August 6th, 2012
TAVR and the Price of Innovation
In an article published in Circulation, Daniel Mark and Robert Mentz of the Duke Clinical Research Institute and University Medical Center examine the economic and policy implications of the recent validation of transcatheter aortic valve replacement (TAVR) in the PARTNER trials. CardioExchange is fortunate and pleased to present Dr. Mark’s responses to questions about his work posed by Rick Lange and David Hillis, co-moderators of the Interventional Cardiology blog.
Lange and Hillis: In the “inoperable” AS cohort (cohort B in the PARTNER trial), TAVR extended life expectancy (from 1.2 to 3.1 years) and improved quality of life at a cost of $62,000 per quality-adjusted life year, according to a recent study by Reynolds and colleagues. Is this cost effective? How is cost effectiveness determined when no alternative treatment is available? To put it in perspective, how does TAVR compare with recently approved biological cancer therapies?
Mark: Cost effectiveness is a calculation of the efficiency with which a particular therapy of interest produces desired health outcomes. One key assumption in this calculation is that both the health benefits and the relevant costs are incremental. Thus, at least two strategies are always considered in any cost-effectiveness analysis. In the case of inoperable AS, the reference strategy could be termed “usual care,” which is essentially palliative symptomatic therapy. In PARTNER B, a significant proportion of the “usual care” arm also received aortic balloon valvuloplasty. To perform a cost-effectiveness analysis of TAVR in this population, it is necessary to identify all the health benefits that were changed and all the resource use that was affected by TAVR compared with the control arm. This is what we mean when we say that the analysis is incremental. We must also identify the time horizon of the analysis, which, for most chronic disease therapies, is the lifetime of the cohort. Once we know the lifetime incremental health benefits, typically expressed in quality-adjusted life years (QALYs), and the lifetime incremental costs (the dollar value of the incremental resource consumption), the calculation of the cost-effectiveness ratio is simple.
This ratio has no natural interpretation, so accepted benchmarks must be used. One strategy economists use to come up with such benchmarks is to find other therapies for which a clear consensus about value for money exists, and use those to set the context for the new therapy. For many years, hemodialysis was used as such a benchmark because it is a legislatively mandated entitlement and, in at least a loose sense, represents a societal consensus on spending healthcare dollars to save lives. The benchmark is often cited as less than or equal to $50,000 per QALY, but taking the increased cost of hemodialysis therapy and inflation into account, the benchmark probably is more like $100,000 per QALY today. Thus, against this benchmark, TAVR for inoperable AS represents reasonable value for money, or a reasonably efficient method of producing extra health benefits for the population. Notice that the method is indifferent to who benefits from the therapy. How we feel as a society about spending healthcare dollars on particular segments of the population raises all sorts of extra-economic value issues that tend to make physicians uncomfortable. We are trained to relieve suffering and prolong life when we can. In the past, this was done without regard for cost or economic efficiency. We are clearly moving into a new era in which discussions about cost and cost effectiveness are becoming a critical part of the debate about new technologies. The PARTNER investigators have done an excellent job of giving us high-quality comparative effectiveness and cost-effectiveness data to help direct the deployment of TAVR into clinical practice.
Lange and Hillis: New medical technologies are typically evaluated by their cost effectiveness, but in your article, you urge us to consider affordability (i.e., how much will TAVR add to the total healthcare bill, and are funds available to pay for it?). Is TAVR an example of a cost-effective medical breakthrough that is not affordable? Should TAVR be “rationed”?
Mark: The question of what we in the U.S. can afford in the way of effective new therapies has no easy answer. Unlike some countries, we do not have a defined budget for all of healthcare. Without a budget, there is no way to decide what we can afford to spend. The amount we spend increases each year, seemingly without limit. In addition, the Institute of Medicine has recently estimated that about 30% of current health spending (about 5% of GDP) is wasted. We are a wealthy country, and we can afford to spend more than many others on healthcare. But we cannot spend without limits, and we cannot throw money at things that provide no value.
Rationing is, of course, an inflammatory word in this country. But, as many have pointed out, we already ration care in a variety of ways. I believe that what people who overreact to the idea of rationing are really concerned about is the loss of control. Some faceless bureaucrat will tell them that they cannot have lifesaving therapy because it is too expensive, or they are too old or in some other way of insufficient value to society, and they will have no recourse, no way to appeal. But, of course, this already happens in the U.S. A fundamental axiom of economics is that we live in a world of scarce resources, and we cannot give everyone what they want or think they need. We can do amazing things in medicine in 2012, things of which we should be very proud. But we have not been good stewards of social resources, and we are paying the price for that neglect in terms of increasing loss of autonomy. The ACC/STS/FDA/CMS-controlled rollout planned for TAVR with mandatory participation in a national registry is actually an exciting new step for us, and I hope it demonstrates a way for us to adopt new technologies into clinical care without the irrational exuberance that has characterized our adoption of some earlier technologies.
Lange and Hillis: In high-risk AS patients who are eligible for surgical AVR or TAVR (cohort A in the PARTNER trial), you refer to the choice as “a therapeutic and policy toss-up,” because a small, nonsignificant early difference favoring TAVR (0.065 more life-years and $2200 lower cost) was noted at 1 year of follow-up. On a population — rather than individual — basis, wouldn’t healthcare costs be lower if all high-risk patients had TAVR rather than surgical AVR?
Mark: Our estimated $2200 per patient might seem like a lot, but it was a very small percentage of total 1-year costs and was not statistically significant. Thus, it remains uncertain whether, when TAVR is deployed across the country, we will continue to see a $2200 1-year cost savings with TAVR. In addition, before we are ready to send surgical AVR to the Smithsonian, we need to be very certain that we understand the lifetime health implications of TAVR in the target population. Unless we are very confident that the health benefits of TAVR are at least as good as surgical AVR over the lifetime of the population, we are not ready to mandate a change in standard of care, regardless of potential cost-saving opportunities.
Lange and Hillis: A comprehensive assessment of a novel therapy looks beyond the questions of economic efficiency and assesses changes created by that technology in the healthcare system. What changes in the healthcare system are created by TAVR that should be considered?
Mark: Ideally, we would understand all the changes that TAVR will create in the healthcare system before we roll it out for general clinical use. Of course, this is not the way things happen in real life. New therapies get introduced on the basis of one or two pivotal clinical trials involving hundreds or thousands of patients, and it is only after clinical use in hundreds of thousands of patients that we become aware of some of the unintended consequences of the technology. Look at all we have learned and continue to learn about stenting for coronary artery disease. Even now, are we ready to say we understand all the ways in which percutaneous coronary intervention has altered the healthcare system? All that we can be certain of at this stage is that we have much more to learn about TAVR and how best to use it in our patients.