Editor’s Spotlight/Take 5: Does Knee Prosthesis Survivorship Improve When Implant Designs Change? Findings from the Australian Orthopaedic Association National Joint Replacement Registry

Evidence-based medicine skeptics would say that every once in a while, something important gets published in an orthopaedic journal.

I disagree. It’s a regular occurrence; at least it should be in the journals that deserve your attention. Limiting myself only to those discoveries published in this journal, only to those covered on the Editor’s Spotlight page, only to arthroplasty papers (this being the month that Clinical Orthopaedics and Related Research® is publishing the selected proceedings of the International Society of Arthroplasty Registries), and only to the last year, here are a few samples:

• Cemented fixation in THA is increasing globally, even in older patients, despite clear evidence that this surgical choice harms patients [3].

• Even after adjusting for relevant confounding variables, black patients with hip fractures wait longer not only for radiographs but also for potentially life-saving surgery than do white patients [1].

• In terms of predicting which patients would experience a clinically important improvement after TKA, a coin toss would do about as well as expert surgeons at a high-volume, internationally known arthroplasty center [6].

In fairness, this last one was published a couple of years ago. I only include it on this list because a letter to the editor published in this issue described a finding in that paper—and not even the finding I mentioned above—as “probably the biggest news Clinical Orthopaedics and Related Research® has ever published” [2]. The letter-writer considered that paper the orthopaedic equivalent of Watson and Crick’s DNA paper, and pointed to another paper he believed to be similarly important [5] that was covered with an Editor’s Spotlight/Take 5 [9] within the past year. So I’ll call both of those articles and the letter tying them together [2, 5, 6] fair game for a year-in-review list like this.

Regardless, all of those papers [1, 3, 5, 6], as well as the interviews that accompanied them [9-12] deserve your attention, as does this month’s Editor’s Spotlight article, “Does Knee Prosthesis Survivorship Improve When Implant Designs Change? Findings from the Australian Orthopaedic Association National Joint Replacement Registry” [13].

This remarkable work found that of the 11 TKA implants in the Australian registry that underwent design modifications over a nearly 20-year span ending in 2017, five delivered implant survivorship that was either no different from or inferior to the design they replaced [13]. Another coin toss. Decide for yourself if you think the new implants were less expensive or more so.

This is a devastating indictment of our specialty and how we integrate new technology into practice. I just don’t know another way to say it. Why toss a coin on a new arthroplasty implant when alternatives with proven track records—both in terms of safety [17] and cost [14]—are available?

Yes, this is the issue of CORR® that contains the selected proceedings of the International Society of Arthroplasty Registries, and yes, this is an arthroplasty paper. But this problem is not limited to lower-extremity arthroplasty. Premature adoption of new technologies, implants, drugs, and approaches has harmed patients in every subspecialty of orthopaedic surgery; local anesthetic catheters have caused chondrolysis after arthroscopy [16], bone-morphogenic protein has resulted in all manner of complications after spinal fusions [4], we’ve seen too many bad ideas in hallux valgus surgery even to name, and I cannot explain our continued warm embrace of interventions both old and new that almost completely lack supportive evidence [7, 8]. For these reasons, this article [13] and especially the behind-the-discovery Take 5 interview that follows with Peter L. Lewis MBBS, FRACS(Orth), FAOrthA (Fig. 1), first author of this essential study, deserve the attention of every surgeon in our specialty.

Peter L. Lewis MBBS, FRACS(Orth), FAOrthA

Keep your eyes on CORR®. We’ll continue to bring you the discoveries you need to practice safely and effectively. And not just once in a while.

Take 5 Interview with Peter L. Lewis MBBS, FRACS(Orth), FAOrthA, first author of “Does Knee Prosthesis Survivorship Improve When Implant Designs Change? Findings from the Australian Orthopaedic Association National Joint Replacement Registry” [13]

Seth S. Leopold MD: Congratulations on this extremely well done, if sobering, registry study. Findings like yours have obvious policy implications—if surgeons don’t start making more-responsible choices, it’s easy to imagine other stakeholders stepping in to start making those choices for us (and for our patients). At the same time, I was not terribly surprised by your top-line finding; it’s practically the history of the specialty of arthroplasty. Why are we so slow to learn this lesson?

Peter L. Lewis MBBS, FRACS(Orth), FAOrthA: Members of the community (which includes surgeons and their patients) are attracted by the latest technology, techniques, and implants, as it is a widespread belief that “new” is better. Additionally, our medical educators have instructed us to continually update our knowledge to keep up with what is new or state-of-the-art, and in the area of arthroplasty, this has often carried over to prosthesis choice.

Surprisingly, longevity is only one factor surgeons use when deciding which implant to choose. We need to retrain our arthroplasty surgeons about the importance of looking beyond the time it takes for patients to recover from their operations. In an international survey of knee arthroplasty surgeons, we found that the risk of revision only ranked tenth of 17 attributes regarding prosthesis choice, and, alarmingly, 20% of surgeons responded that the revision risk did not influence their prosthesis choice [19]. The importance of revision risk needs to be emphasized. If we can do this well, there should be no need for other stakeholder involvement.

Dr. Leopold: Given what you learned about our specialty’s batting average in terms of choosing implants—to say nothing of our apparent inability to guess which patients are likely to benefit from TKA [6]—why shouldn’t policymakers step in with a heavier hand to set standards for implant selection and surgical indications? Or should they?

Dr. Lewis: While some arthroplasty innovations have resulted in harm (such as large-diameter metal-on-metal hip replacement), not all changes have been disasters. For example, a hugely successful new technology has been the wide adoption of highly cross-linked polyethylene, particularly for hip replacement. As our study [13] points out, it is difficult to predict which changes will be beneficial.

It is now 20 years since Malchau [15] proposed his stepwise introduction of new technology. While this utopia is often discussed by surgeons, unfortunately, implant manufacturers want their products to return profits rapidly and don’t have the patience for this more-considered approach (nor are they required to). As a group, surgeons should continue to strive for a more-responsible initial release of new technologies, with clinical trials of prostheses before widespread use.

Market regulation is already happening with changing (and more-stringent) requirements coming into effect in Europe. Regulators from many countries in the past have given approval for prostheses based on “substantial equivalence” rather than requiring specific evidence of clinical effectiveness before a product is sanctioned for general use. Some countries, insurers, health sectors, and hospitals already limit prosthesis choice. Regulatory control for patient safety reasons makes sense, but how far this permeates into limitations on prosthesis selection is a difficult question to answer. Unless policymaking committees are well-informed by orthopaedic surgeons, they also have the potential to make ill-considered choices, restrictions, and limitations.

Dr. Leopold: What findings surprised you most about your study?

Dr. Lewis: I think two things stand out most. First, I was surprised at the speed at which some of the new prostheses were adopted, when no or limited evidence about their use was available. Second, I was surprised by the finding that revisions for instability have not decreased in any new system, despite changing concepts of femoral curvature, introduction of technology to improve the precision of component placement (such as computer navigation or image-derived instrumentation), addition of single-millimeter increment increases in polyethylene insert thickness, and more choices of polyethylene shape (such as anterior lipped or conforming variants). As an aside, an earlier—and as-yet unpublished—study suggested that only about 25% of total knee systems in Australia are still in use 8 years after their introduction. Sometimes insurers or hospital policies can restrict choice and force change.

Dr. Leopold: Your study demonstrates the power of well-run national registries to answer the big-picture questions that we can’t answer any other way. It is for this reason I’m so proud that CORR® is the official journal of the International Society of Arthroplasty Registries. What unanswered questions of this size and scope do you see registries answering next or do you hope to tackle using a registry database?

Dr. Lewis: It is important to understand that registries are quality-assurance mechanisms with the primary purpose of improving outcomes in the communities they serve, but this role is constantly evolving. The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR), like other registries, has realized that its activities need to be expanded. We can no longer just focus on revision and death as primary outcome measures, but need to consider other outcomes such as patient-reported outcomes. By linkage to electronic health records and administrative datasets, registries can obtain more-detailed information on local and systemic complications, identify best practices in arthroplasty surgery, help to reduce the length of stay, determine the role of rehabilitation, and of course, help to minimize cost and maximize benefit.

Registries can also use their infrastructure to develop new approaches such as nested clinical trials. Registry nested trials can answer more-directed or specific questions and do this cheaper and more effectively than current approaches. As an example, the AOANJRR is currently running a cluster randomized trial to compare aspirin with low-molecular-weight heparin for the prevention of venous thromboembolism in patients undergoing hip and knee arthroplasty [18]. This study, which seeks to enroll more than 14,000 patients, would be difficult to undertake without embedding it in a registry-nested trial format, but with the registry’s support, it does not require commercial sponsorship.

Dr. Leopold: So many orthopaedic subspecialties don’t—and won’t ever—have national registries to guide them. How might those specialties, and the patients they serve, get the answers to the big questions?

Dr. Lewis: Not surprisingly, I am a believer in registries and what they can achieve. There is emerging use of registry approaches for soft-tissue orthopaedic conditions, such as ACL injuries or rotator cuff pathology, and I think we will see an expansion of registry use in many new areas of orthopaedics in the coming decade. While there are challenges in obtaining adequate population coverage and measurement of outcomes that may not be as well-defined as in arthroplasty, I feel these can be largely overcome.

While registry analyses provide one method, I don’t think “one size fits all.” In many instances, we will obtain these answers by well-constructed clinical studies, as well with approaches like meta-analyses and systematic reviews. Increasing use of electronic health records and the availability of “big data” sets, such as administrative or discharge data, will offer more opportunities to provide answers through data-linkage studies.