Robotic TKA showed no clinical advantage at one year: the RACER-Knee trial

Polish version: Robotyczna TKA bez przewagi po roku — wyniki badania RACER-Knee

Greater technical precision does not necessarily produce better clinical outcomes. RACER-Knee, a pragmatic multicentre randomized trial published in The Lancet on 20 August 2026, compared Mako-assisted total knee arthroplasty (TKA) with conventional surgery. Robotic assistance did not produce a clinically important improvement in the primary endpoint at 12 months.

Why this study matters

Component positioning and soft-tissue balance are important technical measures, but patients primarily experience pain, function, safety and durability. RACER-Knee tested whether robotic assistance translated into a perceptible clinical benefit. The trial involved 33 surgeons at 10 UK hospitals. Patients and assessors were masked, with measures including additional dressings and sham incisions to conceal allocation.

Study design

Patients with advanced knee osteoarthritis undergoing primary TKA received conventional instrumentation or Mako-assisted surgery. Exclusions included inflammatory arthropathy, previous fractures in the operated limb and cases requiring more complex implants.

Of 807 people assessed, 339 were randomized 1:1: 168 to robotic and 171 to conventional TKA. The primary outcome was Forgotten Joint Score (FJS) at 12 months, analysed by intention to treat. The prespecified minimum clinically important difference was 12 points.

Main results

At 12 months, mean FJS was 49.2 in the robotic group (SD 28.4; n = 154) and 50.2 in the conventional group (SD 29.9; n = 158). The adjusted mean difference was −1.5 points (95% CI −7.5 to 4.5; P = 0.62).

The confidence interval remained well short of the prespecified 12-point clinically important difference. Thus, the finding was more informative than a simple nonsignificant P value: it provided no signal of a noticeable patient benefit at 1 year.

Sixteen patients in each group experienced at least one serious adverse event. Robotic surgery was more expensive, although the public abstract did not report the exact cost difference.

Clinical interpretation for an orthopaedic surgeon

In this population, with this system used pragmatically, greater technical precision did not translate into better patient-reported outcomes at 12 months. Evaluation of robotics should therefore consider clinical outcomes, safety, costs and service delivery alongside component alignment and balancing accuracy.

Limitations

• The trial evaluated Mako, so findings cannot automatically be generalised to all robotic systems or future technology.
• The primary assessment was at 12 months; longer follow-up is ongoing.
• More complex cases, previous fractures and inflammatory arthropathies were excluded.
• The trial was not designed to establish long-term implant survival or revision risk.
• It does not determine whether selected subgroups benefit more from robotics.

Implications for practice

RACER-Knee weakens the claim that using a robot alone ensures a better TKA outcome. It does not show that robotics has no value, exclude benefits in selected indications, establish identical long-term results or resolve differences between all alignment strategies. At 1 year, Mako-assisted TKA did not offer a clinically important advantage over well-performed conventional surgery.

References

Parsons H, Metcalfe A, Griffin J, Khatri C, Bradley H, Blyth M, et al.; RACER-Knee trial team. Robotic-arm-assisted versus conventional total knee replacement (RACER-Knee): a pragmatic, multicentre, participant-masked and assessor-masked, superiority, randomised controlled trial. The Lancet. Published online 20 Aug 2026; issue publication Sep 2026. DOI: 10.1016/S0140-6736(26)00986-4. PubMed PMID 42624811.

The publication is open access under CC BY 4.0.

Related resources

For Doctors and Physiotherapists · Research Updates, Literature and Evidence-Based Medicine

Content prepared by: Tomasz Poboży - orthopaedic surgeon.