference themselves so they didn't have to switch. Another thirty percent of them were given a special exemption. Eighteen percent switched to the non—restricted drug, and then there was a rump of people who switched to an entirely different class of drug and another rump of people who stopped therapy completely. So you only had a mere 18 percent of the total patients switching drugs in response to reference pricing. There wasn't some great enlightenment where people decided to take the cheapest ACE inhibitor. Very few did. Most chose to stay on the newer, more expensive drugs.
Now let us look at the other side of this equation: the health consequences of reference pricing. Recent research by Professor Grootendorst, now of the University of Toronto shows that patients who were switched from restricted cardiovascular class drugs to cheaper ones after reference pricing had higher death rates within 20 weeks. In the short run, there was an increase in the likelihood of admission to hospital for cardiovascular surgery for those on the ACE inhibitor or calcium channel blockers who switched. For those patients on nitrates, there were longer stays in the hospital, more visits to physicians, more visits to the emergency room in the short ran.
Professor Grootendorst is worried that his findings have a statistical problem, and they remain unpublished for the moment, except in a letter to his funder, and available on the Internet. But he hasn't explained what the statistical problem was, and it's not readily apparent.
Now another researcher who has looked into this, Dr. Schneeweiss of Harvard, reported no health consequences for reference drug pricing of ACE inhibitors in his published paper. But in that paper he notes that people who switched from the restricted to the reference ACE inhibitor, from the more expensive to the less—expensive ACE inhibitor, had a 19—percent increase in hospital admissions in the first two months on the new drug. So how do you say that there's no health consequence when you observe a one—fifth increase in hospital admissions?
Well, the answer is that the confidence interval was minus—1 percent to 42 percent, and it takes a lot of heavy lifting to get a statistically significant statement out of that. I interpret this to mean that his research didn't have a lot of statistical power, maybe because there weren't that many people switching drugs, so there wasn't that much to observe. This is the trick with statistics; you're forced to basically say it's not there, when really you just weren't able to observe it. And with a confidence interval that wide, I would say there's more investigation to be done on this issue.
I'm going to wrap up by comparing B.C.'s experience specifically with a couple of other provinces. Quebec increased their deductibles and the co-payments for everyone. And remember that unlike B.C., Quebec did not have a universal program before making these changes. It had a program for seniors and a program for welfare recipients, and there were people who had private insurance, but there was a large segment of the population that fell through the cracks. So the justification for raising the deductible was that it was in anticipation of making the program universal, with the budget increases that would necessitate.
But in the five years since the reforms were introduced, even though Quebec made their program universal and B.C.'s was already universal, Quebec's total prescription drug costs, private and public combined, increased by less. It's an amazing finding. B.C. couldn't even control their costs relative to a province that had introduced universality in its benefit and brought roughly a fifth of the population onto the program.
Admittedly, this is another finding that's not a smoking gun because I haven't been able to do the level of research that Professor Horn has, for example. But if you look at all other health care costs aside from prescription drugs, Quebec has a much better outcome — 16 percent growth over the period versus 29 percent growth in British Columbia. So while it would be very much jumping the gun to confidently say that Quebec has done a better job by spending more on prescription drugs, thereby keeping people out of the hospital and reducing overall costs, it certainly looks that way based solely on the expenditure data.
The other province I compared to B.C. is Manitoba. Manitoba did something that I think is far from perfect, or even as perfect as any health program is likely to be, but it is something interesting. They had had a program for seniors that charged a trivial yearly deductible. And then they decided to means test it. For people who had under $15,000 a year income, the deductible would be 2.5 percent of their income, and for people who were over $15,000 it would be 3.5 percent. It was an interesting approach because it gave a means—tested benefit, but it avoided a welfare trap because it was an almost flat rate. There was a jump at $15,000, and so everybody above it felt the same level of pain when they were buying a prescription drug. Whether you were a low-income earner or a high—income earner, you had 3.5—percent of pain to make you think about what was being spent.
And as with Quebec, this produced vastly different budgetary outcomes from B.C. There hasn't been a lot of research on the health outcomes in Manitoba, so I can't comment on that, but the budgetary outcomes are hugely different. You've got a 74 percent increase in British Columbia's total prescription drug spending, both public and private, and only a 47 percent increase in Manitoba’s.
Also, when you look at the costs in the hospitals and physicians, it doesn't look like Manitoba's drug deductible means test pushed people to use less drugs and more visits. The growth in the physician and hospital costs was pretty much the same in both provinces. So there's something to think about. I think there are some negatives to the Manitoba program, but the key issue here is that both Manitoba and Quebec's systems appear to have resulted in significantly more effective cost containment than the reference pricing approach tried in British Columbia.
Dr. Ralph Hawkins
I’m going to talk about my experience from practicing medicine in British Columbia. When the reference—based pricing system was introduced, not only was I a private—care physician in a smaller community in British Columbia, but I also served on an advisory committee to the Pharmacare program, giving them advice about how to avoid problems with introducing reference—based pricing. Unfortunately, our advice was ignored.
I'll start with a definition of what reference—based pricing is. This is the definition from the Pharmacare program, and I will read it verbatim from one of their newsletters:
“Reference—based pricing is a Pharmacare initiative designed to support first—line prescribing consistent with the best evidence of comparative effectiveness and safety. Under the program, Pharmacare will establish reference products and will provide full reimbursement for them. That reference price may be applied to any other product in that class, but full reimbursement will require a special authority.
To seek a special authority, a physician submits a request by fax or telephone indicating that a patient is unable to tolerate or does not therapeutically benefit from the reference drug. Pharmacare reviews the application and, if it is accepted, grants the special authority, usually within 24 to 48 hours, for another drug in the class to be fully funded.
When a physician prescribes a nonreference drug without a special authority, the PharmaNet computer alerts the dispensing pharmacist, who informs the patient and/or physician of the policy and suggests the following options:
- if there is a patient—specific reason for the use of a nonreference drug, the physician can apply for a special authority;
- if there is no patient—specific reason for the use of a nonreference drug, the physician can change the prescription to a reference drug; or,
- the patient can choose to pay the difference in price between the prescribed drug and the reference price drug."
Now, in the year 2002 there was a review of the Reference Drug Program that was undertaken by the British Columbia Government, and in that review, concerning the special authority process, it was pointed out that 95 percent of all special authorities are approved. While some see this as a positive comment on the efficiency of the system, others question whether the administrative process is really necessary or whether physicians should simply be provided with criteria, and asked to make appropriate judgments and monitor for compliance. With that introduction, what has reference—based pricing done to the health care stakeholders? Well, for patients, they do not get access equally to the best—possible medicines, and they perceive that government is intruding on their doctor—patient relationship. For physicians, and this is the perspective that I'm going to focus on, there's an undue focus on the prescription medicines that are not necessarily the most medically appropriate.
The academic health community has limitations upon their funding for their ongoing research activities, and the government is perceived to obstruct, rather than to facilitate, delivery of health care. It's a very confrontational system, and the government is not perceived as a partner in providing care. The cost savings are, at the very best, time limited and very modest.
Now, in looking at a physician perspective on the reference—based pricing system, I'm going to make use of a study that was conducted in 1997 by MarkTrend Research. A total of 254 telephone interviews were conducted with a random sample of British Columbia general practitioners, internists and cardiologists, and all of the interviews were conducted between February 28th and March 13th of 1997. This was approximately two months after reference—based pricing for blood pressure medications was introduced.
On the total sample of 254, the maximum margin of error is plus or minus 6.2 percent at the 95—percent level of confidence. So what were the problems that were perceived to have resulted from the mandated policy changes?
Amongst doctors who have changed prescriptions due to reference—based pricing, 95 percent reported that their patients have experienced at least one problem because of a medication switch. The most common problem, as reported by 75 percent of doctors who have changed due to the policy, was an acceleration or worsening of symptoms.
Other problems included confusion or uncertainty by the patients (reported by 72 percent), side effects of the new medications (66 percent), patient noncompliance (51 percent), and emergency room admission or hospitalization (20 percent). Only 5 percent of physicians in this group reported they saw no incidence of harm to patients from the policy.
When there was an inquiry into whether the physicians had seen additional patient visits, more than three—quarters of the surveyed doctors said they had had additional patient visits due to the reference—based pricing. The percentage stood at 81 percent for general practitioners. The Ministry of Health, of course, would pay for these increased visits from the Medical Services Plan budget. The terminology they employ is “increased utilization”.
In terms of cholesterol medications, which are not included in reference-based pricing, the physicians were solicited for their opinion about whether they would be appropriate for reference—based pricing, and opposition by doctors outnumbered support by more than two—to—one. Only 30 percent of doctors prescribing these types of medications gave their approval, while 61 percent were opposed, including 39 percent who expressed strong opposition. To this date, cholesterol medications are not subject to reference drug pricing.
In terms of approval for inhaled steroids for asthma, the doctors also oppose reference—based pricing for inhaled steroids. Only 30 percent of doctors gave their approval, while the majority, 62 percent, opposed, typically to a strong degree. Rather than reference price asthma medications, Pharmacare totally delisted a number of nebulized asthma medications from coverage in 1999 to contain costs for that type of medication. With respect to the suitability of the special authority process, most doctors did not believe that special authorization was suitable or effective mechanism for ensuring patients receive the appropriate drug therapy, while 31 percent considered that it was both suitable and effective, and 11 percent weren't sufficiently familiar with the form to give an opinion. Presumably, that 11 percent weren't using the process.
One can certainly question both the suitability and the effectiveness of a process that appears to merely be a rubber stamp with 95—percent approval. How was this process superior to simply honoring the doctor's original prescription as written in the first place?
Moreover, a number of problems with the special authorization process are identified. Nearly nine in ten doctors considered the delays in getting drugs to patients to be a problem with the form. Eighty—one percent of doctors were concerned about patients' confidentiality when doctors were required to include personal medical information on the form and fax it to the government.
More than two out of three doctors believe that reference—based pricing transfers the authority to make appropriate prescribing decisions from the physician to the government. Reflecting their deep concern about the policy, 81 percent of doctors believe that the government should consider other alternatives in its efforts to manage drug costs better, while ensuring appropriate drug care for patients. In the review that was done in 2002, doctors expressed concern about the workload involved in explaining government policy to patients and having to apply for special authorization when required.
The review panel also heard concerns about the loss of professional independence and the desire to ensure access to new drugs for patients. The review panel concluded, on this point, that while the reference drug program does, indeed, impose a constraint on the doctor—patient relationship, in their opinion, this was not unreasonable nor excessive. I wonder if the patients would agree?
The Review Panel also accepted that reference drug pricing is not evidence based in the scientific sense of the term; that is, it is not based on meta—analysis or synthesis of any quantitative data, but on published evidence examined by the Reference Drug Price Expert Advisory Committee. The membership of the Advisory Committee is, to the best of my understanding, not made public.
Good compliance with blood pressure medication has been significantly associated with the use of newer medications, such as ACE inhibitors and calcium channel blockers, as compared to the older medications such as diuretics. Restricting access to newer drugs with better side effect profiles simply causes some patients to stop taking their medication altogether.
I'd like to shift gears at this point and review a tangible example of the importance of looking not just at drug acquisition costs, but also at the downstream cost of disease prevention, patient survival, and therapy—associated morbidity. This approach is sometimes called an integrated or holistic approach, recognizing that drug costs are only one part of overall health care costs.
Let's start with something that should be familiar to everyone in this audience. In the body of the press release announcing the results of the ALLHAT trial is a quotation attributed to Dr. Lenfant, announcing that "diuretics are the best choice for hypertension treatment for both medical and economic reasons."
I point out that this claim is an extrapolation beyond the scope of the actual study. What the ALLHAT study showed about diuretics and kidneys was that there was a significantly faster decline in kidney function in the group receiving diuretics compared to the other medications studied.
The ALLHAT study that was published in JAMA on December 18th, 2002, reported that there was no significant difference in the incidence of end—stage renal disease between the diuretic and the newer drugs, consistent with findings from earlier studies. However, only a single sentence in the report was dedicated to an extraordinarily important observation; namely, that analysis demonstrated a slower decline in kidney function with the newer drugs than with the diuretic.
The ALLHAT report did not elaborate on that observation, and it did not elaborate on the observation that kidney function in the diuretic group went down at a rate faster than you would expect to see in those types of patients through aging alone.
A number of studies have shown that diuretics can adversely affect kidney function. In the European Working Party on High Blood Pressure in the Elderly trial, the patients who had active treatment received diuretic therapy, while a placebo group did not. The risk of mild kidney dysfunction was 23—percent higher in the diuretic—treated group in that trial. Diuretics and beta blockers are the combination where the main drugs used in the Hypertension and the Elderly in Primary Care trial and the Swedish Trial in Old Patients with Hypertension, and both studies showed significant decrease in kidney function in the actively treated group.
In the SHEP trial, after three years of follow—up, the placebo group had no change in kidney function tests, but the actively treated diuretic group had a mild, but statistically significant decline in their kidney function. Analysis of the Syst—EUR trial showed that patients treated with diuretics had a significant decline in kidney function compared to patients treated with calcium channel medication; one of the so—called newer medications. Analysis of the NHANES III Survey Report shows that diuretics were the only blood pressure medicine category associated with abnormal kidney function tests in patients with high blood pressure who were also getting high blood pressure therapy.
The INSIGHT trial compared diuretics with calcium channel—blocking drugs and demonstrated a statistically significant increase in the observation of kidney function impairment in patients on diuretics, compared with the other group. Finally, ALLHAT has joined the other seven studies with the observation that diuretic—treated patients had evidence of significantly faster kidney—function decline, compared with other groups of drugs.
Now, I'm obliged at this point by my code of ethics to explicitly advise you that what follows reflects conclusions that are not generally known or held by the majority of the medical profession. Such a disclaimer is required when nontraditional medical data are presented to an audience predominantly made up of lay people.
With that disclaimer, the repeated observation that diuretic therapy was associated with impaired kidney function provoked my curiosity to try to find data to describe the relationship between diuretic use and kidney injury. I worked very closely with my associate, Dr. Mark Houston, at the Hypertension Institute in Nashville, to come up with this data. And utilizing the technique that is called data fusion, where concurrent observations of the same population, using different measurement tools, is undertaken we have taken the total population observations from the United States Renal Data Service and from a commercial database reflecting medication distribution to the United States for the period 1991 to 2000.
The annual percent change in drug supply for diuretics was accepted to reflect the annual percent change in diuretic drug consumption by the public. This premise permits analysis of the annual variations in disease incidence rates, concurrent with changes in population—wide drug consumption. The fluctuation in annual prescription diuretic use shows a similar variability to the annual changes in the incidence rate for all cause end—stage renal disease in the United States. The interesting thing is that this relationship occurs with a time lag of two years.
Statistical analysis of this relation shows that these variables are strongly and directly related. This relationship allows for the derivation of a very simple linear model that can predict the change in the acceleration rate for end—stage kidney disease if the change in population—wide diuretic consumption is known two years earlier. This linear model has a correlation coefficient, which means that the model would conservatively be expected to explain over 50 percent of the fluctuation in end—stage renal disease behavior from year—to—year.
It is a fact that United States diuretic use increased by 9 percent from 2000 to 2001, and that reflects an actual cost increase of $48 million. Using the linear model, we would project the incident population of renal failure to increase in 2003 by 7.4 percent. If diuretic use in the United States had not increased at all in 2001, we would predict the incident population of renal failure would increase by only 1.2 percent in 2003. The incident population of renal failure for the most recent data year was 95,000.
So the 6.2—percent difference in the observed diuretic rate compared to what might have been would be about 6,000 patients, roughly speaking. At a cost of $45,300 per dialysis year, the Medicare cost, the dialysis cost for these extra patients would be on the magnitude of $272 million per year. In other words, an increment of $48 million for diuretic therapy in 2001 would be predicted to generate $272 million of additional Medicare cost for dialysis in 2003, a cost ratio of 5.6 to 1. If non—Medicare cost for end—stage renal failure care is also considered, the cost ratio gets up close to 8 to 1.
Even if only 56 percent of the excess renal failure is attributable to increases in diuretic use, this data suggests that for every dollar spent on increased diuretic use in the United States, that we would have $4.50 in diuretic—attributable end-stage renal costs two years from now. Furthermore, since the median life expectancy of a typical new patient starting dialysis is three—and—a—half years, for every dollar spent on new diuretic therapy this year, the model would predict at least $16 of attributable lifetime health care expense generated for renal failure care. That's very simply the 4.5 cost ratio multiplied by 3.5 years of survival.
Therefore, with due respect, I would put forward that the NIH recommendation to expand diuretic use for high blood pressure is neither medically nor economically proven. In a society that has banned asbestos insulation and cigarette smoking in the workplace for very valid public health reasons, the data pointing to diuretics as a noxious substance should not be ignored. End—stage renal failure is a public health priority today.
Is reference—based drug pricing the best care for patients? The committee I sat on in 1996 recommended unanimously against reference—based pricing for antihypertensive medications. Reference—based pricing of antihypertensive medications commenced in British Columbia in 1997 anyway. In 2002, the review panel looked at the evidence presented to it and concluded the effect of the program on patient health, particularly in regards to hypertension, warranted further investigation.
Even after more than five years of reference drug pricing, it is entirely uncertain if the policy is beneficial to the patients. The review panel also noted that reference drug pricing is a time—limited cost—containment device. It noted that there are limited areas to expand the program, and if the program is not expanded into other therapeutic categories and does not capture new drug therapies, the cost savings will decline over time as patents expire and generic products enter the market.
In my view, the time—limited savings do not warrant the endangerment of patients, nor the alienation of the health care providers. And, quite frankly, I would not want to see a restrictive drug formulary introduced with the type of rudimentary cost-benefit justification represented by the B.C. experience. Timely access to the optimal medications will improve outcomes, and that should reduce overall costs. Restricted access to any aspect of the health care system simply represents a false economy.
Dr. Joseph Antos
We're very fortunate to have a very distinguished panel, with Bob Goldberg of the Manhattan Institute, Grace Marie Turner, who is the President of the Galen Institute, and Ron Pollack, who is Executive Director of Families USA.
I'll just say one word, before we hear from the panel. Today, the President will release his budget, and we'll get to see whatever is in the budget about the mysterious $400-billion proposal to strengthen Medicare, including a prescription drug benefit.
I don't think we'll see very many details, but we will be getting to this issue this year, and there remains a tremendous desire on the part of policymakers and on the part of Americans, to see that some appropriate kind of drug benefit be included in the Medicare program, and it is exactly that our second panel will be discussing.
Grace Marie Turner
The creation of a prescription drug benefit in the context of overall Medicare reform is a moving target for policymakers and political leaders. The President's proposal is strong on vision and weak on details so far, but certainly it is clear that the President believes that Medicare, as it is currently constructed, is not sustainable, both because of its increasing costs and its limited access to benefits.
The website of the Centers for Medicare and Medicaid Services says that in 1999, Medicare paid for only 53 percent of beneficiaries' medical care, making it substandard by comparison to any modern insurance policy. Further Medicare is placing a huge burden on tomorrow's taxpayers to pay for the program as it exists today, even without adding a drug benefit. So the President has said we've got to change this. In order to be able to make changes, he wants to use the drug benefit to convince members of Congress to take on the issue of transforming the Medicare program into something much more modern, a program that looks a lot more like the consumer choice programs that are becoming increasingly attractive to working Americans.
So the President and the White House asked, as many have before, what is the best health care program in the country today? And not surprisingly, members of Congress and federal officials have designed the best health care program in the country for themselves, the Federal Employees Health Benefits Program (FEHBP).
Within the FEHBP, consumers have a choice of plans that are competing for their business -- trying to provide the best value for the dollar and hoping that they will be able to attract participants both initially and for the long term. That is certainly not the case with the great majority of health plans used by those who have job-based health coverage today. Their employers’ shift plans sometimes as often as every year to find a better deal and a lower price, causing turmoil and confusion for their workers.
The reason the Federal Employee Health Benefit Program works so well is because individual consumers get to make the choices, and health plans are competing for their business. So, the president asked, why not give those same choices to Medicare beneficiaries, creating a new structure, and, importantly, doing so in a way that integrates the Medicare drug benefit with the overall health program, rather than having it be a freestanding benefit.
So, as the President has proposed that Medicare beneficiaries would either be able to stay with the current Medicare program or they could choose from among private, competing plans, and he's put $400 billion on the table in his budget to show that he's serious about his proposal.
Not surprisingly, it has already begun to receive a firestorm of criticism, certainly from the left and even from the right. Senator Debbie Stabenow said the President's proposal says that if seniors want to have help with their prescription drug costs, they would have to lose their choice of doctors and go into a private—sector HMO. Absolutely not true. Seniors could get their drug coverage, keep the doctor of their choice, and not be forced into an HMO. But, nonetheless, the political arrows already have been sharpened.
So let's look at what the options really are for Congress to consider in the coming session. The President's plan is supported by Senate Majority Leader Bill Frist and is based upon a concept that Ways and Means Chairman Bill Thomas advocated as part of the National Bipartisan Commission on the Future of Medicare. So I believe it's clearly going to be taken seriously.
There are also others that I think are going to be given equally serious consideration. The Democratic leadership of the Senate already has reintroduced a bill that would add a drug benefit to Medicare as a part of the benefit structure as it exists today. I believe that this would quickly lead to price controls and restrictions on access to drugs, with the likely result being similar to those described by the earlier panelists.
There also could be a revised Thomas plan, and a revised Senate bi-partisan bill in which there would be a freestanding drug benefit available as an option to all Medicare beneficiaries. It very likely would have some gaps in coverage, but yet be designed to provide some benefits for all and a more generous benefit to lower income seniors. One of the challenges of the Thomas bill is to continue to improve it to incorporate more Medicare reforms.
Another option on the table is a more limited benefit that would provide catastrophic coverage for all, financed by government. The real danger that I see in this is that it would likely wind up being a government—run catastrophic benefit. This is not just the camel's nose under the tent but its head as well in leading to the same kind of price controls and restrictions on access that you would see ultimately with a standard drug benefit add—on to Medicare.
Joe Antos of the American Enterprise Institute and I have been working very hard on a proposal called the Prescription Drug Security Plan, and we've got some key principles that I think provide an important template to look at as these different options are being considered.
First, does it help those who need it most, and does it provide some help to everybody?
Second, would it be run by private competing plans? I believe that's the only way that you're going to get away from government price controls and restrictions on access. We need to have the flexibility of consumers making choices and having them in a conversation with the health plans to provide the coverage that provides value and that works for consumers.
Third, we believe that it's incredibly important to avoid price controls or government formularies.
Fourth, it's important to not crowd out existing coverage. At least two—thirds of seniors have coverage now, and many of them find, when they look at the coverage they have compared to the coverage that they could be offered through one of the various proposals considered last year, that they like the coverage that they have now. We should not crowd that out.
Finally, we need to build a foundation for a competitive Medicare program for the future. That is the only way that this program is going to be saved for future generations and the only way that the taxpayers are going to be saved from a crushing burden.
The Prescription Drug Security Plan is designed to meet this set of principles, to engage consumers in making good choices about the prescription drugs that they need and that will provide them the value and the treatment that they need without having government bureaucrats or private—sector bureaucrats make decisions for them about the best drugs that will be available to them, now and in the future.
There are three crucial points that I would like to make today. One focuses a little bit on the structural debate that took place with the leak of the President's proposal. Secondly, I want to focus on what resources are available to deal with prescription drug coverage and how that squares with the amount of money seniors will be spending over the next 10 years. And thirdly, I want to talk about where I think this debate ultimately will go, in terms of structuring a benefit.
As a preface to that, we talk a great deal about what portion of the senior population has or does not have prescription drug coverage, and, roughly, about one third of America's seniors at any point do not have prescription drug coverage. That percentage, really, masks the number of people who are affected by the lack of drug coverage because some who have it no longer have it some part of the year, and because a fair number of people who have prescription drug coverage have very limited coverage that’s limited by certain payment caps, like $500 or $1,000 in payments per year, and once you reach that cap, you are, in effect, uninsured for prescription drug coverage.
I think what we're going to find over the next few years is more and more seniors are going to be uncovered for prescription drugs. For each of the three predominant ways that seniors get their coverage today, there is a risk that coverage will be reduced.
The predominant way that seniors have coverage is through their retiree health benefits through their previous employers. And as employers feel the pinch of huge health care cost increases and look to pare those costs down, they really are dealing with three constituencies: current workers, the dependents of those workers, and retirees. Of those three, the group that is farthest removed from the workplace is retirees, and so not surprisingly, we are seeing and I think will continue to see significant reductions in the coverage that retirees receive. This will have a significant impact on whether or not they have prescription drug coverage.
With respect to Medicare+Choice, I don't think I need to explain what its history has been, at least over the last few years. Quite a few communities have found that Medicare+Choice plans have left their areas and left people high and dry. Mind you, those people who joined a Medicare+Choice plan largely because they got prescription drug coverage often have some real difficulties gaining that coverage once Medicare+Choice left, because their best option then is to go back to traditional Medicare and then seek Medigap coverage, and there is no guarantee that Medigap will provide coverage with a prescription drug plan. Of the 10 Medigap policies, only three cover prescription drugs. And while patients, forced back into traditional Medicare, are allowed to gain access to a Medigap plan, they are not guaranteed the option of joining a Medigap plan that has prescription drug coverage.
With respect to Medigap coverage, only a small percentage of those people who purchase Medigap coverage have purchased one of the three policies that include prescription drugs, and that's because the premium costs for prescription drug coverage in Medigap plans has escalated very substantially because the only people who tend to buy those policies are those likely to have high use of drugs.
So I think we're going to find that this crisis that has brought us to the brink of developing a prescription drug proposal that can pass Congress and be signed by the White House is going to get worse, and, clearly, we need to focus on getting the job done this year. Having said that, I suggest to you that it is going to be very, very difficult to pass a prescription drug proposal in this Congress. We have seen the difficulties in past years. To the extent that the President's proposal carries the kind of freight and baggage that the news leaks suggested the administration was contemplating, I think it would transform an already up—hill climb to a Himalayan expedition. And so I think the administration would be well advised to avoid linking that kind of reform to a prescription drug benefit.
There are a good number of us who are certainly willing to see a system of competition arise from legislation that is enacted, but I don't view what we've heard of the administration's plan as competition. That is, they are tying the hands and feet behind the back of the public plans, in terms of their competing with private plans, and, given the extraordinary percentage of America's seniors who have opted to remain in the public plan, I don't think that that is going to receive a great deal of favor. Certainly it's going to make it far more difficult to pass a prescription drug benefit in this Congress.
Let me get to the resource question because I think it tells us a lot about potential structure of a benefit. As the President indicated in the State of the Union message, and as presumably we'll be able to confirm when each of us looks at the budget that was released today, the administration plans to put on the table approximately $400 billion over the next 10 years for a prescription drug benefit.
Well, that's real money, and it is over double what the administration proposed the previous year, which was $192 billion. But it is important to understand that this is in comparison with approximate total expenditures on prescription drugs for seniors over the next 10 years of approximately $2 trillion. That tells you something about what can and can't be done and how it will be perceived in terms of a new prescription drug benefit. If the President places on the table $400 billion in new money for prescription drugs, that amount is approximately one—fifth of the total expenditures that seniors already are projected to spend.
What this indicates is that, to the extent that you take that money and you try to spread it across the entire senior population, the benefit is not exactly going to be heralded as satisfactory by the senior population. I think it's going to provide too little a benefit for seniors for it to be adequately appreciated by its intended beneficiaries.
If that assumption is accurate then, I think, as we move further along in the debate we're going to have to pare down our vision of who gets helped and how we help them. My presumption is that we are going to have to focus pretty soon on trying to make sure that the neediest portion of the elderly population gets served first, and everyone has got their own definition of that what neediest population is.
I think we should start with those of lower and moderate incomes. Obviously, in the mix will be those with higher incomes who experience catastrophic drug costs, but I think that with the real limitation in resources, there is going to have to be some careful targeting. I certainly would like to see all seniors have adequate coverage. I believe in social insurance—type principles, but given the realities of the budget, and given the realities of the expenditures for America's seniors, I think we should make sure we do the job well for the portion of the population that needs it the most, and I believe the debate will move in that direction.
Lastly, let me conclude by saying a few words about what I mean by doing that job well. There have been a number of different proposals concerning how to provide drug assistance for low—income populations. Some of these proposals have been part of a larger proposal designed to cover everybody, and this, in effect, would be a wrap-around policy. Others have been more freestanding, such as the administration's initial proposal when it came into office, the so—called immediate helping hand.
I think that irrespective of the structure, whether it's part of a broader package or is a stand—alone, there are some very important things that need to be done to make sure that this help to low-income seniors is genuine. First, the income eligibility standards need to be constructed to make sure that the benefit really reaches the people who need help the most. Remember that the poverty line for a senior living alone is approximately $8,000 in annual income, not very much. And so my hope is that a policy that targets needier seniors is going to reach into eligibility levels that are somewhat close to double the poverty line, which still is not exactly a princely sum.
Secondly, the benefits need to be structured in such a way that they are realistically affordable. Now, when various proposals were considered last year, and the year before, with respect to an overall prescription drug benefit, many of those policies had so—called doughnuts in them, gaps in coverage where after the first dollars were covered, then next dollars were not covered at all, and then coverage picked up again at some catastrophic amount. Whatever you think about that kind of a structure, it really does not make sense for low—income seniors because the gap in coverage, and the thousands of dollars that may be the gap in the coverage, simply make prescription drugs unaffordable for a low—income population. Even if one likes that kind of a structure for everybody else, it really will not work for low—income seniors.
Thirdly, there is likely to be a significant debate that will emanate from the governors who want to see any kind of funds for a low—income benefit captured as much as possible for fiscal relief. While I have some sympathies with the plight of governors who are experiencing fiscal crises and who are paring back their Medicaid programs, to the extent that more and more money is given to the states for fiscal relief, there will be less money to provide genuine relief for low—income seniors. My hope is that dollars allocated for a prescription drug benefit are retained as much as possible so that additional low—income seniors get coverage, rather than simply providing fiscal relief to states across the country.
Lastly, we've got to make sure that the benefit system and the system for gaining access to it is realistic and user friendly. With respect to making it user friendly, we've got to make sure that the systems for signing up seniors really work. It is often the case that means tested programs fail to reach a significant portion of those who are eligible for assistance. There are a variety of mechanisms that can be used to reverse that trend. They need to be employed in such a system, and we've got to make sure that the benefits that are provided to seniors, particularly low—income seniors, are comprehensive.
Some of the Medicaid programs that provide prescription drug coverage limit the number of prescriptions that a low—income senior can get. In some states, you can only get three prescriptions a month. Certainly, that is going to provide far too limited relief to low—income seniors. We need to do better.
Dr. Robert Goldberg
I'm going to discuss the status quo a little bit, i.e. what is possible in this congressional session, and then present what I think we should be focusing on in light of those realities.
First though, let's take a step back and look at the issue of targeting the people that really need help. For example, 60 percent of the seniors that are eligible for Medicaid, who could get drug coverage in that fashion, are not signed up. That's something that Congress has not focused on because of the larger debate over Medicare reform and prescription drug coverage.
We should be focusing on that situation because that can be immediately addressed. Again, I look at the number of seniors who are eligible for drug coverage under the Medicaid program and aren't getting it, and that's something where additional funding, if structured properly, could be made available quickly.
Another point is that all of this talk about a drug benefit and putting people into health plans versus the fee—for—service is really just scratching the surface of the larger financial burden that Medicare is going to claim on current taxpayers and future generations. My standard line is why should the woman that is waiting tables pay for George Steinbrenner's prescription drug benefits? I think there's going to be a realization that whatever we do going forward, people who are of means must bear more responsibility for their own health care costs. That way we can focus government resources and tax dollars on those who really can't provide for themselves, and I think that's going to be part and parcel of any kind of Medicare reform.
Having said that, there's even a larger issue which will continue to come back to confront us in our deliberations. That is, the way we pay for health care is totally inconsistent with the technology of health care. Setting aside the fact that we don't always pay for prescription drugs, even though prescription drugs are our most dynamic health technology, and the most important part of managing people's health care, even when we do cover prescription drugs, it's in a haphazard fashion. I think Dr. Horn has eloquently demonstrated the importance of treating and managing disease at the patient level, and yet we have not grasped that fact yet — we are not doing that effectively either through the Medicare program or in the private sector.
The President has really laid some larger questions on the table: should the government or private plans provide drug coverage and health care management for seniors? And how and by whom should it be paid for?
The question is, which is a more humane and effective way to promote access to medical progress? For over the last two decades, for example, disability among seniors has declined 25 percent, largely because of new technologies and new medicines.
Most of that spending has been taking place in the private sector through private health plans. As for Medicare+Choice, Robert Berenson had a very illuminating article in Health Affairs showing that the problem with Medicare+Choice is that the plans were lashed to the price controls of Medicare and really couldn't compete. There were some other reasons too, such as that all managed care plans were on the down swing of their underwriting period. But in any event, Medicare+Choice expanded prescription drug coverage to millions of seniors, albeit poorly.
As much as the private sector has been restricting medications, they have also been very aggressive in promoting the idea of managing care by trying to give people drugs to prevent further diseases. So now they're feeling a backlash from their success in trying to limit access to drugs, and I think there is a more openness to the notion of the need for open access. This is shown by the fact that most of the formularies are now open in these managed care plans, and they're trying to manage costs through disease management and through copayments instead. I contrast that with the behavior of Medicare, where they've basically done reference pricing on biotech drugs in the last year. Veterans Affairs limits the use of SSRIs to one drug — you have to get prior authorization for everything else. They limit use to one atypical antipsychotic. They've been doing all of the kinds of things that drive up health care costs and just don't make sense.
We stand at the edge of an age where illnesses still stalking the elderly, like Alzheimer's, muscular degeneration, incontinence, diabetes and heart disease, may become not only curable, but also predictable or even preventable. Right now it takes two to five years for Medicare to make some of these drugs and devices accessible to people, and some diagnostic tests. I know for a fact that treatments that are available to the private sector aren't available through Medicare. So we have to find a way to pay for things differently. Moving stuff into the private sector is one way to do it.
I don't have the ultimate answer. But some things are clear. Integrating medical technologies that provide the most value makes the most sense for Medicare reform. Giving people greater freedom for investing in and paying for their own health care is the best way to promote better use those technologies.
In the end, we want to choose a system that costs less and adds more value to our lives. I think we've seen, in the first panel, that open access and consuming more new technologies is a good investment. So I'm all in favor of spending more on prescription drugs.
To the extent that we can move these decisions into the hands of individuals, and to the extent that we can target our resources to the poorest and the neediest, not just people that are of low income, but those with chronic illnesses, I think we'll make great strides. If we can do that, perhaps we'll begin to look at Medicare as a way for enhancing lives and not as an albatross around weighing down future generations.