Author + information
- Received March 7, 2003
- Revision received August 5, 2003
- Accepted August 25, 2003
- Published online February 18, 2004.
- ↵*Reprint requests and correspondence:
Dr. David R. Holmes, Jr, Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA.
Cardiovascular medicine is changing rapidly with the development, testing, and introduction of new diagnostic and therapeutic methods. New interventional techniques such as the use of drug-eluting stents have important implications for the care of individual patients and the delivery and economics of health care in general. Drug-eluting stents have been shown to improve outcomes among patients undergoing percutaneous coronary intervention by significantly reducing restenosis rates. Two randomized trials have documented that per 100 patients treated with the sirolimus drug-eluting stent, 12.5 to 13.6 patients avoided the need for subsequent target lesion revascularization, when compared with patients treated with conventional stents. The economic effect of the introduction of these stents, which are projected to be two to three times as expensive as conventional stents, is complex and depends on which segment of health care is considered. These stents will be favorably received by patients, physicians, employers, and society as well as payers. However, hospitals may be adversely affected by having increased procedural costs for the stents, along with fewer procedures for evaluation and treatment of restenosis and probably decreased surgical volumes. Drug-eluting stents are only the first of many new technologic advances that will affect cardiovascular care. These procedures have many features in common, including: 1) replacement of major surgical procedures with less invasive approaches; and 2) redistribution of costs, with a decrease in hospital profits but potentially lower costs of health care delivery for society as a whole.
Health care costs have attracted considerable attention from the federal government, consumer organizations, the public, and health care professionals. In 2001, the estimated average outlay for health care by every person in the U.S. was $5,035. Total outlays in hospitals were $451.2 billion, and total outlays for physicians and their services were $313.6 billion (1). A substantial percentage of this budget was spent on cardiovascular diseases, the leading causes of morbidity and mortality throughout the Western world. Health care costs for the treatment of cardiovascular diseases are complex, involving patients, employers, third-party payers, the federal government, physicians, paramedical personnel, hospitals, and both device and pharmaceutical industries.
Rapid changes within cardiovascular care are expected to stress the various components of this complex system. These changes have resulted from the development, testing, and introduction of new diagnostic and therapeutic methods that extend treatment opportunities for patients with established disease and other treatment approaches before symptomatic disease develops. New interventional techniques have important implications not only for the care of individual patients but also for the delivery and economics of health care in general.
This article explores the potential effect of new technology on the delivery of health care and defines the paradigm shift, using the drug-eluting stent as an example. The drug-eluting stent for the treatment of coronary artery disease is probably only the first of several technologic advances that will redefine this paradigm.
Coronary artery bypass graft surgery (CABG) is performed frequently; it is estimated that in 2001 there were approximately 520,000 of these operations in the U.S. (2). The indications for surgical revascularization vary. Surgical therapy, compared with medical therapy, has improved survival in several subsets (e.g., in patients with two- or three-vessel disease, particularly with left anterior descending coronary artery involvement or with abnormal left ventricular function) (3). Compared with conventional angioplasty, surgical therapy has been found to improve survival in diabetic patients (4–7). In the Bypass Angioplasty Revascularization Investigation (BARI) trial, at seven years after treatment, 76.4% of diabetic patients treated with CABG were alive, compared with 55.7% treated with angioplasty (4). The extent to which stents, particularly the drug-eluting stent, will mitigate this difference and improve survival remains to be determined (8)and is scheduled to be tested in a randomized multicenter clinical trial. In other groups, surgical revascularization is performed to improve refractory angina or ischemia that cannot be controlled by either medical therapy or percutaneous coronary intervention (PCI).
Surgical revascularization has often provided excellent results, although it has a number of limitations (Table 1), including costs associated with preoperative and postoperative care, the need for several days of hospitalization (including intensive care), rehabilitation, and delayed or inability to return to work. Because the useful life of venous bypass grafts is seven to 10 years, some patients may need to be re-treated, sometimes with results that are poorer than after the initial surgery. For the payer, the cost of these procedures is high—a minimum of $20,000 to $30,000 for each hospital stay. For the hospital, however, these procedures constitute a substantial source of both revenue and net income (9). In cardiovascular surgeons' practices, CABG often provides the bulk of their income. Finally, for patients, CABG is a major surgical procedure with attendant risks, including the well-recognized potential for neurologic complications or a decline in neurocognitive function (10,11).
Percutaneous coronary intervention provides a less invasive approach for revascularization and offers a shorter hospital stay and faster recovery. During the first decade of PCI, the quality of results was limited by the lack of user-friendly and lesion-suitable technical equipment. Dissection and acute or threatened closure (12), resulting in major morbidity and mortality, were sufficiently frequent and unpredictable so that in some institutions, an operating room was kept empty but staffed in case complications required emergency surgery. An additional problem was restenosis. In one of the earliest studies of percutaneous transluminal coronary angioplasty, restenosis occurred in 33% of patients undergoing angioplasty from 1978 to 1980 (13).
In the largest multicenter trial of PCI with stenting—the Prevention of REStenosis with Tranilast and its Outcomes (PRESTO), in which approximately 11,500 patients were enrolled from 1998 to 2000—the angiographic restenosis rate was also 33%, although this included more complicated types of lesions than those that are now being treated (14). Conventional stents improve the short-term outcome of PCI, mainly by decreasing the frequency of acute or threatened closure and obviating the need for urgent/emergent CABG. In the most recent American College of Cardiology National Cardiovascular Data Registry experience of 146,907 patients treated with PCI, surgery was required or performed in 1.9% of the patients (15). In another recent series, from 1992 to 2000, emergency CABG was performed in only 0.61% (16). It has also been repeatedly documented that angiographic restenosis rates with stent implantation are approximately 30% to 50% lower than with conventional percutaneous transluminal coronary angioplasty (17). Restenosis has not been eliminated by stent implantation because, although stenting essentially eliminates acute recoil and late vessel remodeling (2 components of the restenotic process), it may aggravate neointimal hyperplasia (18–21).
In-stent restenosis may be recalcitrant. Treatment of in-stent restenosis with conventional techniques still requires subsequent intervention in 35% to 80% of patients, depending on the specific lesion morphology. The effect of restenosis on outcomes among patients who were candidates for either PCI with stenting or CABG has been evaluated in several studies (22–24). In the largest such study—Arterial Revascularization Therapy Study (ARTS)—1,205 patients with multivessel disease were randomly assigned to be treated with either conventional stenting or CABG. Although death, stroke, and myocardial infarction rates were not different between the two groups at one, two, and three years, the event-free survival at one year was significantly greater with CABG (87.8%) than with stenting (73.8%). This difference was related to restenosis and the need for subsequent procedures, which were required in only 3.8% of the surgical group, as compared with 21.0% of the stent group. At three years, the only change was a slightly larger difference in the need for subsequent interventions. Because of these additional procedures, the initial cost savings of stent implantation relative to CABG decreased over time. On the basis of these and other studies, some patients and their physicians may still choose CABG rather than PCI with stenting to avoid the need for subsequent procedures. Historically at the Mayo Clinic, treatment of restenosis has accounted for 15% of the total number of yearly interventional procedures. The need for several different procedures for the treatment of coronary restenosis spawned the development of new techniques such as vascular brachytherapy (25,26), which are effective but costly because they require the expertise of not only interventional cardiologists and their staff but also radiation oncologists and radiation physicists.
Treatment with the drug-eluting stent
Treatment of restenosis has complex economic implications: 1) it is expensive for patients and payers; and 2) it is a source of substantial revenue (and profit) for physicians and medical care facilities, particularly fee-for-service institutions, because of the return visits if restenosis is suspected or present and because of the subsequent angiography and interventions (either PCI alone, PCI with adjunctive vascular brachytherapy, or CABG), which are highly profitable. A new entrant—the drug-eluting stent—will dramatically change the treatment of coronary artery disease in terms of cost, quality, and value. At least two devices are well along in development and testing. For this discussion, we highlight one—the sirolimus-eluting stent—which was recently approved by the U.S. Food and Drug Administration (FDA) on the basis of two randomized, double-blinded, controlled, multicenter clinical trials (27–30): 1) the RAndomized Study With the Sirolimus-Eluting Bx VELocity Stent (RAVEL) trial (Cordis Corp., Miami, Florida), with 238 patients and a primary end point of angiographic restenosis at six months (late lumen loss); and 2) the Sirolimus-Eluting Bx VELOCITY Stent (SIRIUS) trial, with 1,100 patients and a primary end point of target vessel failure (cardiac death, myocardial infarction, or target vessel revascularization) at nine months.
The SIRIUS trial included somewhat higher risk patients and lesions (Table 2). There was a dramatic effect on angiographic restenosis both within the stent itself and within the treated segment in each trial, and patients treated with a drug-eluting stent had a significantly decreased need for subsequent intervention (Fig. 1) (27,30). Analysis of various clinical and angiographic subsets of the SIRIUS trial documented a remarkably consistent and concordant improvement in outcomes on the basis of angiographic and clinical end points with the drug-eluting stent. Per 100 patients treated with drug-eluting stents, 12.5 patients in the SIRIUS trial and 13.6 patients in the RAVEL trial avoided subsequent target lesion revascularization.
Drug-eluting stents have important implications, not only for patients but also for the entire health care delivery system. These implications were recognized by the U.S. Department of Health and Human Services and the Centers for Medicare and Medicaid Services in the final rule issued August 1, 2002, which created a new International Classification of Diseases-Ninth Revision (ICD-9) code for drug-eluting stents (code 36.07) and two new diagnosis-related groups (DRGs): drug-eluting stents with acute myocardial infarction (DRG 526) and without acute myocardial infarction (DRG 527). This step of creating a new ICD-9 code and new DRGs even before the use of a new device was approved by the FDA is unprecedented. It occurred for two reasons: 1) compelling clinical data; and 2) evidence that these stents may be no more expensive at the end of one year, because the additional initial expense may be offset by decreased subsequent interventions and the potential shift from CABG to PCI. In addition, many groups, including physicians and medical societies, supported the need for additional reimbursement to hospitals as close to the time of FDA approval as possible to blunt the effect of the increased cost of these devices.
Effective April 24, 2003, the use of the two new drug-eluting stent DRGs (nos. 526 and 527) will result in an additional payment to hospitals by Medicare of approximately $2,100 as an average base increase over the national average base payments by Medicare in fiscal year 2002 (October 2, 2001, to September 30, 2002) for bare-metal stents (DRGs 516 and 517). There is an additional adjustment to this differential based on geographic location, indirect medical expense, and disproportionate care of indigent patients. The result is a national average increase of 22% to 24% (range 0% to 50%), which varies by hospital. Thus, the average total additional reimbursement from Medicare is approximately $2,400 to $2,600. Because the base payment for bare-metal stents (DRGs 516 and 517) increased from fiscal year 2002 to fiscal year 2003 by approximately $300, the base additional payment for a drug-eluting stent within fiscal year 2003 is approximately $1,800 plus the applicable adjustment, as mentioned earlier (Fig. 2). The financial effect of drug-eluting stents on each hospital depends on four variables: 1) additional reimbursement per procedure by Medicare and other payers; 2) additional cost per stent; 3) number of stents used per procedure; and 4) number of procedures performed, which may change if the threshold for intervention changes. The financial impact will depend on the mix of private pay, health maintenance organization, and Medicare patients and the pricing of stents.
One might assume that these stents would be used in all eligible patients as soon as they are available. It is projected that the stents will be used in approximately 80% of eligible patients by the fourth quarter of 2003. A fundamental problem is that drug-eluting stents are approximately three times more expensive than bare-metal stents. This will be a particular problem if the number of stents used per patient increases substantially.
The paradigm shift
In the U.S., the central elements of the paradigm shift in the treatment of cardiovascular disease are related to patient, payer, physician, hospital, and societal expectations of health care financing.
Patients expect to have access to the latest technology, although they may have little knowledge of the cost or little willingness to pay for it. A new stent that appears to have no more adverse side effects than a conventional stent, but which leads to markedly improved outcomes with fewer subsequent revascularization procedures, will stimulate patient requests and increase expectations. In this era of widespread distribution of medical knowledge to the public and through the lay press and increased direct-to-consumer advertising by industry, the pressures from patients and families will most likely be intense. Attempts by physicians or hospitals to limit the use of drug-eluting stents on the basis of projected probabilities of restenosis may be received poorly by patients. Depending on the reimbursement from the patient's health care plan, a patient may be asked to pay more of the extra costs related to the stent itself. Although FDA approval will most likely be for the treatment of patients who have one lesion in one vessel (as in the randomized trials), drug-eluting stents will undoubtedly be used to treat multiple lesions in multiple vessels. Thus, the number of stents per patient may increase substantially and exceed the average number of stents used in patients, which is the number used in Medicare's calculation of the average prospective payment to hospitals.
There are several third-party payers for cardiovascular care (Fig. 3). All of these groups are concerned with rising health care costs, and the up-front higher initial costs of these new stents will be problematic. However, as is true for Medicare, these increased costs will be at least partially offset by a substantially reduced need for subsequent procedures, even if the third parties pay more for these devices. In addition, the rates of CABG, traditionally a much more expensive procedure, will decrease as more patients with multivessel disease, including left main coronary artery disease and diabetes, are treated with drug-eluting stents.
The predicted cost savings related to a reduction in restenosis rates may be rapidly offset by changes in practice, an unknown factor. These changes will occur as more patients are treated with stents instead of with medical therapy alone. This trend is expected to occur with the spread of knowledge about a safe therapy for the treatment of coronary stenosis with single-digit restenosis rates. As indicated previously, on the basis of the SIRIUS and RAVEL trial data, an additional 125 to 136 patients per 1,000 treated will avoid needing subsequent target lesion revascularization within six to nine months when treated with a drug-eluting stent. This situation is highly favorable for payers, and, even with the additional initial costs, it is likely to result in a cost-neutral or even cost-saving scenario.
The challenge for cardiologists will be to determine how to implement drug-eluting stents economically and efficiently. For surgeons, the problem will be substantially different because surgical volumes will probably decrease as more data become available to support the use of this therapy instead of CABG. Increasingly, surgeons will be consulted only about patients in whom these stents cannot be placed; also, the overall outcomes of surgery may worsen because of changes in case-mix as low-risk patients will receive interventional treatment and only high-risk patients will receive surgical treatment.
Hospitals will face major changes. The costs of PCI will increase because of the higher cost of stents; this increase will probably surpass the increase in the DRG payment. The economic impact on hospitals with respect to procedures performed on patients covered by private payers will depend on the contractual arrangements between the hospital and private payers. At the same time, cardiac surgery volumes will decrease, as will the number of percutaneous procedures performed to evaluate and treat restenosis. The costs of surgical procedures will also increase as sicker patients are treated, further eroding the margins of prospective payments. These changes will create a significant shortfall for many hospitals, particularly those with a large Medicare population.
From the societal standpoint, the use of drug-eluting stents should offer substantial advantages. Major surgical procedures will be replaced by less invasive procedures. Patients will need fewer hospitalizations and will have fewer recurrent symptoms. They may return to work considerably sooner, with enhanced productivity, than if major surgery had been required. Cost considerations are, however, complex: in the short term, this therapy is likely to be cost-neutral or even cost-saving; in the long term, if survival is enhanced, additional spending related to noncoronary disease may increase.
Clearly, drug-eluting stents will be received favorably by patients and physicians. However, hospitals may be somewhat less enthusiastic because of the potential for a negative effect on hospital budgets. The additional reimbursement from Medicare is an important first step to making this technology affordable. However, hospitals receive approximately half of their reimbursement from private payers, who reimburse hospitals on the basis of individual contracts that range from being advantageous for introducing new techniques (e.g., percentages of charges and adjustments for use of new procedures) to disadvantageous (e.g., per diem rates). Therefore, it behooves hospitals to review and restructure these contracts, when possible, to maximize the hospital's reimbursements for the use of new technologic advances. Hospitals will need to ensure that physician documentation is appropriate and that coding is correct.
The delivery of health care is increasingly complex. With new developments, paradigms of delivery must change. Drug-eluting stents are only the first of many new technologic advances that will affect cardiovascular care. Other examples of new cardiovascular device-based procedures that will affect the various constituencies differentially include carotid stenting with distal protection instead of surgical carotid endarterectomy, abdominal aortic aneurysm endografts instead of open aneurysm repair, and intravascular coil placement instead of open craniotomy for cerebrovascular aneurysms. It is difficult to ignore the great appeal of procedures that entail local rather than general anesthesia, less invasive approaches, shorter hospital stays, and more rapid return to work. These procedures have many features in common: 1) They will replace major surgical procedures with less invasive approaches after documentation that these procedures are safe and yield either the same or better outcomes than open surgical procedures. 2) Redistribution of costs will occur. Hospital profits are likely to decrease because the direct cost of the procedure will likely increase, and it may not be adequately reimbursed, while total costs for society (comprising direct and indirect medical and nonmedical costs, which include enhanced productivity due to more rapid return to work) may stay approximately the same or even decrease. 3) The use of new treatments, which patients will request, will be good for patients and society but may be problematic for hospitals and specific physician groups, especially surgeons, unless they become engaged in these new procedures.
The drug-eluting stent is only the most immediate and obvious example of a major paradigm shift in the device-based treatment of cardiovascular disease. The successful incorporation of drug-eluting stents by a hospital may therefore be a harbinger of how successfully a specific institution may assimilate other new technologic developments and manage the changing treatment paradigm. This example also has important implications for health care resourcing and for people who will provide the required training or credentialing of physicians, nurses, technicians, and other caregivers.
- Arterial Revascularization Therapy Study
- coronary artery bypass graft surgery
- diagnosis-related group
- Food and Drug Administration
- International Classification of Diseases- Ninth Revision
- percutaneous coronary intervention
- Randomized Study with the Sirolimus-Eluting Bx VELOCITY Stent trial
- Sirolimus-Eluting Bx VELOCITY Stent trial
- Received March 7, 2003.
- Revision received August 5, 2003.
- Accepted August 25, 2003.
- American College of Cardiology Foundation
- ↵Opinion page. Wall Street Journal 2003;Jan. 10:A-10
- American Heart Association
- The BARI Investigators
- Bypass Angioplasty Revascularization Investigation Investigators,
- Detre K.M.,
- Lombardero M.S.,
- Brooks M.M.,
- et al.
- the Emory Angioplasty versus Surgery Trial (EAST) Investigators,
- King S.B. III.,
- Kosinski A.S.,
- Guyton R.A.,
- Lembo N.J.,
- Weintraub W.S.
- Northern New England Cardiovascular Disease Study Group,
- Niles N.W.,
- McGrath P.D.,
- Malenka D.,
- et al.
- ARTS Investigators,
- Abizaid A.,
- Costa M.A.,
- Centemero M.,
- et al.
- ↵Centers for Medicare and Medicaid Services Data Compendium. Available at: http://www.cms.hhs.govAccessed December 30, 2003
- Co-Investigators of the National Heart, Lung, and Blood Institute's Percutaneous Transluminal Coronary Angioplasty Registry,
- Detre K.M.,
- Holmes D.R. Jr..,
- Holubkov R.,
- et al.
- Holmes D.R. Jr..,
- Savage M.,
- LaBlanche J.-M.,
- et al.
- Seshadri N.,
- Whitlow P.L.,
- Acharya N.,
- Houghtaling P.,
- Blackstone E.H.,
- Ellis S.G.
- Farb A.,
- Sangiorgi G.,
- Carter A.J.,
- et al.
- Mehran R.,
- Dangas G.,
- Abizaid A.S.,
- et al.
- the ERACI II Investigators,
- Rodriguez A.,
- Bernardi V.,
- Navia J.,
- et al.
- Popma J.J.
- RAVEL Study Group,
- Regar E.,
- Serruys P.W.,
- Bode C.,
- et al.
- RAVEL Study Group,
- Serruys P.W.,
- Degertekin M.,
- Tanabe K.,
- et al.
- Leon MB, Moses JW, Popma JJ, Kuntz RE, Fitzgerald P, for the SIRIUS Investigators. SIRIUS: a U.S., multicenter, randomized, double blind study of the sirolimus-eluting stent in de novo native coronary lesions. N Engl J Med 2003;349:1315–23