Robotic-arm–assisted total knee replacement may not demonstrate a clinically meaningful improvement in the Forgotten Joint Score at 12 months compared with conventional total knee replacement among patients with advanced knee osteoarthritis.
In the pragmatic, multicenter, participant- and assessor-masked RACER-Knee trial, researchers randomly assigned patients with knee osteoarthritis undergoing total knee replacement at 10 hospitals in England, Scotland, and Wales to robotic-arm–assisted replacement using the Mako system (n = 168) or conventional replacement (n = 171). Surgeons could use their usual surgical techniques and alignment strategies, reflecting routine practice.
Additional draping, sham incisions in the conventional group, and masked operative notes aided in masking. Both treatment groups underwent preoperative computed tomography and surgical planning to preserve masking and minimize differences attributable to preoperative planning.
The primary outcome was the 12-month Forgotten Joint Score, ranging from 0 (constant awareness of the replaced knee) to 100 (no awareness). The researchers prespecified a 12-point target between-group difference and analyzed the primary outcome according to the intention-to-treat principle.
Among 807 patients screened from December 2021 to December 2023, 339 underwent randomization. Primary outcome data were available for 312 patients at 12 months. Mean Forgotten Joint Scores were about 49 in the robotic group and 50 in the conventional group, with an adjusted difference of 1.5 points favoring conventional surgery. The findings excluded the prespecified 12-point target difference. Planned per-protocol and as-treated analyses yielded similar findings.
Measures assessed during hospitalization—including postoperative pain, opioid consumption, and discharge timing—were similar between the treatment groups. Out-of-hospital measures, including knee function and EuroQol 5-Dimension 5-Level utility scores showed no clear between-group differences. The researchers noted that these secondary findings warrant caution because the analyses did not account for multiple comparisons. Serious adverse events occurred in 16 patients in each group.
In exploratory radiological analyses, robotic assistance produced greater precision relative to planned hip-knee-ankle alignment but increased mean operating time by 10.5 minutes.
A prespecified economic analysis found higher costs with robotic-assisted replacement and a 4% probability of cost-effectiveness at the willingness-to-pay threshold. The researchers noted that results from the Mako-based trial should not be assumed to apply to other robotic technologies, which require separate evaluation. Follow-up beyond 12 months is ongoing.
The study was funded by the National Institute for Health and Care Research Health Technology Assessment Programme. Full disclosures can be found in the published study.
Source: The Lancet