Robotic vs Traditional Knee and Hip Replacement: What the 2026 BMJ Registry Studies Found

As of 15 October 2026, two large UK registry studies of about 1.36 million hip and knee replacements found no clear difference in early implant survival or overall revision between robotic and conventional operations. The one exception was hip revisions for implant malposition, which were fewer after robotic surgery. The studies are observational, follow-up is short, and they say nothing about pain or function.

This guide explains what the 2026 BMJ studies compared, what they can and cannot tell you, and what to ask when one quote includes robotic assistance and another does not. It takes no side. Robotic-arm assistance is a choice for you and your surgeon, and nothing here says to choose it or to avoid it. For how the systems work, see our earlier explanation of robotic knee replacement.

Is robotic knee replacement better than traditional knee replacement?

So far, the evidence has not shown that it is. In the BMJ knee study, estimated five-year survival of a total knee replacement was 98.5% with conventional surgery and 98.6% with robotic surgery, and the revision risk did not differ (hazard ratio 1.03, 95% confidence interval 0.86 to 1.26). Cause-specific revisions and complications during surgery did not differ either.

A randomised trial points the same way. The RACER-Knee trial, published in The Lancet in August 2026, randomly assigned 339 people to a total knee replacement with conventional instruments or with the Mako robotic arm at ten hospitals in Great Britain. After 12 months the groups scored similarly on the Forgotten Joint Score (49.2 robotic, 50.2 conventional; adjusted difference -1.5 points, 95% confidence interval -7.5 to 4.5), a questionnaire about how aware people are of their new joint. The authors concluded that the robotic operation, as delivered in routine practice, was more costly and gave no clinically meaningful patient benefit at 12 months. They reported similar safety in the two groups.

That trial tested one robotic system, in total knee replacement only, and followed people for a year. Long-term follow-up is ongoing.

Is robotic hip replacement better than traditional hip replacement?

The registry data show no clear difference in overall outcome. In the BMJ hip study, estimated five-year implant survival was 98.8% in both the conventional and the robotic group, and overall revision risk did not differ (hazard ratio 0.96, 95% confidence interval 0.75 to 1.22). Patient survival and complications during surgery did not differ either.

One specific result did differ, and it is covered below. We found no published results yet from the randomised hip trial, RACER-Hip, so there is no trial evidence on how a robotic hip feels or functions.

What did the 2026 BMJ studies compare, and how many people were in them?

Two papers (hip and knee), published on 30 September 2026 by researchers at Queen Mary University of London and the University of Oxford, compared robotic and conventional operations using data from the National Joint Registry. The BMJ Group press release summarises them. The registry covers England, Wales, Northern Ireland, the Isle of Man, Guernsey and Jersey, and both public and private hospitals. The operations took place between 2018 and 2024.

Robotic operations were a small share: about 1.5% of the hips and 3.2% of the knees. The researchers used a method called target trial emulation, which designs a registry analysis to mimic a randomised trial, and matched robotic and conventional patients on recorded characteristics. The press release lists age, sex, body mass index, diagnosis, physical fitness, implant type and surgeon volume.

	Hip replacement	Total knee replacement	Partial knee replacement
Operations in the study	666,283 (656,080 conventional, 10,203 robotic)	697,145 total and partial knees together (675,034 conventional, 22,111 robotic)	Included in the 697,145
Estimated five-year implant survival, conventional vs robotic	98.8% vs 98.8%	98.5% vs 98.6%	97.8% vs 96.4%
Revision risk, robotic vs conventional	Hazard ratio 0.96 (0.75 to 1.22), no significant difference	Hazard ratio 1.03 (0.86 to 1.26), no significant difference	Hazard ratio 1.03 (0.75 to 1.42), no significant difference

A hazard ratio compares the risk of an event over time. A value of 1 means no difference, below 1 means a lower risk in the robotic group and above 1 a higher one. The range in brackets (the 95% confidence interval) shows how uncertain the estimate is.

Does a robot make a joint replacement last longer?

Nothing in these studies shows that it does. The headline figure is "five-year implant survival", but the average follow-up was only about 2.5 years. The five-year number is an estimate from survival analysis, not five years of observation of every patient, so many people contributed less than five years of data.

The press release lists limited follow-up as a limitation that may mask longer-term outcomes. The sensible reading is that robotic and conventional joint replacements looked alike early on, and that the long-term picture is not yet known.

Also, "no significant difference" is not the same as "proven equal". A wide interval, such as 0.75 to 1.22 for hips, leaves room for a modest difference in either direction.

What does the finding about fewer malposition revisions mean?

In the hip study, revisions for implant malposition (the authors list dislocation, leg length discrepancy and malalignment) were less common after robotic surgery. The hazard ratio was 0.53 (95% confidence interval 0.30 to 0.93; P=0.03). It was the only outcome that differed significantly.

It is worth reading carefully:

• It is a lower risk of one cause of revision, not fewer revisions overall. Overall revision risk was not different (hazard ratio 0.96).
• It is one of several outcomes compared, and its confidence interval runs close to 1, so it is a signal to test rather than a settled fact.
• It is an average across thousands of patients. It does not tell any one person what will happen in their hip. Revisions of any kind were uncommon in both groups, and the figures we could read do not give the number of malposition revisions in each group.
• It is observational. Surgeons and hospitals that use robots may differ in other ways from those that do not, and matching can only account for what was recorded.

Whether a robotic arm lowers the chance of a hip that dislocates or ends up with unequal leg lengths is a question for a randomised trial and longer follow-up.

Does robotic assistance matter for a partial knee replacement?

The registry data do not show that it does either way. For partial (unicompartmental) knee replacement, the revision risk did not differ significantly between robotic and conventional operations (hazard ratio 1.03, 95% confidence interval 0.75 to 1.42; P=0.86).

You may notice that the estimated five-year survival was 97.8% for conventional and 96.4% for robotic partial knees. The robotic figure comes with a very wide interval, 92.4% to 98.3%, which overlaps the conventional one, so the number is uncertain and the study does not show that one approach does worse. The revision-risk comparison is the more reliable guide here, and it showed no difference.

What do these studies not tell you?

Registry studies count implants and revisions. They cannot tell you:

• How the joint feels. Pain, walking, stairs, return to activities and quality of life were not measured. RACER-Knee measured joint awareness at 12 months for total knees only.
• Which robotic system was used. The press release and abstracts do not say which platforms were included, so the results cannot be read as being about one brand.
• How much an individual surgeon's experience matters. The press release lists surgeon volume among the matching factors, but the studies compare groups of operations, not individual surgeons, so a surgeon's own experience with a given operation is still a separate question to ask.
• Costs. The authors say robotic surgery has substantially higher capital and procedural costs and call for careful evaluation in publicly funded systems.
• What happens outside the UK. This is UK registry data. It is not data from Türkiye, and it says nothing, favourable or unfavourable, about robotic surgery in any single hospital or country.

Should I pay extra for robotic surgery, and what should I ask the clinic?

Whether to pay for robotic assistance is a decision for you and your surgeon, and the evidence above does not show it to be either necessary or unnecessary. What you can do is ask questions that make the offer clear. Whether you are comparing clinics in Istanbul, or your own country's clinics in the US, Canada, the UK or elsewhere in Europe, the same questions apply:

1. Which system will be used, and why is it suggested for me? Ask for the reason in your case, not only a general description.
2. Is robotic assistance, including any CT planning scan, a separate line in the quote? Then you can see what it adds and compare like with like.
3. How many of this operation does the surgeon perform a year, and how many with and without the robot?
4. What happens if the robot is unavailable on the day? Ask whether the surgeon routinely does the conventional operation and whether the plan changes.
5. How does the hospital track its outcomes? Ask whether it takes part in a joint registry or audits revisions and complications.
6. What does the written quote include, and does it prevail? The partner hospital's written quote is the reference for what is covered, not a summary on a website.

What does this mean if I am planning a knee or hip replacement in Istanbul?

Some of our partner hospitals use robotic-arm assistance for joint replacement. The operation and the choice of technique are decided by the surgeon and you. Our job as a coordinator is to put the offers, the written quotes and the questions in front of you.

Because the evidence on how the joint feels is still thin, planning around the operation deserves as much attention as the technique. Our procedure pages suggest allowing about 2 to 3 weeks in Istanbul, with your surgeon confirming when you are fit to fly. Ask how physiotherapy is arranged there, who reviews your progress after you are home, and how your questions reach the surgeon. Our guides on JCI-accredited hospitals and checking the surgeon and planning aftercare cover how to verify those parts.

What to do next

Take these questions to your own orthopaedic surgeon, and ask them how the evidence applies to your joint, your anatomy and your health. If you are comparing offers abroad, send us your imaging and a short medical history through the contact page, and we will pass them to our partner orthopaedic teams for a written opinion and quote.

Med World Türkiye is an Istanbul-based, independent health-travel coordinator that arranges treatment trips with JCI-accredited partner hospitals; it is not a hospital or clinic and is not affiliated with Medworld Clinic in Antalya. See the pages on knee replacement and hip replacement for how the process works.

This page reflects evidence available on 15 October 2026. We will update it when the RACER-Hip trial or longer registry follow-up reports.