The Lancet August 20, 2026

Robotic-arm-assisted versus conventional total knee replacement (RACER-Knee): a pragmatic, multicentre, participant-masked and assessor-masked, superiority, randomised controlled trial

Helen Parsons, Andrew Metcalfe, James Griffin, Chetan Khatri, Helen Bradley, et al.

Bottom Line

In patients with advanced knee osteoarthritis, robotic-arm-assisted total knee replacement did not improve joint awareness or functional outcomes at 12 months compared to conventional surgery, while adding operative time and procedural costs.

Key Findings

1. At 12 months, the mean Forgotten Joint Score (FJS) was 49.2 (SD 28.4) in the robotic-arm-assisted group and 50.2 (SD 29.9) in the conventional group.
2. The adjusted mean difference in FJS was -1.5 (95% CI -7.5 to 4.5; p=0.62), demonstrating no clinically meaningful benefit and failing to meet the prespecified target difference of 12 points.
3. Safety profiles were identical between the two arms, with exactly 16 participants in each group experiencing one serious adverse event.
4. Robotic-arm-assisted procedures took an average of 10.5 minutes longer to perform than conventional surgeries.
5. The robotic approach cost approximately £950 more per operation and was not considered cost-effective within the first year of follow-up.

Study Design

Design
RCT
Double-Blind
Sample
339
Patients
Duration
12 mo
Median
Setting
Multicenter, UK
Population Adult patients with advanced knee osteoarthritis requiring primary total knee replacement.
Intervention Robotic-arm-assisted total knee replacement (rTKR) using the Stryker Mako system.
Comparator Conventional total knee replacement (cTKR) utilizing standard manual instrumentation, with sham incisions used to maintain participant masking.
Outcome Forgotten Joint Score (FJS) at 12 months post-randomization.

Study Limitations

Follow-up is currently limited to 12 months; longer-term follow-up (planned up to 10 years) is necessary to determine if robotic precision reduces late implant wear, loosening, or revision rates.
The pragmatic design allowed surgeons to utilize their preferred alignment strategies, which mimics real-world practice but might dilute potential benefits if a specific robotic alignment philosophy is inherently superior.
The study specifically evaluated the Mako robotic system, and results may not completely generalize to other robotic or navigation platforms.

Clinical Significance

The RACER-Knee trial provides rigorous Level I evidence challenging the widespread commercial claims that robotic-assisted knee arthroplasty yields superior early functional recovery. While robotic systems reliably improve radiographic precision, this technological advantage does not translate into tangible, patient-reported benefits regarding joint awareness or pain at 1 year. Consequently, healthcare systems and orthopedic centers must critically weigh the lack of early clinical superiority against the substantial capital investments, per-case costs, and increased operative times associated with routine robotic use.

Historical Context

Over the past decade, robotic-arm-assisted joint replacement has experienced explosive growth and aggressive marketing, achieving massive global market penetration (utilized in over 40% of cases in Australia and nearly 20% in the US by 2026). Proponents argued that millimeter-level precision in implant positioning and dynamic soft-tissue balancing would fundamentally improve clinical outcomes, addressing the historical 15-20% patient dissatisfaction rate in conventional knee arthroplasty. Prior to RACER-Knee, the literature was largely composed of observational cohorts, retrospective reviews, or small, unblinded trials subject to heavy placebo effects. As the largest participant- and assessor-masked randomized controlled trial in this space—employing sham marker incisions to maintain blinding—RACER-Knee serves as a landmark reality check, highlighting the dissociation between technical precision and actual patient-perceived benefits.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the primary pathophysiological indication for total knee replacement, and what is the theoretical mechanism by which robotic-arm assistance was hypothesized to improve postoperative joint awareness?

Key Response

Total knee replacement is indicated for end-stage knee osteoarthritis refractory to conservative management, characterized by cartilage loss, osteophyte formation, and joint space narrowing. Robotic assistance was theoretically hypothesized to improve outcomes by providing sub-millimeter precision in bone cuts and improved soft tissue balancing, which surgeons believed would result in a more natural feeling knee and decreased joint awareness, though the RACER-Knee trial proved this does not translate to clinical superiority at 12 months.

Resident
Resident

How should you counsel a patient with severe knee osteoarthritis who specifically requests robotic-assisted total knee replacement because they saw an advertisement claiming it provides a faster, less painful recovery?

Key Response

Residents must practice evidence-based patient counseling. Based on the Level I evidence from the RACER-Knee trial, the patient should be informed that while robotic-assisted surgery is safe, it does not provide superior functional outcomes, better pain relief, or improved joint awareness at 12 months compared to conventional methods, and it is associated with longer operative times and higher procedural costs.

Fellow
Fellow

While the RACER-Knee trial demonstrated no difference in functional outcomes at 12 months, how might the improved radiographic accuracy and implant alignment typically achieved with robotic assistance impact the 10- to 15-year survivorship of the prosthesis?

Key Response

Fellows must distinguish between short-term functional outcomes and long-term mechanical survival. While 12-month patient-reported outcomes show no difference, improved coronal and sagittal alignment might reduce eccentric wear and aseptic loosening over decades. Therefore, the true value of robotic assistance may only become apparent in long-term registry data regarding revision rates, rather than early functional recovery.

Attending
Attending

Given that the superiority design of the RACER-Knee trial failed to show clinical benefit at 12 months while demonstrating increased operative time and procedural costs, how do you justify the continued capital investment and institutional utilization of robotic systems in a value-based healthcare model?

Key Response

Attendings must balance clinical evidence with hospital economics and market dynamics. This question highlights the tension between empirical evidence showing a lack of short-term clinical superiority and external pressures like market competition and patient demand. It emphasizes the need for critical value analysis before institutional adoption of expensive new technologies that add procedural time without improving early patient outcomes.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The RACER-Knee trial utilized participant and assessor masking but could not mask the operating surgeons. How does this partial masking, combined with a pragmatic, multicentre superiority design, influence the risk of performance bias and the interpretation of patient-reported primary outcomes?

Key Response

Surgical trials face inherent challenges with masking. Unmasked surgeons might alter their postoperative care, rehabilitation protocols, or surgical speed based on their enthusiasm for the technology. While the use of a patient-reported outcome with participant masking helps mitigate placebo effects, surgeon-induced subtle biases in perioperative protocols could still theoretically occur, making rigorous assessor masking and objective secondary endpoints crucial for trial validity.

Journal Editor
Journal Editor

As a peer reviewer, how would you critically evaluate the decision to use a 12-month primary endpoint for assessing the superiority of robotic-assisted total knee replacement, and what potential threats does this timeline pose to the definitive conclusion that the technology lacks clinical utility?

Key Response

A rigorous reviewer would argue that while 12 months is sufficient for assessing acute recovery and early functional plateau, it is completely inadequate for detecting differences in implant longevity or late-stage revisions, which are the primary theoretical benefits of micromillimeter precision. Publishing this as a definitive negative trial requires careful editorial framing to avoid over-extrapolating short-term equivalence to long-term parity.

Guideline Committee
Guideline Committee

How should the results of the RACER-Knee trial influence upcoming orthopedic clinical practice guidelines (such as those by the AAOS or NICE) regarding recommendations for the routine use of robotic assistance in total knee arthroplasty?

Key Response

Guideline committees rely on high-quality RCTs to establish standards of care. The RACER-Knee trial provides Level I evidence that robotic assistance should not be recommended as a routine standard of care over conventional TKA for the purpose of short-term functional gains, due to increased cost and operative time without clinical benefit. Guidelines should reflect a strong recommendation that conventional TKA remains the gold standard, while noting that long-term survivorship data for robotics is still pending.

Clinical Landscape

Noteworthy Related Trials

2013

Song et al. RCT

n = 100 · JBJS

Tested

Active robotic-assisted TKA

Population

Patients with severe knee osteoarthritis

Comparator

Conventional manual TKA

Endpoint

Mechanical axis alignment and implant positioning

Key result: Robotic-assisted TKA significantly reduced the number of mechanical axis outliers compared to conventional surgery.
2018

Kayani et al. Prospective Study

n = 80 · Bone Joint J

Tested

Robotic-arm assisted TKA (Mako)

Population

Patients undergoing primary TKA for osteoarthritis

Comparator

Conventional jig-based TKA

Endpoint

Early functional recovery and hospital discharge time

Key result: Robotic-arm assistance was associated with reduced postoperative pain, lower opiate use, and faster time to hospital discharge.
2021

Batailler et al. RCT

n = 100 · KSSTA

Tested

Robotic-arm assisted TKA

Population

Patients with end-stage knee osteoarthritis

Comparator

Conventional TKA

Endpoint

Clinical outcomes (Knee Society Score) and radiological accuracy

Key result: While robotic assistance improved implant positioning accuracy, there was no significant difference in patient-reported clinical scores at medium-term follow-up.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis