This article is more than 5 years old. Information may no longer be current.
Technology reimbursement and CMS: The struggle to deliver ‘relevant’ outcomes in orthopedics
Click Here to Manage Email Alerts
Peter J. Neumann and Sean R. Tunis titled their recent controversial New England Journal of Medicine article; “Medicare and medical technology — The growing demand for relevant outcomes.” In this paper, the authors argued that a move to higher requirements for evidence in payment decisions for new technology is “necessary and beneficial.” They also highlighted several buzzwords that likely represent the new gold standard in medical evidence and reimbursement policy: comparative effectiveness; cost effectiveness; relevant populations; and coverage with evidence development. Several concerns regarding this situation were acknowledged, including impeding access to important medical advances, interference by “big government” in the doctor-patient relationship, and a cost-containment agenda. Notably lacking from the discussion was any suggestion of how the expense of higher evidence level studies should be funded.
To reflect on the implications of this new approach to reimbursement policy, we have assembled a Round Table of orthopedic surgeons with expertise of both the clinical and research aspects of these issues.
David A. Wong MD, MSC, FRCSC
Moderator
|
||||||||
David A. Wong MD, MSC, FRCSC
Kevin J. Bozic MD, MBA
Steven D. Glassman, MD
Julie A. Dodds, MD
Andrew N. Pollak, MDDavid A. Wong MD, MSC, FRCSC: Do the clinical aspects of your subspecialty lend itself to the level of research suggested in the article?
Kevin J. Bozic MD, MBA: Prospective, randomized controlled clinical trials (RCTs) are the gold standard study design for evaluating the efficacy — defined as the extent to which the intervention improves health under ideal circumstances — of a specific treatment intervention or technology. However, the effectiveness — defined as the extent to which an intervention improves health in real practice settings — of many interventions and technologies may, and often does, differ from its efficacy, and therefore the results of RCT’s may not be generalizable to the larger population.
For interventions like total joint arthroplasty, which is a commonly performed procedure whose outcomes are measured over a period of decades, RCTs may not be the most appropriate study design to evaluate the comparative effectiveness of new techniques and technologies for use in clinical practice. Clinicians may be skeptical as to whether or not the practice environment and patients included in a RCT are similar and therefore generalizable to their own practice.
Clinical registries and administrative claims databases can provide useful insights into the effectiveness of total joint arthroplasty procedures and implants. Although they are limited by the amount of clinical detail that can be included, and in some cases, the clinical specificity of the data, particularly true for administrative claims databases, they have the advantage of including large numbers of patients from diverse demographic, geographic, and socioeconomic settings, which enhances the generalizability of the results and conclusions derived from research studies which utilize these data sources.
Julie A. Dodds, MD: In sports medicine, our outcomes are often based on ability to return to sport or an active lifestyle, which while it definitely has an influence on quality of life, is hard to quantify as such. Many of our technical advances, such as arthroscopic rotator cuff repair, may be effective in areas such as postoperative pain relief and have short term advantages, but may not change the big picture when subjected to outcomes research.
Steven D. Glassman, MD: Spinal diseases, particularly lumbar and cervical degenerative disorders, are remarkably common and the array of potential treatment options seems almost unlimited. This represents a major challenge for study design, but also underscores the critical need for comparative effectiveness and cost effectiveness research. The range of pathology also contributes to what is perhaps the greatest barrier to comparative effectiveness research (CER), the lack of a useful diagnostic classification or coding scheme for lumbar spinal disorders. This deficiency severely restricts the present usefulness of large administrative databases for all but a small group of diagnostic entities.
Andrew N. Pollak, MD: Comparative effectiveness research, by its very nature, produces the type of information that allows clinicians to make the best possible decisions about their patients’ health care. This is just as true in the world of orthopedic trauma as it is in any other subspecialty. Unfortunately, good high quality CER is not feasible for every permutation of every injury pattern that orthopedic trauma specialists are forced to treat.
Certain injury patterns are common and lend themselves to rigorous scientific investigation into optimal treatment. Unfortunately, good high-quality research definitively demonstrating the superiority of one form of treatment over another still is not always easy. An excellent example of this taken to the extreme is the treatment of femoral shaft fractures. In the United States and in most parts of the world in 2010, the treatment for femoral shaft fractures in most patients is closed intramedullary nailing. We know the results of this treatment are good and that with this treatment, patients can be expected to resume high level activity. We also know that closed management of these fractures with skeletal traction is associated with prolonged morbidity and long-term disability in the form of leg length discrepancy and malalignment with resultant poor functional outcome. Unfortunately, or fortunately, when closed femoral nailing first became commonly utilized for treatment of these injuries, nobody performed a prospective, randomized comparison of short- and long-term outcomes after treatment of femoral fractures with nailing vs. skeletal traction. Today, if somebody were to try and do such a study, it would likely be impossible to recruit patients and difficult to convince an IRB that such a study was ethical even given the lack of any level 1 evidence demonstrating that the nailing is better than traction. The reason is that it is inherently obvious that patients treated with nails do well and that they do much better than anything that was ever historically demonstrated in the population of patients treated with traction years ago.
A similar problem with expecting CER to answer our questions. with regard to appropriate treatment. exists in situations where we actually have level 1 evidence comparing two forms of treatment. For example, the management of displaced diaphyseal clavicle fractures: There is now level 1 evidence available comparing the effectiveness of operative vs. nonoperative management of these injuries concluding that patients treated operatively did better than those treated nonoperatively. Nonetheless, careful examination of those results suggests that they are applicable only to a narrow population of patients with clavicle shaft fractures and that the actual differences in outcome between the operative and nonoperative groups, while statistically significant, are relatively small.
Thus, while the idea of using high-quality evidence to direct decisions about treatment is theoretically a good one, in many common and uncommon situations in orthopedic trauma, it is neither possible nor feasible.
Wong: Are there sufficient sources of research funding available in your subspecialty for CER/outcomes research?
Dodds: Unfortunately not. In spite of the generosity of industry and private donors, these studies tend to be very costly. The Arthroscopy Association of North America (AANA) is currently undertaking a large study to determine arthroscopic complications, and we are realizing how labor intensive and costly these studies can become.
Glassman: I suppose that the obvious answer is that funding is never sufficient, but lack of funding is certainly not the only factor limiting CER in spine. A particular problem in spine research is that much of our standard clinical practice consists of “off-label” use of medical devices. The restrictions on research funding for widely used, but off-label, interventions such as pedicle screw instrumentation for pediatric deformity or lateral mass screws for cervical trauma poses a significant hurdle to necessary evidence development.
Pollak: The way to obtain more information that will allow us to make better evidence-based decisions with regard to selecting treatments for our patients is to do more scientifically rigorous studies comparing treatment options for various commonly encountered injuries. Unfortunately, such studies are extremely expensive. While there was talk about including “some money” for such studies in both the House and Senate versions of recently passed health care reform legislation, it is going to take a lot more than “some money” to make such studies happen. Rather than facilitating this research by making more money available, the federal government routinely erects extensive and expensive roadblocks to such research. A great example of such a roadblock is HIPAA. There is no question that compliance with HIPAA regulations increases the cost of doing any patient-related research. Furthermore, essentially all universities in the country, in remaining compliant with federal regulations, have adopted IRB procedures and policies that are not only onerous, but extremely expensive. Any department engaging in clinical research necessarily needs to hire one if not more staff personnel purely to address the regulatory challenges with obtaining and maintaining IRB approval for any ongoing prospective research.
Additional sources of research funding are definitely necessary in order to make good CER occur. Such sources of funding are far from readily available in any subspecialty of which I am aware.
Another important barrier to generating the comparative effectiveness data that impedes performance of good clinical research is the National Institutes of Health (NIH) salary cap structure. Under that structure, the maximum salary for an investigator in 2010 is $199,700. That number is grossly inconsistent with the salary of most orthopedic surgeons in academic centers who are performing research. Thus, for every hour of a surgeon’s time devoted to NIH-funded research, his department or his family (or perhaps both) cost shares (an equivalent term to loses) money. This is a recipe for driving clinician scientists out of the scientific world and into the clinical world. If the Centers for Medicare & Medicaid Services (CMS) and other government entities are serious about encouraging clinician scientists to do such research, they must encourage the agency that is going to fund such research to eliminate the salary caps and allow research grants to cover real investigator costs for participation.
Wong: What are the technologies/procedures that you feel would be appropriate for CER in your area of interest?
Bozic: Total joint arthroplasty procedures have been shown to be highly effective in terms of reducing pain, improving function, and quality of life in patients who suffer from hip and knee degeneration. The implant design concepts originally put forward by pioneers like Sir John Charnley and John Insall have withstood the test of time. Therefore, any incremental changes in implant design have a relatively high barrier to overcome in order to demonstrate incremental improvement in clinically meaningful outcomes. One example would be hard-on-hard bearing surfaces used in total hip arthroplasty. Although they have been shown in the laboratory to reduce bearing surface wear rates compared to metal-on-polyethylene bearings, there is no evidence that they are associated with lower revision rates compared to conventional implants. These types of incremental design innovations lend themselves to CER.
Dodds: One procedure that I can think of is rotator cuff repair. In patients who get are alleviated of their night pain and are able to return to doing overhead activities, it would be fairly straightforward to demonstrate an improvement in quality of life. An emerging technology amenable to outcome research may be biologics, such as platelet-rich plasma, which may increase healing rates and thus, increase long term outcomes.
Glassman: An important criticism of the spine literature is that most level 1 studies compare one surgical technique to another, but few studies compare surgical management to nonsurgical alternatives. In fairness, it is difficult to design and execute randomized studies of surgical vs. nonsurgical treatment because in most cases failure of nonsurgical treatment is a prerequisite for surgical selection. At this point, it appears that the development of a spine registry is the most promising mechanism to facilitate comparison of surgical and nonsurgical modalities. Within the realm of surgical treatment, one important question is; for which specific patient will the addition of an interbody fusion (ALIF or TLIF) generate a better outcome than posterolateral fusion alone?
Pollak: For some new technologies, the results are so obviously superior to those associated with older techniques that massive changes in practice occurred before high-quality comparative effectiveness studies could be done. An example of this is the use of locked plates for treatment of distal femoral, proximal tibial, and some other perimetaphyseal and periarticular fractures. The new technology was so clearly superior from a clinical standpoint that surgeons did not wait for comparative effectiveness data to become available. Obtaining it now would be impossible because an inadequate number of surgeons would be available to demonstrate equipoise with the older techniques. Thus, had data been initially required to support the transition from the old to the new in the context of increased expense for the new implants, patients would have been blocked from access to the new implants and progress would have been impeded.
Wong: Should medical or subspecialty societies play a significant role in organizing CER studies?
Bozic: The American Academy of Orthopaedic Surgeons (AAOS) has made a substantial investment of both human and financial resources in the area of comparative effectiveness research. The Evidence Based Practice Committee (EBPC) and the Guidelines and Technology Oversight Committees (GTOC) are actively involved in evaluating and synthesizing the evidence related to specific diagnostic and therapeutic treatment algorithms and technologies. The practice guidelines and information statements put forth from these committees have been helpful to many stakeholders, including surgeons, hospitals, policymakers, and most importantly, our patients.
The formation of the American Joint Replacement Registry (AJRR) is another example of AAOS’ involvement in comparative effectiveness research. The AJRR will eventually provide the largest body of information regarding the comparative effectiveness of joint replacement procedures that exists in the world. As such, I believe AAOS’ resources are better spent organizing CER, for example AJRR, and synthesizing the available information, for example EBPC/GTOC, rather than sponsoring investigator-initiated research projects and/or performing CER in-house.
Dodds: I think it is important for organizations to encourage and possibly organize outcomes-based research. Payments for future surgeries will be highly dependent on these studies.
Glassman: For spine, there is a broad group of stakeholders including AAOS, the neurosurgical societies and spine-specific Societies such as North American Spine Society (NASS) and Scoliosis Research Society (SRS). Also, a really comprehensive effort should involve a partnership with nonsurgical groups representing family practice, pain management and physiatry. While it is not necessary that these groups collaborate on individual studies, a common set of metrics would greatly facilitate the overall goal of evidence-based decision making. In an effort to promote this concept, an Agency for Healthcare Research and Quality (AHRQ) funded multi-stakeholder conference is scheduled for July 2010 in Madison Wisconsin. Additionally, the above mentioned surgical societies are in the early stages of discussion regarding a combined effort on registry development.
Pollak: Medical societies such as the AAOS and subspecialty societies such as the Orthopaedic Trauma Association (OTA) must play a significant ongoing role in convincing Congress and the administration of the necessity for providing adequate funding for CER across the entire spectrum of orthopedic surgery. The levels of research funding that become available need to be proportionate to the burden of disease. The musculoskeletal system represents the largest source of long-term disability for both the young and old, yet the amount of research funding available for orthopedics is grossly disproportionate to that of other areas.
Wong: Kevin, you have published several papers in the CER arena. For new technology, should cost effectiveness or clinical effectiveness, or a combination, be the major criteria for approval/reimbursement?
Bozic: When evaluating any new health care technology, it is important to consider the potential clinical risks and benefits compared with the gold standard technology. However, it is also important to recognize that those results are achieved at a cost. Ultimately, the value of any new intervention or technology is a function of the improvement in patient outcomes relative to the total — both initial and downstream — cost of attaining those outcomes.
The consideration of cost in the evaluation of the comparative effectiveness of new health care technologies has drawn a great deal of attention and criticism from various stakeholders. However, given rampant and uncontrolled health care inflation in the United States, it is naïve to think that cost cannot or should not be explicitly considered. If cost is not explicitly considered by providers and patients when evaluating new treatment options, the inevitable rationing of finite health care resources will force implicit consideration of cost on the back end.
Wong: Julie, as you know, I trained in Toronto when Bob Jackson brought the Watanabe 21 arthroscope to North America. Would orthopaedic surgeons be performing arthroscopy today if that new technology had to meet the criteria suggested in the article?
Dodds: Probably not. While the arthroscope was truly revolutionary, it simply made many of the procedures that were previously performed doable with less morbidity and more easily. Comparative effectiveness research does not usually recognize these types of benefits.
Wong: Steve, establishing Minimum Clinically Important Difference (MCID) in clinical research has been an interest of yours. Do we have well-recognized criteria in this area as yet?
Glassman: Historically, statistical significance implied that a finding was clinically important, but as studies grow larger very small changes in outcome can become statistically significant without being clinically relevant. Therefore, defining thresholds of clinical importance has become necessary, and the MCID is the most widely used threshold. For any given measure, MCID represents the smallest change that a patient recognizes as being clinically significant. Conceptually, this might be thought of as the amount of change above the statistical noise. While MCID is an important benchmark, it is certainly not an ideal target for clinical success. More recently, we have identified a threshold of Substantial Clinical Benefit (SCB) which represents a change that patient recognizes as a major rather than a minimum level of improvement.
Although there may never be a single measure of success, there is increasing agreement that Health Related Quality of Life (HRQOL) measures are a critical component of the evaluation. Moving forward, we are likely to see a less familiar measure, health utilities, play an increasingly important role. Health Utilities, such as EQ-5D and SF-6D, include an assessment of patient preference as to the relative value of any given intervention. Health utilities allow the calculation of Quality Adjusted Life Years (QALYs) which serve as the basis for cost effectiveness analysis.
References:
- Neumann PJ, Tunis SR. Medicare and medical technology–The growing demand for relevant outcomes. New Engl J Med. 2010; 362(5):377-379.
- Kevin J. Bozic, MD, MBA, is associate professor and vice chair of the Department of Orthopaedic Surgery at the University of California, San Francisco and Core Faculty, Philip R. Lee Institute for Health Policy Studies. He can be reached at 500 Parnassus Ave. MU-320W, San Francisco, CA 94143; 415-476-3900; e-mail: BozicK@orthosurg.ucsf.edu
- Julie A. Dodds, MD, can be reached at Sparrow Health Science Pavilion, 2900 Hannah Blvd., Suite 212, East Lansing, MI 48823; 517-319-1831; e-mail: Julie.Dodds@hc.msu.edu.
- Steven D. Glassman, MD, can be reached at Department of Orthopaedic Surgery, University of Louisville School of Medicine, Louisville, KY 40202; e-mail: steven.glassman@nortonhealthcare.org.
- Andrew N. Pollak, MD, can be reached at University of Maryland School of Medicine, 22 S Greene St., Room 3TR54, Baltimore, MD 21201-1544; 410-328-6280; e-mail: apollak@umoa.umm.edu.
- David A. Wong, MD, can be reached at Denver Spine Center, 7800 East Orchard Road, Suite 100, Greenwood Village, CO 80111; 303-783-1300; e-mail: ddaw@denverspine.com.