In 2007, the British Medical Journal reported the fifteen most important medical milestones since the inception of that journal in 18401. Included were the discovery of DNA, the development of vaccinations and of antibiotics, the use of anesthetics for surgery, and evidence-based medicine. The practice of evidence-based medicine can be conceptualized as the integration of the best available research evidence, the clinical circumstances, and the values and preferences of the patients. Clinical expertise is critical to the practice of evidence-based medicine, allowing the sensible and skilled application of best evidence to patients. The use of best evidence implies a hierarchy. This hierarchy, described by Sackett et al.2 in 1996, can be visualized as a pyramid of evidence with randomized trials on top (Level I) and expert opinion on the bottom (Level V). For a therapy, Level-I evidence obtained from randomized controlled trials remains the highest standard for the most valid information.
Yet, there is a discrepancy between the apparent validity of medical knowledge and the actual facts. Only about 15% of medical interventions are supported by verifiable scientific data3. Even when evidence is available, physicians are often slow to adopt changes in thinking. There have been ten trials of bed rest after spinal puncture that have demonstrated no change in headache with bed rest and an increase in back pain. However, 80% of the protocols in neurology units in the United Kingdom continued to recommend bed rest after lumbar puncture as of 1998, despite earlier evidence-based reports4.
In the case of orthopaedic medicine, any drug or implantable device to be used in humans should logically undergo the same scrutiny of evidence before widespread application. It would seem appropriate that high-quality confirmatory evidence about safety and efficacy from randomized trials be obtained prior to patient application. This approach is standard in the drug development cycle with a series of phases (Phases I through IV) with early safety and dosing studies followed by pivotal randomized trials. Regulatory approval of a drug in orthopaedics mandates a Phase-III randomized controlled trial.
Currently, however, the regulatory approval process for devices diverges substantially from that of drug development. The U.S. Food and Drug Administration (FDA) 510K pathway is a process that includes the need to demonstrate equivalency to a prior approved device. Without such approval, additional evidence of safety and efficacy is required. However, under a pathway termed the 510K approval mechanism, if the new device is considered to be substantially equivalent to an existing device, it does not require any additional evidence for safety and efficacy, and it will be regulated the same way as the existing device. In principle, this simply means that a new device that is biomechanically equivalent to an existing device can be marketed and implanted in patients without any further clinical evidence of safety or efficacy. Therefore, in the hierarchy of evidence for regulatory approval, drug trials are typically held to Level-I randomized trial standards, whereas devices may in some cases require only Level V (experimental evidence). However, a leading journal in the field has begun to address the issue of identifying levels of evidence more directly for orthopaedic studies. In 2000, an evidence-based medicine section that appeared quarterly was added to The Journal of Bone and Joint Surgery (American Volume). Then, in 2003, all published articles were required to have a level-of-evidence rating, as follows: Level I indicated a randomized controlled trial; Level II, a prospective cohort study or lesser-quality randomized controlled trial; Level III, a case-control study or retrospective cohort study; Level IV, a case series; and Level V, expert opinion. Even so, the role of and controversy surrounding evidence-based medicine will continue to increase over the coming decades.
The aim of this symposium, therefore, is to elucidate some of the critical issues related to the development and regulation of orthopaedic devices with use of evidence-based medicine.
On June 7, 2008, at a combined meeting of the American Orthopaedic Association (AOA) and the Canadian Orthopaedic Association (COA), a survey of the participants at a clinical symposium on evidence-based medicine in orthopaedics indicated some of the prevalent perceptions and habits of orthopaedic surgeons regarding the development, regulation, and application of orthopaedic devices in their clinical practice5. While a majority of surgeons believed that an evidence-based medicine framework was well suited to best evaluate both the safety and efficacy of new orthopaedic devices (see Appendix), over half did not think that they had adequate information to make informed decisions regarding the use of new technologies in their practice (see Appendix). Additionally, over 80% felt pressure from their own institution to be more cost conscious in using new technologies in their clinical practice (see Appendix) and that many orthopaedic devices were overpriced and did not do much to improve patient conditions (see Appendix). As a result of the symposium, a large majority thought that they were now better informed as to evidence-based medicine in orthopaedics, and 87% believed that the AOA should explore this topic further as a critical issue. Consequently, orthopaedic surgeons in North America believe that there is a need for more standardized assessment of new technologies for their own clinical practices, which may be assisted by the combined efforts of the AOA, the COA, and device manufacturers.
The Importance of Preclinical Data
Obtaining relevant and repeatable preclinical data is a key element in the product development cycle that aims to bring new and improved treatments to the marketplace. Despite the fact that the product development cycle has evolved over the last several decades and that many transformative products and technologies have been brought to the market, the question "How much experimental data are needed to move to clinical studies?" is quite difficult to answer in the general case. The amount and nature of experimental preclinical data required prior to moving to clinical studies will vary as a function of the disease being treated, the availability of treatment modalities, and the novelty of the technologies being employed. Throughout the product development cycle, patient welfare must always be the first consideration: Are the products safe and free of undue risks to the patients? Moreover, it must be determined whether the products are effective in treating the target disease.
Classifying the Challenges
Modern preclinical testing protocols can often provide answers to many of the safety questions sought. To a large extent, this is why short-term failure events are rarely seen following the introduction of new orthopaedic devices. Preclinical testing modalities include wear testing, biomechanical tests, electrochemical tests for corrosion, tissue culture studies, animal models, and finite element analysis. However, it is important to differentiate between the requirements for preclinical testing in an established technology compared with a new technology in which there is limited clinical experience and, more importantly, limited knowledge of the relevant failure mechanisms. For example, the FDA employs a paradigm in which products are classified according to their complexity and risk. In this scheme, Class-1 products are common low-risk devices that require general regulatory controls, such as labeling requirements and conformance with Good Manufacturing Practice regulations. Class-2 devices are more complex and carry a higher risk. These require both general and special controls (which may include performance standards or postmarket surveillance) as well as premarket notification (510K). Preclinical testing is often an important component of the regulatory pathway for Class-2 devices. Class-3 devices are the most complex and highest risk devices. To bring these to the marketplace, extensive preclinical data are required as well as a premarket application. Furthermore, it is important to differentiate between diseases with currently available effective treatments compared with diseases with only ineffective treatments available. In the former circumstance, there would generally be less tolerance for risk in the introduction of new devices, whereas in the latter case, a greater degree of risk is tolerable.
Medical Device Standards and Guidelines
It is important to emphasize the role of standards in preclinical testing. Consensus standards developed by organizations such as the ASTM International (ASTM) and the International Organization for Standardization (ISO) are particularly valuable in establishing reproducible protocols for such testing. These standards6 are developed by committees that include health-care practitioners, design engineers, and other interested parties. These standards specify the composition of materials used in medical devices as well as the methodologies for preclinical testing. By standardizing test methodology, a direct comparison of the results obtained in different laboratories is possible. The use of standardized test methods will also minimize potential bias. Through the Food and Drug Administration Modernization Act of 19977, the FDA now recognizes standards in the regulatory approval process that certainly can facilitate and accelerate the product development cycle.
Orthopaedic device standards are generally developed by a consensus process. While this can lead to a "softer" standard initially, these standards are continuously revised and updated on the basis of advancements in the field and improvements in the understanding of the clinical performance of the devices. Generally, orthopaedic device standards follow the clinical introduction of a device. However, in certain circumstances (e.g., tissue-engineered medical products), these standards can actually anticipate the clinical introduction of a new technology. Within ASTM, there are numerous standards that cover osteosynthesis devices, biocompatibility testing, metallurgical materials, ceramic materials, polymeric materials, material test methods, medical and surgical instrumentation, joint replacement devices, spinal instrumentation, and tissue-engineered medical products6. The most relevant and useful standards to help to conduct preclinical testing are developed in conjunction with orthopaedic surgeons with knowledge of the performance of the devices, as well as knowledge of the common failure mechanisms. In addition, implant retrieval studies can greatly inform the standards development process. The ISO also provides consensus standards that can be useful in preclinical testing.
The FDA has developed guidance documents that can facilitate the regulatory review process8. These documents are based on evaluation of the extant literature and on the experience of the agency in regulating substantially similar products. They are intended to be scientific position papers that suggest important evaluation criteria, preclinical testing procedures, and end points that are believed to be necessary to provide reasonable assurance of substantial equivalence and/or safety and effectiveness. Manufacturers find these guidance documents to be helpful in order to negotiate the product approval process. In fact, there are internal FDA data showing that guidance documents can shorten the review times for devices introduced into the marketplace9.
Limitations of Preclinical Data
Preclinical testing has often been highly effective in preventing short-term failures and in predicting long-term behaviors of medical devices. However, this was not always the case, especially when the nature of the failure mechanisms of the device was unknown or could not be predicted and because relevant input parameters were only incompletely understood. In many of these circumstances, appropriate preclinical testing may have predicted the unsatisfactory outcomes; however, this was only known retrospectively after the clinical introduction of the device and a thorough failure analysis was conducted.
How much experimental data are needed to move to clinical studies? For established, mature technologies in which failure mechanisms are well understood, a fairly limited panel of preclinical testing, which can be based on consensus standards that have been developed and refined over years of clinical use, will be required. However, for newer technologies for which there is limited clinical experience and no (or only anticipatory) standards, the scope of the preclinical testing required will of necessity be greater. Furthermore, the bar for introducing new technology should be higher for disease entities for which there are successful and established technologies (e.g., treatment of osteoarthritis in the knee) than it is for diseases or injuries for which there are less effective treatments available (e.g., the treatment of complex wartime extremity injuries).
Finally, while preclinical testing has a key role to play in the product development cycle, these tests are no substitute for clinical trials. It is only through clinical experience that a full understanding can be gained of the parameters governing implant performance and the relevant failure mechanisms.
The default position regarding evidence for many clinicians is generally that a randomized controlled trial is the gold standard of evidence for the introduction of new medical devices or interventions. A randomized controlled trial is certainly an appropriate requirement when the efficacy of the control is unknown or hard to define or measure, when the intervention has high cost or risk, and when acceptable alternatives exist. On the other hand, a pivotal case series may be an acceptable substitute for the randomized controlled trial when the efficacy of the control is well defined, the new intervention has low risk, and/or the alternatives are suboptimal.
There are at least two situations when a pivotal case series is likely to provide adequate evidence to introduce a new medical device: first, when the intervention is not expensive, it has low risk, or there is no good alternative; and, second, when the efficacy of the comparison and/or control intervention is very clearly established and easily measured.
Conversely, there are two situations when a pivotal case series is likely not to provide adequate evidence to introduce a new medical device: first, when the intervention is costly, has risk, or could be less effective than current interventions; and, second, when the efficacy of the comparison and/or control intervention is unclear or poorly defined or measured. One recent example is the case of platelet gel concentrates that were marketed to be used to enhance local bone or iliac crest bone-healing in spinal fusions. There were limited animal studies performed, often not simulating the exact use intended in humans. These products were aggressively marketed as so-called natural growth factors and as cheaper alternatives to well-documented products containing bone morphogenetic proteins (BMPs) such as demineralized bone matrix and recombinant BMPs. The presence of BMPs in platelet gels was not documented, yet often implied to surgeons. Despite early case series claiming improved healing success and handling properties of the graft, several controlled series demonstrated not only a lack of enhancement of autograft healing but also an inhibition of spine fusion healing, ranging from 14% to 29%10-12.
Total hip replacement has revolutionized the care of patients with end-stage hip arthritis, providing durable pain relief and restoring function13. Unfortunately, not all attempts to improve total hip replacement outcomes have been successful and, in some instances, patients have been harmed (i.e., titanium-alloy femoral heads, Boneloc bone cement, carbon-reinforced polyethylenes, and metal-on-polyethylene surface replacements)14-18. The question is: How can these disasters be prevented? Malchau proposed a stepwise introduction of new orthopaedic technologies, including preclinical testing, randomized clinical trials, multicenter studies, and postmarket surveillance19. The strengths and weaknesses of randomized controlled trials and postmarket surveillance (national joint replacement registries) as tools to provide safety and efficacy data on novel orthopaedic implants should be recognized.
Randomized Controlled Trials
Chalmers et al. clearly demonstrated the power of randomized controlled trials20. In a meta-analysis to compare new and conventional treatments of acute myocardial infarction, with death as the end point, they found that the beneficial effect of the new treatment depended more on the study design than on the medical treatment. When new treatments were compared with historic controls, they found a 58% improvement, which decreased to 24% when nonblinded, randomized controlled trials were used and further diminished to only 9% when a double-blind, randomized controlled trial was employed.
In the case of assessing new orthopaedic devices, consider, for instance, a prior investigation that utilized a blinded, randomized controlled trial to compare cemented and cementless fixation of total hip replacements13,21,22. In the study, 250 patients were randomized to receive either a cemented or cementless total hip replacement and were stratified by implant type (cemented or cementless), age, and surgeon. Patients were assessed by a single observer, using validated disease-specific, patient-specific, global health, functional capacity, and cost-to-utility outcome tools. Highly significant and long-standing health-related quality of life improvements were noted for both cemented and cementless stems. Both cemented and cementless implants provided highly cost-effective treatments comparable with other medical and surgical interventions. Kaplan-Meier survivorship revealed that the cementless total hip replacement studied had significantly better ten-year survivorship than the cemented counterpart.
The first question is whether one can generalize the outcomes of this study to the many other cemented and cementless total hip replacements introduced to the marketplace. The answer is probably no. The second question is whether the outcomes of the randomized controlled trial are relevant after ten or more years of clinical follow-up, considering the advent of new procedures and technologies. Once again, the answer is no, since in the study only the cementless femoral stem was still in use by the ten-year follow-up mark. The metal-backed cemented socket, cementless socket, gamma-in-air-sterilized polyethylene, ion-implanted titanium femoral heads, and titanium-alloy cemented stem had all been removed from the market, but not until tens of thousands of patients had these implants inserted.
National Joint Replacement Registries
Postmarket surveillance with use of national joint replacement registries to monitor the performance of hip and knee replacements originated in Sweden19. The concept behind a registry was to link the implants, surgeons, and hospitals with large numbers of hip and knee arthroplasty patients, recognize revision trends, provide feedback, and monitor the subsequent decision-making of orthopaedic surgeons. As a result, Sweden has very low crude revision rates (i.e., the annual number of revisions divided by annual number of revision and primary procedures), which are the envy of the rest of the world.
To be successful, national joint replacement registries should ideally track all primary and revision procedures to ensure that all revisions are truly captured. Various national registries have had varying levels of success with complete data capture. Voluntary surgeon participation, the need for patient consent, confidentiality issues, medicolegal concerns, and the burden of extra work for the surgeon have been impediments that need to be worked through. Countries that have mandated registry data as a quality assurance tool and have legislated the necessary protections to cover confidentiality and medicolegal issues have had the most success. In addition, many of these successful registries have been able to bypass the surgeon for data collection, capturing hospital data that link the patient with a surgeon, implants, and operating-room environment. The resulting minimum data set, such as patient identifier, sex, age, diagnosis, body mass index, anesthetic type, comorbidity, implant bar codes, and operating-room environment, has been the most successful.
While a minimalist registry in terms of data capture has been the most successful model, it has been recognized that revision of a component is a rather crude outcome measure, excluding health-related quality-of-life outcomes and radiographic assessments. Several attempts have been made to add these assessment tools to national joint replacement registries. The use of additional outcomes tools such as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)23 and Jaeschke et al. transition ratings in a registry can also be considered24 (Fig. 1).
The American and Canadian health-care systems face many challenges that threaten their long-term viability and sustainability. Changing population demographics, rising costs, and highly variable quality have all been cited as causes for concern and declining public confidence in the health-care systems. Orthopaedic surgery has played a central role in the health-care debate. Increases in procedure volumes, the continual introduction of high-cost and relatively unproven devices into the marketplace, and wide regional variations in practice patterns25,26 have led to increased public scrutiny of orthopaedic surgery by hospitals, payers, public reporting agencies, and the government27,28.
Part of the Problem or Part of the Solution?
An important philosophical debate centers on the question of whether technology is part of the problem or part of the solution to the current health-care crisis. Certainly, there are many examples of technologies whose introduction has improved the quality of life for orthopaedic patients, including the discovery of penicillin, the poliomyelitis vaccine, and the Charnley low-friction arthroplasty. However, many of the technologies used by orthopaedic surgeons today lack evidence regarding their comparative effectiveness vis-à-vis established gold-standard technologies. Furthermore, there is widespread evidence that adoption and use of new technologies is one of the primary drivers of rising health-care costs29.
Overshooting the Need
Another concern among health-care policy makers is that, in many cases, new orthopaedic technologies may have "overshot" the needs of the average orthopaedic patient. As Christensen et al. demonstrated, the functionality of today's health-care technologies, although impressive, often outstrips the ability of patients to absorb it30. An example of this is seen with so-called high-flexion knee replacements. These technologies were developed to improve knee kinematics and minimize wear and impingement in patients who are able to flex the knee beyond 130° of motion. Although this design represents an improvement over conventional total knee replacement designs biomechanically, the most reliable predictor of postoperative range of motion following total knee replacement is preoperative range of motion31. Since only a small minority of patients with end-stage osteoarthritis of the knee who are considered candidates for knee replacement have a preoperative range of motion of >130°, most patients managed with a total knee replacement would not be able to realize the theoretical benefits of high-flexion knee implants32. However, they have been aggressively marketed to patients and surgeons and have achieved substantial market penetration at an important price premium over conventional total knee replacement designs, despite a lack of evidence regarding any true benefits in terms of improved patient outcomes.
Most so-called advances in orthopaedic technology are evolutionary, rather than revolutionary. This opinion is supported by the fact that the orthopaedic device industry spends roughly 4% to 5% of annual revenue on research and development, which is substantially less than that spent by the rest of the medical device sector (10% to 12%) and pharmaceutical industry (15% to 20%) on research and development. Furthermore, many high-cost orthopaedic devices are introduced into the marketplace with little if any data regarding their comparative effectiveness vis-à-vis the gold-standard technology. An example is the widespread adoption of locking plates for use in the operative treatment of routine fractures, despite a dearth of evidence regarding any benefit over conventional plates, which are much less costly.
Moral Hazard
Another problem that contributes to rising costs in both the American and Canadian health-care systems is what health-care economists refer to as the so-called moral hazard. This means that the primary decision makers, namely physicians and patients, are relatively price-insensitive, since they are not financially responsible for the care they provide or receive. An example of this can be seen in the field of total joint arthroplasty, where implant costs have risen steadily over the past fifteen years33, with no demonstrable corresponding improvement in patient outcomes. This is distinctly different from other technology-intensive industries, such as the personal computing industry, where the cost of technology has decreased over time, while the functionality of the technology has improved dramatically over the same time period.
The high cost of orthopaedic care, largely driven by the widespread adoption and use of new, unproven orthopaedic technologies, has created friction between orthopaedic surgeons and other health-care stakeholders. In response to declining profit margins in their traditionally profitable orthopaedic service lines, hospital administrators have stepped up their efforts to limit the use of new orthopaedic devices in clinical practice. Both government and private payers have labeled many new orthopaedic devices and procedures as experimental or investigational, and thus denied payment for these services. Public policy makers have increased their scrutiny of orthopaedic providers through public reporting of both quality and so-called efficiency measures related to orthopaedic procedures. Also, recently, the U.S. Department of Justice and the media have called into question the relationships between orthopaedic surgeons and orthopaedic device manufacturers27,28.
Orthopaedic surgery will play a prominent role in health-care delivery and reform in the next decade. An aging population and an increased emphasis on bone and joint health will continue to contribute to increasing costs and procedure volumes. Innovation and improvement in the quality of care to patients must continue. However, a higher level of evidence must be demanded when the adoption of new technologies into clinical practice is considered. It is important for orthopaedic surgeons and their patients to recognize that newer is not always better. As Emanuel et al. pointed out, the value of an intervention or a device resides not in its newness, but rather in its ability to improve patient outcomes and reduce costs34. The future viability of the orthopaedic specialty depends on the ability of surgeons to work with other health-care stakeholders to create positive-sum competition by improving the value and quality of care provided to patients.
A number of challenges are posed today by misperceptions on the part of some orthopaedic surgeons as to how orthopaedic devices ideally should be (compared with how they actually are) developed and regulated. Moreover, a host of market forces create conflict regarding clinical decisions made on the basis of cost-cutting pressures on the one hand and scientific evidence on the other. Thus, it may be appropriate to propose a preliminary scheme that would more rigorously monitor the development and clinical usage of new orthopaedic technologies. One might conceptualize a new definition for product development in orthopaedics as follows: The conscientious delivery and use of best evidence to guide orthopaedic product development toward the introduction of devices and/or drugs that have a positive impact on patient-important outcomes. Conscientious delivery implies the need for surgical expertise and a cost-conscious approach. Best evidence implies the use of the hierarchy of evidence during implant development, and patient-important outcomes imply the use of relevant and important measures of safety and efficacy. Similar to drug development, a so-called orthopaedic pyramid, in which the development of orthopaedic implants begins with preliminary studies of safety toward the goal of a definitive trial of efficacy prior to a wide-scale launch in the orthopaedic marketplace, is proposed (Fig. 2). Thus, Phase 1 would be a laboratory evaluation of a device or technique involving biomechanical studies, basic-science investigations, and expert opinion. Phase 2 would provide evidence from early cases through case series and case control studies. Phase 3 would employ comparative assessment through cohort studies. Finally, Phase 4 would seek to perform pivotal, randomized controlled trials that would be the final arbiter of whether the new technology or methodology should be proposed for widespread usage in humans.
It should, however, be pointed out that this research and development pyramid is vulnerable in three respects. First, the pyramid can be time-sensitive. By the time a procedure or technology goes through all of the phases and reaches the pivotal, randomized, controlled trial phase, there may have been other developments in the health-care field that could render the randomized, controlled trial findings and, hence, all of the findings from the previous phases, clinically moot and/or commercially impractical. For example, the immunological finding that a novel implant material is detrimental to patient health trumps the randomized, controlled trial finding that it has superior wear properties in total joint replacement and good short-term clinical performance. Second, the pyramid can be limited in scope. It is not possible to answer all of the clinical and technical questions related to the procedure or technology being assessed even after going through all of the phases, since longitudinal patient follow-up will subsequently be required to determine further pros and cons. For instance, a novel total joint arthroplasty may require fifteen to twenty years of follow-up to address certain concerns, even after the completion of a randomized controlled trial and governmental approval. Third, the pyramid can be overly scrupulous. It may be that all of the phases for a new product or methodology have helped to show its statistical superiority for a number of outcome measures compared with an older product or prior methodology. However, whether this difference is truly meaningful and results in a clear benefit to the patient or clinician needs to be weighed carefully, given the time, energy, money, and resources expended on its development.
In conclusion, this symposium attempted to examine some of the more important aspects of the development and clinical introduction of new orthopaedic technologies. It is apparent that North American surgeons recognize the gap that exists between the ideal situation of evidence-based decision-making and the current lack of information that is being disseminated in their clinical community for such decisions. Market forces, in addition, will continue to apply pressure to the health-care sector to only introduce new orthopaedic devices that have been proven to be safe and effective. In this regard, it is proposed that innovative techniques and technologies will need to pass the rigors of biomechanical and other experimental evaluation, but also to be finally subjected to randomized controlled trials before widespread implementation clinically.
The audience responses to questions posed at the 2008 combined meeting of the American Orthopaedic Association and the Canadian Orthopaedic Association are available with the electronic version of this article on our web site at jbjs.org (go to the article citation and click on "Supporting Data").
Watts
G. Let's pension off the "major breakthrough". BMJ.
2007;334
Suppl 1:s4.[PubMed][CrossRef]
Sackett
DL;
Rosenberg
WM;
Gray
JA;
Haynes
RB;
Richardson
WS. Evidence based medicine: what it is and what it isn't. BMJ.
1996;312:71-2.[PubMed]
Serpell
MG;
Haldane
GJ;
Jamieson
DR;
Carson
D. Prevention of headache after lumbar puncture: questionnaire survey of neurologists and neurosurgeons in United Kingdom. BMJ.
1998;316:1709-10.[PubMed]
Schemitsch
EH;
Bhandari
M;
Boden
S;
Bourne
B;
Bozic
KJ;
Jacobs
J. Evidence-based approach to bringing new orthopaedic devices to market. Read at the 2nd Joint Meeting of the American Orthopaedic Association and Canadian Orthopaedic Association; 2008; Quebec City, QC, Canada. .
ASTM standards: medical and surgical materials and devices, anesthetic and respiratory equipment, manufacture of pharmaceutical products. Vol 13.01. West Conshohocken, PA: American Society for Testing and Materials International; 2008.
U.S. Food and Drug Administration. Food and Drug Administration Modernization Act (FDAMA) of 1997. .
U.S. Food and Drug Administration. Guidances. .
Buch
B, Deputy Director, Division of General, Restorative and Neurological Devices, Center for Devices and Radiological Health, Food and Drug Administration. .
Weiner
BK;
Walker
M. Efficacy of autologous growth factors in lumbar intertransverse fusions. Spine (Phila Pa 1976).
2003;28:1968-71.[PubMed][CrossRef]
Anekstein
Y;
Glassman
S;
Puno
R;
Carreon
L. Platelet gel (AGF) fails to increase fusion rate [abstract]. Spine J.
2003;3:89S.[CrossRef]
Castro
FP
Jr. Role of activated growth factors in lumbar spinal fusions. J Spinal Disord Tech.
2004;17:380-4.[PubMed][CrossRef]
Bourne
RB;
Rorabeck
CH;
Laupacis
A;
Feeny
D;
Wong
C;
Tugwell
P;
Leslie
K;
Bullas
R. A randomized clinical trial comparing cemented to cementless total hip replacement in 250 osteoarthritic patients: the impact on health related quality of life and cost effectiveness. Iowa Orthop J.
1994;14:108-14.[PubMed]
Riegels-Nielsen
P;
Sørensen
L;
Andersen
HM;
Lindequist
S. Boneloc cemented total hip prostheses. Loosening in 28/43 cases after 3-38 months. Acta Orthop Scand.
1995;66:215-7.[PubMed][CrossRef]
Bankston
AB;
Faris
PM;
Keating
EM;
Ritter
MA. Polyethylene wear in total hip arthroplasty in patient-matched groups. A comparison of stainless steel, cobalt chrome, and titanium-bearing surfaces. J Arthroplasty.
1993;8:315-22.[PubMed][CrossRef]
Berry
DJ;
Barnes
CL;
Scott
RD;
Cabanela
ME;
Poss
R. Catastrophic failure of the polyethylene liner of uncemented acetabular components. J Bone Joint Surg Br.
1994;76:575-8.[PubMed]
Norton
MR;
Yarlagadda
R;
Anderson
GH. Catastrophic failure of the Elite Plus total hip replacement, with a Hylamer acetabulum and Zirconia ceramic femoral head. J Bone Joint Surg Br.
2002;84:631-5.[PubMed][CrossRef]
Willert
HG;
Bertram
H;
Buchhorn
GH. Osteolysis in alloarthroplasty of the hip. The role of ultra-high molecular weight polyethylene wear particles. Clin Orthop Relat Res.
1990;258:95-107.[PubMed]
Malchau
H. On the importance of stepwise introduction of new hip implant technology [PhD dissertation]. Göteborg, Sweden: Göteborg University; 1995. p 55.
Chalmers
TC;
Celano
P;
Sacks
HS;
Smith
H
Jr. Bias in treatment assignment in controlled clinical trials. N Engl J Med.
1983;309:1358-61.[PubMed][CrossRef]
Laupacis
A;
Bourne
R;
Rorabeck
C;
Feeny
D;
Wong
C;
Tugwell
P;
Leslie
K;
Bullas
R. Costs of elective total hip arthroplasty during the first year. Cemented versus noncemented. J Arthroplasty.
1994;9:481-7.[PubMed][CrossRef]
Rorabeck
CH;
Bourne
RB;
Laupacis
A;
Feeny
D;
Wong
C;
Tugwell
P;
Leslie
K;
Bullas
R. A double-blind study of 250 cases comparing cemented with cementless total hip arthroplasty. Cost-effectiveness and its impact on health-related quality of life. Clin Orthop Relat Res.
1994;298:156-64.[PubMed]
WOMAC 3.1 Index: Knee and Hip Osteoarthritis Index. .
Bourne
RB. Measuring tools for functional outcomes in total knee arthroplasty. Clin Orthop Relat Res.
2008;466:2634-8.[PubMed][CrossRef]
Weinstein
JN;
Bronner
KK;
Morgan
TS;
Wennberg
JE. Trends and geographic variations in major surgery for degenerative diseases of the hip, knee, and spine. Health Aff (Millwood). 2004;.
Wennberg
JE;
Fisher
ES;
Skinner
JS. Geography and the debate over Medicare reform. Health Aff (Millwood). 2002;.
Associated Press. Medical device makers settle fraud case. 2007.
Associated Press. Orthopedics makers subpoenaed by U.S. 2006.
Congressional Budget Office, The Congress of the United States. Technological change and the growth of health care spending. .
Christensen
CM;
Bohmer
R;
Kenagy
J. Will disruptive innovations cure health care?Harv Bus Rev.
2000;78:102-12, 199.[PubMed]
Lingard
EA;
Katz
JN;
Wright
EA;
Sledge
CB; Kinemax Outcomes Group. Predicting the outcome of total knee arthroplasty. J Bone Joint Surg Am.
2004;86:2179-86.[PubMed]
Kim
YH;
Sohn
KS;
Kim
JS. Range of motion of standard and high-flexion posterior stabilized total knee prostheses. A prospective, randomized study. J Bone Joint Surg Am.
2005;87:1470-5.[PubMed][CrossRef]
Mendenhall
S. Implant costs not slowing, hospitals say. OR Manager.
2006;22:28.
Emanuel
EJ;
Fuchs
VR;
Garber
AM. Essential elements of a technology and outcomes assessment initiative. JAMA.
2007;298:1323-5.[PubMed] [CrossRef]