BACKGROUND: Benchmarking seeks to evaluate implant survivorship at certain follow-up time points against predetermined standards. The International Society of Arthroplasty Registries (ISAR) International Prosthesis Benchmarking Working Group (IPBWG) developed a predefined set of standards in which receiving a benchmark indicates that an implant has acceptable revision-free survivorship, whereas not receiving a benchmark suggests that the revision rate exceeded these established standards. Although benchmarking of TKA implant systems has been performed in select international countries, applications of benchmarks to TKA systems used in a US-based cohort are lacking. QUESTIONS/PURPOSES: We sought to apply the ISAR IPBWG benchmarking standards to our US-based healthcare system's total joint replacement registry and identify TKA implant systems meeting (1) early 2- and 5-year benchmarks, (2) midpoint 10-year benchmarks, and (3) late 15-year benchmarks. As a secondary aim, we sought to evaluate the consistency of these benchmarks when stratified by patient gender and age. METHODS: Data from our US-based healthcare system's Total Joint Replacement Registry were used to conduct a retrospective cohort study. The registry is well suited for benchmarking analyses because it captures all primary and revision TKAs performed within a closed, longitudinal healthcare system, enabling near-complete outcome tracking and comprehensive implant tracking through barcode-based documentation, with routinely audited administrative and clinical coding. Patient follow-up was 96%, 91%, 85%, and 80% at 2, 5, 10, and 15 years, respectively. Surgeons complete intraoperative registry forms via the electronic health record (EHR) for all TKAs, which are crosschecked against implant information-including manufacturer, implant name, and catalog number-recorded in the EHR via barcode scan and extracted into the registry. Patients are longitudinally followed for revision surgery until healthcare plan termination or death, and all identified revisions are manually validated through chart review. We identified all primary TKAs for osteoarthritis performed from January 1, 2001, through December 31, 2024. The final sample included 288,584 primary TKAs performed by 605 surgeons at 68 facilities. The cohort had a mean ± SD age of 68 ± 9 years, mean ± SD BMI of 31.2 ± 5.5 kg/m2, was predominantly women (61% 176,641 of 288,584), and largely comprised patients with ASA class I to II (59% 171,306 of 288,584). Each unique implant system, defined as the femoral and tibial component brand, fixation for each component, insert stability, and tibial insert mobility, was evaluated separately; systems with < 250 TKAs recorded into the registry were excluded. All-cause revision incidence during follow-up was evaluated using 1 - the Kaplan-Meier estimate and reported as cumulative percent revision (CPR) and 95% confidence interval (CI). Benchmark standards proposed by the IPBWG were used to evaluate the 2-, 5-, 10-, and 15-year time points; at each time point, at least 250 TKAs had to remain at risk to receive a benchmark. Implant systems received an early benchmark at 2 and 5 years after the primary TKA when the lower bound of the CPR 95% CI was less than or equal to the revision rate benchmark standards of 2% and 3%, respectively. Superiority and noninferiority relative to the benchmark standard were determined for 10-year midpoint and 15-year late benchmarks; the noninferiority margin was set at a 20% higher relative revision rate than the benchmark standard. For the 10-year time point, the system was considered superior when the upper bound of the 95% CI was less than or equal to 5%; when the upper bound of the 95% CI was less than or equal to 6% (20% above the 5% benchmark standard, or a 1% noninferiority margin), the system was considered noninferior; when it was above 6%, the system did not receive a 10-year benchmark. For the 15-year time point, systems with the upper bound of the 95% CI less than or equal to 6.5% were considered superior, and those in which the upper bound was less than or equal to 7.8% were considered noninferior; systems in which the 95% CI was above the noninferiority margin did not receive a benchmark. RESULTS: Of the 45 TKA implant systems evaluated, 93% (42) received a benchmark at 2 years, representing 99% (284,929 of 288,584) of the TKAs included. The three implant systems not receiving a 2-year benchmark included the Smith CPR 3.1 95% CI 2.1 to 4.5); the Smith CPR 2.8 95% CI 2.1 to 3.6); and the Zimmer Biomet cemented, ultracongruent or anterior-stabilized, fixed-bearing Natural Knee®/Natural Knee II (n = 566; CPR 3.9 95% CI 2.6 to 5.9). Of the 42 implant systems evaluated for 5-year benchmarks, 88% (37) received a benchmark, representing 96% (274,056 of 286,020) of the cohort evaluated for this time point. Five implants did not receive a 5-year benchmark: the DePuy cemented, posterior-stabilized, rotating PFC Sigma®/MBT (n = 7899; CPR 3.6 95% CI 3.2 to 4.0); the Smith CPR 5.4 95% CI 4.1 to 7.1); the Smith CPR 5.1 95% CI 4.1 to 6.4); the Zimmer Biomet cemented, cruciate-retaining, fixed-bearing Natural Knee II (n = 410; CPR 4.9 95% CI 3.2 to 7.7); and the Zimmer Biomet cemented, ultracongruent or anterior-stabilized, fixed-bearing Natural Knee/Natural Knee II (n = 566; CPR 5.9 95% CI 4.2 to 8.2). Of 25 implant systems considered for the 10-year benchmark, 52% (13) of systems received a superior benchmark, and 16% (4) of systems received a noninferior benchmark. Eight implant systems did not receive any 10-year benchmark. Implant systems receiving a superior benchmark were used in 91% (204,633 of 224,619) of the eligible TKAs, whereas implants that did not receive any 10-year benchmark were used in only 3% (6385 of 224,619) of the eligible TKAs. At the 15-year time point, 50% (6 of 12) of the implant systems evaluated received a superior benchmark, representing 89% (119,871 of 135,060) of the cohort evaluated at this time point. Forty-two percent (5) of the implant systems received a noninferior benchmark, and 8% (1) of the systems did not receive a benchmark at the 15-year time point. The same implant systems consistently failed to receive benchmarks at various time points in gender- and age-stratified analyses. CONCLUSION: Applying IPBWG benchmarking to a large US-based arthroplasty registry identified TKA implant systems with consistently inferior, adequate, or superior revision performance at mid- and long-term follow-up. These findings support prioritizing implant systems that meet benchmark standards for primary TKA and reevaluating routine use of implants that fail to achieve benchmarks when alternatives with reliable mid- and long-term performance are available. Future studies integrating benchmarking with risk-adjusted analyses are needed to clarify factors driving implant performance and to strengthen postmarket surveillance. LEVEL OF EVIDENCE: Level III, therapeutic study.
Son et al. (Tue,) studied this question.