RACER-Knee Trial: Robotic Assistance Improves Precision but Not Patient Outcomes at One Year
Written By : Raju Vaishya
Published On 2026-08-27 14:30 GMT | Update On 2026-08-27 14:30 GMT
UK: Total knee replacement (TKR) remains one of the most successful operations in modern orthopaedics, yet 10–20% of patients continue to report residual pain, dissatisfaction, or awareness of their artificial joint. Against this backdrop, robotic-arm assistance has been aggressively marketed and rapidly adopted worldwide as a route to greater surgical precision and better outcomes. The RACER-Knee trial, published recently online in The Lancet, delivers the most rigorous test yet of this claim; and its findings should prompt careful reflection across the orthopaedic community.
What Did RACER-Knee Actually Test?
Led by Helen Parsons, Andrew Metcalfe and colleagues, RACER-Knee was a pragmatic, multicentre, participant- and assessor-masked superiority randomised controlled trial conducted across ten hospitals in England, Scotland and Wales. Between December 2021 and December 2023, 339 patients with advanced knee osteoarthritis were randomised to either Mako robotic-arm-assisted TKR (rTKR, n=168) or conventional instrumented TKR (cTKR, n=171). Thirty-three surgeons participated, all of whom had prior experience with the Mako system. Both groups received the same cemented Triathlon implants and standardised rehabilitation pathways. Alignment strategy was deliberately left to the operating surgeon’s preference to mirror real-world practice.
The primary outcome was the Forgotten Joint Score (FJS)—a patient-reported measure of joint awareness—at 12 months. The trial was powered to detect a clinically meaningful 12-point difference.
The Headline Results
At one year, the mean FJS was virtually identical: 49.2 in the robotic group versus 50.2 in the conventional group (adjusted mean difference –1.5, 95% CI –7.5 to 4.5; p=0.62). The confidence interval excluded the pre-specified target difference of 12 points. Secondary patient-reported outcomes—including Oxford Knee Score, EQ-5D-5L, pain scores and activity measures—showed no meaningful differences at 3, 6 or 12 months. Early postoperative pain, opioid consumption and time to discharge were also similar.
Robotic assistance did deliver greater precision. The absolute difference between achieved and planned hip–knee–ankle alignment was smaller with robotics (2.0° versus 2.8°). Operative time, however, was longer by an average of 10.5 minutes, and the robotic pathway cost approximately £950 more per case. Serious adverse events were evenly distributed (16 participants in each arm). Under standard UK willingness-to-pay thresholds, rTKR was not cost-effective within the first year.
Why This Trial Matters
RACER-Knee stands out for several reasons. It is the largest double-masked randomised comparison of a contemporary rTKR versus cTKR. Masking was unusually rigorous for a surgical trial (sham tibial pin-site incisions, identical preoperative CT scans, additional draping and masked operation notes). The pragmatic design—allowing surgeons to choose alignment strategy—means the results reflect how the technology is actually used in day-to-day practice rather than an artificial protocol-driven ideal.
Previous evidence had been dominated by observational series and smaller, often unmasked studies that suggested early advantages in pain, length of stay, or radiographic accuracy. RACER-Knee demonstrates that superior radiographic precision does not automatically translate into better patient-reported outcomes at one year.
Important Caveats
The primary analysis is limited to 12 months. Recovery after TKR generally plateaus by this time, but differences in implant survival, revision rates, or late functional decline may emerge later; the trial team plans follow-up to ten years. The pragmatic approach to alignment also means the trial tested “robotics as currently practised” rather than a single idealised alignment philosophy. If the “correct” target alignment remains incompletely defined, executing imperfect plans more precisely will not improve outcomes. Finally, the trial was conducted in the UK NHS using one implant system and one robotic platform; generalisability to other systems, cementless fixation or different healthcare environments should be considered.
Clinical and System Implications
For individual surgeons and patients, the message is clear: choosing robotic assistance in the expectation of meaningfully better pain relief, function or joint “forgetfulness” at one year is not currently supported by high-quality evidence. The technology is safe and more precise, but those technical gains have not yet translated into the outcomes that matter most to patients in the short-to-medium term.
For health systems, the cost and theatre-time implications are significant. Rapid, widespread adoption driven primarily by marketing claims or competitive pressure risks diverting resources without delivering proportional patient benefit. The value proposition of robotics may ultimately lie in reducing the minority of technical outliers that contribute to the poorly performing knee, or in enabling more sophisticated personalised alignment strategies once those targets are better defined. Neither of these potential benefits has been proven in a randomised setting to date.
Looking Ahead
RACER-Knee does not “kill” rTKR, but raises the evidentiary bar. Longer-term data on revisions and survivorship will be crucial. Future research should also examine outcome distributions (particularly the lower tail of the FJS), identify any subgroups who may benefit disproportionately, and test whether robotics can deliver superior results once alignment philosophy itself is optimised.
Until then, the trial offers a timely reminder that surgical innovation must be judged by its effect on patients, not merely by its technical elegance. Precision is valuable only when it improves lives.
References
Parsons H, Metcalfe A, Griffin J, et al. Robotic-arm-assisted versus conventional total knee replacement (RACER-Knee): a pragmatic, multicentre, participant-masked and assessor-masked, superiority, randomised controlled trial. Lancet. Published online Aug 20, 2026. doi:10.1016/S0140-6736(26)00986-4.
Our comments section is governed by our Comments Policy . By posting comments at Medical Dialogues you automatically agree with our Comments Policy , Terms And Conditions and Privacy Policy .
Disclaimer: This website is primarily for healthcare professionals. The content here does not replace medical advice and should not be used as medical, diagnostic, endorsement, treatment, or prescription advice. Medical science evolves rapidly, and we strive to keep our information current. If you find any discrepancies, please contact us at corrections@medicaldialogues.in. Read our Correction Policy here. Nothing here should be used as a substitute for medical advice, diagnosis, or treatment. We do not endorse any healthcare advice that contradicts a physician's guidance. Use of this site is subject to our Terms of Use, Privacy Policy, and Advertisement Policy. For more details, read our Full Disclaimer here.
NOTE: Join us in combating medical misinformation. If you encounter a questionable health, medical, or medical education claim, email us at factcheck@medicaldialogues.in for evaluation.