|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 2004; 44:1809-1811, doi:10.1016/j.jacc.2004.08.010 © 2004 by the American College of Cardiology Foundation |
Duke Clinical Research Institute, Division of Cardiology, Department of Medicine, Duke University Medical Center, Durham, North Carolina, USA
* Reprint requests and correspondence: Dr. David F. Kong, Duke Clinical Research Institute, DUMC Box 3850, Durham, North Carolina 27710 (Email: kong0008{at}mc.duke.edu).
An economist is a surgeon with an excellent scalpel and a rough-edged lancet, who operates beautifully on the dead and tortures the living.Nicholas Chamfort (1741–1794)
The cost of health care attracts intense attention every decade or so, when professors, presidents, and popular media trumpet declarations of "crisis" and predictions of doom. Although consensus exists that rising health care costs are undesirable, politicians and health care policy experts predictably fail to agree on viable solutions. The resulting friction generates considerable heat, yet a paucity of light. Consequently, little ever changes.
Rising health care costs defy regulation for several reasons. Our aging population will require more medical care than a younger one; this trend has yet to show its full effect. Inflation forces consumers to pay more for the same product. Thus, if materials and labor cost more, then health care produced from those resources will also cost more. Technological innovation represents an essential, often unrecognized third factor. Simply put, health care costs more today than a decade ago because technical advances now offer unprecedented therapeutic capabilities.
Pushing cutting-edge technologies from initial concept to regulatory approval requires tremendous investment—about $800 million for new pharmaceuticals (1). These development costs often make new therapies significantly more expensive than existing treatments for the same condition. Development dollars fund extensive portfolios of evidence, designed to convince health care decision makers and payers that new treatments should be standard therapy and, therefore, reimbursed by whomever is paying the bills. To justify regulatory approval, these collections principally feature "pivotal" randomized trials of clinical effectiveness. Increasingly, portfolios include economic analyses showing that the new technologies provide good value for money (2).
The insistence of health care consumers and purchasers on rigorous substantiations of value is essential for better stewardship of the nation's health care investments. Conducting economic analyses in the setting of major randomized trials reduces the potential for bias and improves credibility. On the other hand, judgment for regulatory approval remains a distant cousin to rational medical decision making. Regulatory trials often answer narrow questions that are only indirectly related to underlying key clinical and health policy issues. The law mandates the U.S. Food and Drug Administration to evaluate safety and efficacy, not to guide clinical medicine or assess incremental value. Thus, industry ordinarily designs pivotal trials to best support approval by demonstrating statistically significant proofs of efficacy. These trial populations are skewed subsets of the eventual recipients of the candidate therapy. For example, women, the elderly, and patients with serious comorbidities are particularly under-represented in most trials (3). Economic analyses of "pivotal" trials may demonstrate the value of new therapies in selected circumstances, but these studies fall short of a generalized stamp of good value for societal dollars. Economists frequently address these limitations by mathematically modeling treatment costs and benefits in the presumed target population. A lack of population-based data characterizing key patient subgroups hampers model construction. The uncertain effectiveness of treatment in patient groups not represented in the trial also constrains these model-based extrapolations. Models containing more educated guesses evoke less credibility among payers and decision makers.
In this issue of the Journal, Cohen et al. (4,5) present economic analyses of two technical improvements to percutaneous coronary intervention (PCI) in selected populations. Both trials presume that the value of performing PCI has already been established in the respective populations. In the Randomized Evaluation in PCI Linking Angiomax to Reduced Clinical Events (REPLACE)-2 trial, bivalirudin, a direct antithrombin, saved money relative to glycoprotein (GP) IIb/IIIa platelet inhibitors for patients undergoing nonemergent PCI procedures, with noninferior outcomes (4). This occurred primarily because bivalirudin costs less to provide to patients than abciximab or eptifibatide. In the Saphenous Vein Graft Angioplasty Free of Emboli Randomized (SAFER) trial, use of a distal protection device in saphenous vein graft PCI incurred a higher net cost than PCI without the device, but significantly improved infarction-free survival (5). In cost-effectiveness analysis, the distal protection device was a very efficient way to generate improved health outcomes (i.e., favorable cost per life year added).
What do these studies teach us about the relationship among technological innovation, assessment of value for money spent, and the ever-increasing cost of medical care? Bivalirudin appears to provide health benefits similar to existing therapies but for less money—a "win" for all parties. Economists refer to a therapy that fulfills these conditions as "dominant" and would expect its wide adoption into general practice. Clinicians, however, have been slow to replace adjunctive GP IIb/IIIa therapy with bivalirudin for several reasons. The REPLACE-2 trial was designed as a non-inferiority trial, which increases the uncertainty surrounding the therapeutic benefit. The early REPLACE-2 trial results suggested that bivalirudin might be associated with a higher rate of non–Q-wave myocardial infarction (MI), creating additional ambiguities about its benefit (6). One-year survival data from this trial recently showed no evidence that the small excess of non–Q-wave MI resulted in a higher mortality for bivalirudin patients. Whether these additional data will translate into adoption of the dominant therapy (bivalirudin) remains to be seen. Although unlikely, complete replacement of GP IIb/IIIa therapy with bivalirudin during PCI would potentially save $80 million annually. This would be insufficient to solve the health care cost problem, but a movement in the right direction.
The embolic distal protection device studied in the SAFER trial contrasts with the case of bivalirudin. In the SAFER trial, use of the device reduced MI significantly at 30 days (7). Although a mortality benefit for distal protection is mechanistically plausible, it has yet to be detected at nominal levels for statistical significance (1.0% vs. 2.5%, p = 0.11). In the economic analysis by Cohen et al. (5), embolic distal protection was associated with a cost offset of $965 due to fewer subsequent ischemic complications. The technology studied costs approximately $1,500 per device. As a result, the clinical benefits of distal protection were obtained at a net incremental cost of $582 per patient. Extrapolating the reduction in death and MI to a lifetime horizon, Cohen et al. (5) estimated that distal protection technology costs $3,718 to save one life-year. Economists generally consider anything <$50,000 per life-year saved economically attractive. Only the most cost-effective therapies, such as coronary artery bypass graft surgery (CABG) for left main disease and medical therapy for severe hypertension, have values <$10,000 per life-year saved. Although the incremental cost (and benefit) to society is attractive, the technology comes at a $1,500 price tag to hospitals. With more than 500,000 patients receiving CABG surgery annually in the U.S., the number of vein graft interventions (and demand for routine distal protection) could rapidly obliterate any savings obtained from converting from GP IIb/IIIa to bivalirudin.
Vigorous competition among disparate therapeutic technologies for health care dollars drives the systemic costs of health care inexorably upward. Although many new therapies may satisfy the conventional criteria of good value for money, as distal protection did in the SAFER trial, few will pay for themselves, as bivalirudin did in the REPLACE-2 trial. So, where should we draw the line on health care expenditures? The U.S. presently spends more than 15% of its gross domestic product on health care. Oddly, economists cannot identify how much health care spending is "too much." Rather, spending limits are defined by the willingness of consumers, employers, and taxpayers to fund care and society's allocations to competing priorities such as education, defense, and the environment. Additional spending for new technologies may be acceptable if these technologies convincingly improve the lives of patients. However, most trials of new technologies fall short of definitive proof in this regard.
Beyond generalizing pivotal trial results to clinical practice, economic analyses of novel technology further assume that the comparator therapy is the standard of care. The value of the newest technical solution is predicated upon whether the comparator therapy is, indeed, good value. These two studies examine the value of improving a certain procedure relative to some established way of doing the same procedure. Technical improvements may produce a better, safer, more expensive procedure, but they should not be confused with the more important system issue of appropriateness—whether we should be doing the original, unimproved procedure in the first place. Whether society chooses the "first class" PCI or the "coach" PCI becomes relevant once procedural appropriateness has been established. Unfortunately, few trials compare PCI with medical therapy, and none have shown superior life expectancy or freedom from major disability (8,9). Similarly, comparisons of PCI and CABG in patients with extensive disease have largely favored surgery (6,10,11). The pioneer vision of PCI as a less invasive form of bypass surgery remains unrealized, while many current PCI procedures fall most easily under the classification of a more expensive alternative to medical therapy. Attempts to address therapeutic appropriateness in large clinical trials are often hampered by rapidly evolving procedural techniques and clinicians' unwillingness to enroll patients in a trial addressing a question they have already answered, at least for themselves.
Given this litany of limitations, what is the proper role for economic analysis? Rigorous assessment of value for money should certainly receive more attention than it does. Moreover, these analyses illuminate the gap between what we would like to believe and what clinical evidence supports. Economic analysis forces analysts to explicitly indicate how a new therapy affects the quantity and quality of life. By formally elaborating otherwise tacit assumptions, these studies silhouette the protean uncertainties dividing the glowing promises of research from the realistic expenses of society.
 |
Footnotes |
|---|
* Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology. 
|
References |
|---|
 Top References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
