BACKGROUND Previously, we conducted a retrospective study of American Joint Replacement Registry (AJRR) data that examined the 2-year odds of revision between robotic-assisted and nonrobotic-assisted TKA, and we found no benefit to robotic assistance. However, proponents of robotic assistance have suggested that robot platforms confer more accurate bone cuts and precise implant sizing that might promote osteointegration of cementless implants by limiting micromotion at the bone-implant interface that could lead to aseptic loosening. Therefore, it seems important specifically to evaluate the odds of revision among patients with cementless implants only within our previous study population. QUESTIONS/PURPOSES (1) After controlling for potentially confounding variables, such as surgeon, institution, and patient comorbidity profile, was robotic assistance associated with a decreased odds of 2-year revision of cementless TKA for any reason compared with cementless TKAs performed without robotic assistance? (2) After again controlling for potentially confounding variables, was robotic assistance associated with a decreased odds of 2-year revision of cementless TKA for particular revision indications (such as aseptic loosening, infection, instability, or pain) compared with the cementless TKAs performed without robotic assistance? METHODS Using the AJRR, a retrospective cohort of patients ≥ 65 years of age with osteoarthritis who underwent primary TKA with cementless femur and tibial components from January 2017 through March 2020 was identified. Procedures performed with hybrid fixation (cement only on the tibia but not on the femur, or vice versa) were excluded. The AJRR was selected because it is the largest arthroplasty registry in the world by annual procedure volume, and it contains a data linkage with inpatient and outpatient Medicare claims data to ensure near-complete 2-year follow-up. A total of 9220 patients were identified, and robotic assistance was used in 45% (4130) of procedures. Patient age did not differ between groups (72 ± 5 years versus 72 ± 5 years; p = 0.29). However, the robotic-assisted cohort had a slightly higher proportion of female patients (56% [2332 of 4130] versus 53% [2693 of 5090]; p = 0.002) and higher Charlson comorbidity index (CCI) (2.9 ± 0.9 versus 2.8 ± 0.9; p = 0.003). Therefore, a mixed-effects model was used to analyze the ORs for all-cause linked revision with robotic assistance and was adjusted for age, gender, CCI, surgeon, and institution. Subanalyses were performed on indications for revision. A power analysis demonstrated the ability to measure a difference as small as one-half SD between risk of revision within each cohort (specifically, moderate effect sizes based on Cohen d). RESULTS After controlling for potentially confounding variables, such as surgeon, location of surgery, and patient comorbidity profile, we found no difference regarding odds of all-cause revision between robotic-assisted and nonrobotic-assisted cementless TKA (OR of robotic-assisted versus nonrobotic-assisted cementless TKA 0.8 [95% CI 0.5 to 1.3]; p = 0.41). There were no differences in reasons for revision between robotic-assisted and nonrobotic-assisted cementless TKA, such as mechanical loosening (OR 3.2 [95% CI 0.8 to 12]; p = 0.09) and infection (OR 1.5 [95% CI 0.8 to 2.6]; p = 0.19). CONCLUSION We found no evidence that robotic assistance improves the odds of cementless implant survival free from revision at 2 years. Importantly, the differences in odds of revision beyond 2 years as a function of robotic assistance in cementless TKA remains unknown and warrants further population-level investigation, but until or unless robotic assistance proves its value in well-designed studies in terms of endpoints that patients can perceive (such as pain, function, or survivorship), we recommend against widespread adoption until further evidence becomes available. Additionally, we encourage surgeons and healthcare facilities using robotic assistance in cementless TKAs to study its impact on patient outcomes, if any. LEVEL OF EVIDENCE Level III, therapeutic study.

中文翻译：

---

机器人辅助与非骨水泥 TKA 的早期翻修减少无关：美国关节置换登记处分析。


背景 此前，我们对美国关节置换登记处 （AJRR） 数据进行了回顾性研究，该研究检查了机器人辅助和非机器人辅助 TKA 之间的 2 年翻修几率，我们发现机器人辅助没有益处。然而，机器人辅助的支持者建议机器人平台提供更准确的骨切割和精确的植入物尺寸，这可能会通过限制可能导致无菌松动的骨-植入物界面的微运动来促进无骨水泥植入物的接骨。因此，仅在我们之前的研究人群中评估非骨水泥植入物患者的翻修几率似乎特别重要。问题/目的 （1） 在控制了潜在的混杂变量（例如外科医生、机构和患者合并症概况）后，与没有机器人协助的非骨水泥 TKA 相比，机器人辅助是否与无骨水泥 TKA 的 2 年翻修几率降低有关？（2） 在再次控制潜在的混杂变量后，与没有机器人协助的无骨水泥 TKA 相比，机器人辅助是否与针对特定翻修适应症（如无菌松动、感染、不稳定或疼痛）的无骨水泥 TKA 2 年翻修的几率降低有关？方法 使用 AJRR，确定了 2017 年 1 月至 2020 年 3 月接受原发性 TKA 的 65 ≥骨关节炎骨关节炎患者的回顾性队列。混合固定（仅在胫骨上粘接，而不在股骨上，反之亦然）进行的手术被排除在外。 选择 AJRR 是因为按年手术量计算，它是世界上最大的关节置换术登记处，并且它包含与住院和门诊 Medicare 索赔数据的数据链接，以确保近乎完整的 2 年随访。共确定了 9220 名患者，45% （4130） 的手术中使用了机器人辅助。患者年龄组间无差异 （72 ± 5 岁 vs 72 ± 5 岁;p = 0.29）。然而，机器人辅助队列的女性患者比例略高 （56% [4130 人中的 2332 人] 对 53% [5090 人中的 2693 人];p = 0.002）和较高的查尔森合并症指数 （CCI） （2.9 ± 0.9 对 2.8 ± 0.9;p = 0.003）。因此，使用混合效应模型在机器人辅助下分析全因相关翻修的 OR，并针对年龄、性别、CCI、外科医生和机构进行调整。对翻修适应症进行亚分析。功效分析表明，能够测量每个队列内翻修风险之间小至 SD 的二分之一 SD 差异（具体而言，基于 Cohen d 的中等效应量）。结果 在控制了潜在的混杂变量，例如外科医生、手术位置和患者合并症概况后，我们发现机器人辅助和非机器人辅助非骨水泥 TKA 之间的全因翻修几率没有差异 （机器人辅助与非机器人辅助非骨水泥 TKA 的 OR 0.8 [95% CI 0.5， 1.3];p = 0.41）。机器人辅助和非机器人辅助非骨水泥 TKA 之间的翻修原因没有差异，例如机械松动 （OR 3.2 [95% CI 0.8， 12];p = 0.09） 和感染 （OR 1.5 [95% CI 0.8， 2.6];p = 0.19）。 结论 我们没有发现任何证据表明机器人辅助可以提高 2 年无骨水泥植入物翻修的存活率。重要的是，在非骨水泥 TKA 中，机器人辅助超过 2 年的翻修几率差异仍然未知，需要进一步的人群水平调查，但除非或除非机器人辅助在设计良好的研究中证明其在患者可以感知的终点（如疼痛、功能或生存率）方面的价值，我们建议在获得进一步证据之前不要广泛采用。此外，我们鼓励外科医生和医疗机构在非骨水泥 TKA 中使用机器人辅助来研究其对患者预后的影响（如果有）。证据级别 III 级，治疗研究。