Knee Replacement by Robot Worked No Better Than the Standard Operation

CATEGORY: MOBILITY, BONES & JOINTS

Does a robot give you a better knee? Hospitals increasingly advertise robot-assisted knee replacement, the operation that swaps a worn, arthritic knee for an artificial one. A trial at ten NHS hospitals has now put the question to a fair test, and so far the answer is no. One year after surgery, 339 adults with advanced knee osteoarthritis, the wear-and-tear arthritis in which the joint’s cushioning cartilage breaks down, scored essentially the same whether a robotic arm had helped or not: 49.2 versus 50.2 on a 100-point measure of how natural the knee feels. That’s a 1.5-point gap in a trial built to detect a 12-point one. The robotic operations also ran about 10 minutes longer and cost roughly £950 more each.¹

Sham Incisions Kept Everyone Guessing

A computer made the assignment: 168 people to robotic surgery with the Mako system, 171 to conventional instruments. Nobody got a different implant. Both groups received the same cemented artificial knee and the same rehabilitation plan, from 33 surgeons who all had robotic experience. Patients weren’t told which operation they’d had. The surgical teams even made small sham cuts where the robot’s guide pins would have gone, so a glance at the scars couldn’t give the secret away. The researchers scoring the results were kept in the dark too. Everyone reported on pain, knee function and daily life for a year, and 93% completed the final questionnaires.¹

Straighter on the X-Ray, the Same in Daily Life

The robot did what it’s sold to do. Implants in the robotic group sat closer to their planned position, off by 2.0 degrees on average against 2.8 degrees with standard instruments. That precision just never reached the patients. Knee scores, pain in the first days after surgery, painkiller use and overall health moved together in both groups at every checkpoint. Serious complications were identical: 16 people in each arm. Think of a dartboard where every spot in the inner ring scores the same. The robot lands the dart a little nearer the bullseye, and the scoreboard doesn’t move.¹

Why a Negative Answer Still Matters

Robots now assist about 6% of knee replacements in the UK, 16% in the US and 42% in Australia, and the Mako is the most widely used system in the world.¹ This trial was designed as a superiority test, built to show the robot was better, and it concluded instead that robotic assistance “did not provide a clinically meaningful patient benefit at 12 months.” Stryker, which makes both the robot and the implants used, covered some treatment costs but had no role in the analysis; the UK’s National Institute for Health and Care Research funded the trial. Earlier evidence points the same way. A 2024 review pooling 17 comparisons and 1,148 patients found satisfaction of 95% after robotic surgery and 91% after conventional, a difference that wasn’t statistically meaningful.² The surgeons involved don’t read any of this as the end of surgical robots. The precision might still pay off once researchers learn what to do with it, such as tailoring an implant’s position to each knee. Follow-up continues for ten years, partly to see whether either group needs fewer repeat operations down the road. Both groups, it’s worth saying plainly, ended the year far better off than before surgery.¹

Key Takeaways

• One year after knee replacement, people whose surgeon used a robotic arm felt no better than people who had the standard operation: 49.2 versus 50.2 on a 100-point score, where a 12-point gap was the smallest considered worth having.
• The robot was genuinely more precise, placing implants within 2.0 degrees of plan versus 2.8, but the extra accuracy didn’t translate into less pain or better function.
• Robotic operations took about 10 minutes longer and cost around £950 more, with identical safety: 16 serious complications in each group.
• Follow-up runs to ten years, so whether that precision eventually means longer-lasting knees is still an open question.

References

1. 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. The Lancet. 2026;408(10558):924–934. link
2. Hoveidaei AH, Esmaeili S, Ghaseminejad-Raeini A, et al. Robotic assisted Total Knee Arthroplasty (TKA) is not associated with increased patient satisfaction: a systematic review and meta-analysis. International Orthopaedics. 2024;48(7):1771–1784. link