Robotic Knee Replacement in Bristol – Mako

The single most important decision in knee replacement surgery is choosing the right operation: partial or total, and which compartment to address. Get that wrong, and no amount of technology will save you. Get it right, and robotic assistance is a precision tool that helps me position implants accurately and consistently.

I use the Mako robotic system for both partial knee replacement (medial, lateral, or patellofemoral) and total knee replacement. This page explains what robotic assistance actually does, where I find it useful, and what the trials and registry studies show it does and does not improve.


What does the robot actually do?

The robot does not perform the operation. I do. What it provides is a planning and guidance system that helps me:

  • Build a precise implant plan based on your individual knee anatomy
  • Prepare bone accurately within the planned boundaries
  • Fine-tune alignment and knee balance during the operation itself

Think of it as a precision instrument, not an autopilot. I use it most often for total knee replacement and patellofemoral replacement, where I value the ability to plan and balance the knee in real time. Trials in total and partial knee replacement show that a robot makes the position of the implant more accurate.1, 2 So far they have not shown that patients do better as a result, and I set out the evidence below.


What I use robotic assistance for

I find robotic assistance most useful for:

  • Accuracy of component positioning – the system enforces the plan, so there is less variability between cases1
  • Knee balancing – I can check ligament tension and alignment through the full range of movement during the operation, and adjust in real time
  • Bone preparation – bone is removed only within the planned boundaries, which matters for both total and partial replacement
  • Complex anatomy – in knees with unusual alignment or previous surgery, the additional planning data is most useful to me

It is still a tool. The outcome depends on the surgeon's decision-making before and during the operation, not the brand of robot in the room.


What the evidence shows

Robotic knee replacement has now been tested in randomised trials and followed in national joint registries, and in 2025 NICE assessed robot-assisted knee and hip replacement.3 NICE allows six robotic systems, including Mako, to be used in the NHS while more evidence is collected. It noted that Mako had the most mature evidence and that the other systems had less, generally of lower quality. The evidence answers three different questions, and the answers are not the same.

Accuracy of implant position

The implant does end up closer to the plan. In a meta-analysis of 12 randomised trials of total knee replacement (2,200 patients), knees operated on with a robot were less than half as likely to end up more than 3° away from the straight leg alignment that was the target (risk ratio 0.43), and were on average about 1° closer to it.1 In a randomised trial of partial knee replacement, the forward and backward tilt of the tibial (shin bone) component was within 2° of the plan in 80% of robotic operations and 22% of conventional ones.2 NICE likewise found implant alignment consistently more precise with a robot, but was uncertain whether more precise alignment results in better clinical outcomes.3

How patients do afterwards

So far, the greater accuracy has not been shown to give patients a better result. NICE concluded that improvements in patient-reported outcomes and complications were similar with robotic and conventional surgery.3 A 2026 meta-analysis of 24 randomised trials of total knee replacement (3,425 patients) found no significant difference in knee function scores, pain, range of movement or length of hospital stay, and rated the certainty of the evidence as low to very low.4 Individual trials do not all agree. One of those 24, in 180 patients and using a handheld imageless robot, reported better knee scores at 12 months.5 RACER-Knee, a UK trial published too late to be included in that analysis, found no difference.6 I describe it below, along with the trials of partial replacement, where some of the results favour the robot.

Revision and complications

The UK evidence on this comes from the National Joint Registry. An analysis of 697,145 knee replacements recorded between 2018 and 2024, of which 22,111 were robotic, compared matched groups of patients. Estimated implant survival at five years was 98.6% for robotic and 98.5% for conventional total knee replacement, and 96.4% and 97.8% for partial knee replacement. There was no significant difference in the risk of revision, or in complications during the operation, for either operation.7 Follow-up averaged 2.5 years, so a difference that only appears later cannot be ruled out.


Robotic total knee replacement

Total knee replacement is where I use robotic assistance most frequently. When arthritis affects more than one compartment, the robot helps me plan and execute a replacement built around your specific anatomy, rather than relying on cutting guides set to standard angles.

In total knee replacement, the robotic system helps me:

  • Plan component positions matched to your specific anatomy
  • Restore alignment and joint line position in a controlled way
  • Balance the knee through flexion and extension – checking that it feels stable and tracks well before closing

This real-time balancing is where I find the robot adds the most value. I can assess ligament tension throughout the full arc of movement and make adjustments during the operation, rather than relying solely on pre-operative planning.

A large blinded trial of the Mako system in total knee replacement, RACER-Knee, was published in The Lancet in 2026. It randomly allocated 339 patients at ten hospitals in Great Britain to total knee replacement with the Mako system or with conventional instruments, and neither the patients nor those assessing them knew which they had had. A year after their operation, patients in both groups had less pain and less disability than before it, with no difference between the groups. At 12 months the Forgotten Joint Score, which measures how aware you are of the knee in daily life on a scale of 0 to 100 (100 is best),8 averaged 49.2 after robotic surgery and 50.2 after conventional surgery. Pain over the first three days and the use of painkillers were also no different. The robot was more precise on some X-ray measures and similar on others. Follow-up is continuing to ten years.6


Robotic patellofemoral replacement

Patellofemoral replacement is the other procedure where I find robotic assistance particularly valuable. When arthritis is confined to the front of the knee – behind the kneecap and in the trochlear groove – kneecap tracking is very sensitive to component position.9 Robotic guidance helps me place the trochlear component where I have planned it.

The published evidence for this operation is still limited. A 2026 systematic review found no randomised trials of robotic patellofemoral replacement. Pooling the comparative studies that exist, it found a lower risk of revision with a robot (relative risk 0.75), a result that was only just statistically significant, and one of the two studies that reported X-ray results found better correction of kneecap tilt than with conventional surgery. Its authors concluded that definitive conclusions could not yet be drawn.10 In the Australian joint registry, the cumulative revision rate at six years was 12.7% after 669 robotic and 11.6% after 1,481 conventional patellofemoral replacements, which is not a significant difference.11


Robotic partial knee replacement (medial and lateral)

A partial knee replacement replaces only the worn compartment, preserving the healthy parts of the knee – bone, cartilage, and ligaments. Robotic assistance can be used for partial replacement, though I do not use it routinely for every case.

Medial (inner side)

Medial unicompartmental replacement is the most common partial replacement. When the arthritis is confined to the inner compartment, robotic guidance can help with implant positioning and preserving the rest of the knee.

Medial replacement has more trial evidence than the other partial operations, and some of it favours the robot. In a randomised trial of 139 patients, the Mako system placed the components more accurately than conventional instruments,2 and pain scores over the first eight weeks were lower.12 At five years there was no difference in any patient-reported outcome, but none of the robotic group had needed further surgery, compared with 9% of the conventional group.13 The two groups received different designs of implant, so the comparison is not of the robot alone. Interim results from a second trial, of 107 patients, showed less pain in hospital, a shorter stay and a better Forgotten Joint Score at two years with the robot, with no difference in the Oxford Knee Score.14 In the National Joint Registry analysis, the risk of revision after partial replacement was no different with or without a robot.7

Lateral (outer side)

Lateral unicompartmental replacement addresses arthritis on the outer side. The lateral compartment differs from the medial side in its shape and in the way it moves, and the operation is performed less often,15 so robotic planning can be helpful for alignment and tracking. In a comparison of 52 robotic and 43 conventional lateral replacements followed for at least five years, clinical scores and revision rates were no different.16

The decision between partial and total replacement is far more important than whether a robot is used. A well-performed knee replacement without robotic assistance will outperform a robotic-assisted operation where the wrong procedure was chosen.


Who is robotic knee replacement right for?

If you need a knee replacement, robotic assistance is worth considering when:

  • You are having a total knee replacement, where I use it to plan the implant position and balance the knee
  • You are having a patellofemoral replacement, where accurate component position directly affects kneecap tracking
  • Your anatomy or alignment is unusual, where I find detailed pre-operative planning most useful
  • You have looked at the evidence and the cost, and would prefer the implant to be positioned with robotic guidance

It has not been shown to give a better outcome, and it is not necessary for every case. NICE noted that robotic surgery may be more beneficial in complex cases.3 RACER-Knee did not include patients with a previous fracture or who needed complex implants,6 so the trials have not yet tested this. The priority is always choosing the correct operation for your knee and getting rehabilitation right afterwards.


How I assess and plan robotic knee replacement

Every robotic knee replacement starts with a thorough pre-operative assessment. The most important question is not "should I have a robot?" but "which operation is right for my knee?"

I assess the pattern of your arthritis with weight-bearing X-rays (including alignment views), examine the knee for stability, movement, and kneecap tracking, and talk through what is actually limiting you day to day. The Mako system also needs a CT scan for planning before the operation.

The CT scan involves a dose of radiation. In a study of 236 planning scans published in 2015, the average effective dose was 4.8 mSv, with a three-fold difference between institutions.17 For comparison, the average person in the UK is exposed to about 2.6 mSv of radiation a year from all sources,18 so the scan in that study was equivalent to a little under two years of that exposure.

Once I am confident we have selected the right procedure, the robotic plan is built around your anatomy.


What happens during the operation?

The exact steps depend on the procedure, but the general sequence is:

  • Before the operation, I create a surgical plan from the CT scan of your knee
  • During surgery, I register landmarks and confirm the plan against your real anatomy
  • The robotic arm guides bone preparation – I control the saw or burr, and the system constrains it to the planned boundaries
  • I place the implants, then check stability, alignment, and range of movement before finishing

The operation takes a little longer than a conventional knee replacement. In RACER-Knee, a robotic total knee replacement took on average 10.5 minutes longer,6 and across comparative studies a robotic partial knee replacement took about 16 minutes longer.19 The anaesthetic and the main incision are the same, but the pins that hold the robot's tracking markers are often placed through separate small cuts.20 In RACER-Knee these were two cuts of 1 cm.8


Recovery and rehabilitation

Recovery after robotic knee replacement follows the same principles as conventional surgery. The main drivers are:

  • Good pain control and swelling management in the first few weeks
  • Early, safe mobilisation – most patients are up and walking the same day or the day after surgery
  • Structured physiotherapy with progressive strengthening
  • Time – the knee takes months to settle fully, and patience is important

Partial knee replacement generally recovers faster than total, regardless of whether a robot is used. In a review of 60 studies that I co-authored, partial knee replacement was associated with a shorter hospital stay and a quicker return to work and to sport compared with total knee replacement.21 Individual recovery also depends on pre-operative fitness, body weight, and how well the knee responds to rehabilitation.


Risks and limitations

Robotic knee replacement carries the same risks as conventional knee replacement, and two of its own:

  • Infection
  • Blood clots (DVT / pulmonary embolus)
  • Stiffness or prolonged swelling
  • Persistent pain or dissatisfaction
  • Instability, loosening, or wear over time
  • Pin-site problems – the pins that hold the robot's tracking markers can cause a wound problem or, very rarely, weaken the bone and cause a fracture. In a review of 15,874 robotic total knee replacements, 0.59% had a wound problem at a pin site and 0.15% (about 1 in 660) had a fracture20 (this is specific to robotic and computer-navigated surgery)
  • Radiation from the planning CT scan

The risks shared with conventional surgery do not appear to be higher or lower with a robot. In RACER-Knee, 16 patients in each group had a serious adverse event,6 and in the National Joint Registry analysis neither complications during the operation nor later revisions differed.7 Studies from other countries are not all consistent. In Ontario, Canada, a major complication within a year (revision, or deep infection or fracture needing surgery) followed 2.0% of robotic total knee replacements and 1.0% of matched conventional ones.22 In Finland, revisions within 12 months were no different overall, and fewer were for infection after robotic surgery.23

There is also a learning curve with any robotic system. Across 31 studies of robotic total knee replacement it was a median of 17 operations, during which operations took up to 17 minutes longer. Complication rates and patient-reported outcomes were no different between the learning phase and afterwards.24

Robotic assistance makes implant positioning more accurate, but it does not remove the risks of surgery and has not been shown to reduce them. I discuss the realistic benefits and limitations with every patient before we decide whether to use it.


What it costs

Robotic assistance adds to the cost. For patients paying for their own treatment, the guide price at Nuffield Health Bristol Hospital is £17,540 for a robotic-assisted (Mako) total knee replacement, compared with £16,540 for a conventional one, and £16,500 for a robotic-assisted (Mako) partial knee replacement, compared with £15,500. These are guide prices only: the hospital provides a fixed-price quote after your consultation, once the exact procedure is confirmed. My costs and fees page has the details.

Whether the extra cost is good value for a health service is not settled. In the NHS, RACER-Knee found that a robotic total knee replacement cost around £950 more than a conventional one and was not cost-effective over the first year.6 For partial knee replacement, a ten-year analysis of one randomised trial found robotic surgery to be cost-saving where more than 100 robotic operations are performed a year, because fewer patients needed further surgery.25


Frequently asked questions

Does the robot perform the operation?

No. I perform the operation. The robot is a guidance system that helps me plan implant positions and prepare bone accurately. I make all the surgical decisions.

Is robotic knee replacement better than standard?

It is more accurate, but it has not been shown to give patients a better result. In randomised trials the implant ends up closer to its planned position with a robot, but when the trials are pooled, pain and knee function are no different. In the RACER-Knee trial of 339 patients having a total knee replacement, the Forgotten Joint Score at 12 months averaged 49.2 out of 100 with the Mako robot and 50.2 without. The most important factor is still the surgeon's experience and decision-making. Choosing the right operation and positioning the implant well matters more than the technology used to get there. Robotic assistance helps with the "how", but the surgeon determines the "what" and "why".

Can I have robotic patellofemoral replacement?

Yes – this is one of the procedures where I most often use robotic assistance. Kneecap tracking is very sensitive to the position of the trochlear component, and the robot helps me place it where I have planned it. There are no randomised trials of robotic patellofemoral replacement yet, so how much difference it makes to the result is not known.

Does robotic surgery speed up recovery?

For total knee replacement, no: the RACER-Knee trial found no difference in pain over the first three days or in the use of painkillers. For partial knee replacement, possibly in the early weeks: two randomised trials found less pain in the first days and weeks after robotic surgery, and one of them a shorter hospital stay, with little or no difference in knee scores later on. Recovery depends mainly on the type of operation (partial versus total), pain and swelling control, and rehabilitation.

Which robotic system do you use?

I use Mako, an established platform for robotic knee replacement. It uses a CT scan taken before the operation to build the plan, and it supports partial (medial, lateral, patellofemoral) and total knee replacement. The most important factor is always the surgeon's planning and execution, not the specific platform.

Can I be assessed for robotic knee replacement in Bristol?

Yes. I offer robotic partial and total knee replacement at my Bristol clinics. The first step is a consultation to work out whether you need a knee replacement, and if so, which type is right for your knee.


References

  1. Ruangsomboon P et al. Clinical and radiological outcomes of robotic-assisted versus conventional total knee arthroplasty: a systematic review and meta-analysis of randomized controlled trials. Acta Orthopaedica. 2023. PubMed
  2. Bell SW et al. Improved Accuracy of Component Positioning with Robotic-Assisted Unicompartmental Knee Arthroplasty: Data from a Prospective, Randomized Controlled Study. Journal of Bone and Joint Surgery (American volume). 2016. PubMed
  3. National Institute for Health and Care Excellence. Robot-assisted surgery for orthopaedic procedures: early value assessment. HealthTech guidance HTG743. 2025. NICE
  4. Abdelaziz O et al. Comparative pain relief and functional outcomes following robotic-assisted versus conventional total knee arthroplasty: a GRADE-assessed meta-analysis with meta-regression of randomized controlled trials. Journal of Robotic Surgery. 2026. PubMed
  5. Bollars P et al. Imageless handheld robotic-assisted total knee arthroplasty showed better clinical outcomes than conventional total knee arthroplasty: A randomized controlled trial with preliminary results at 1-year follow up. The Knee. 2025. PubMed
  6. Parsons H 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. PubMed
  7. Mohammad HR et al. Early national comparison of robotic versus conventional knee replacements for arthritis using National Joint Registry data: target trial emulation study. BMJ. 2026. PubMed
  8. Griffin J et al. Robotic Arthroplasty Clinical and cost Effectiveness Randomised controlled trial (RACER-knee): a study protocol. BMJ Open. 2023. PubMed
  9. Yamawaki Y et al. Internal Rotation, Varus, and Anterior Femoral Component Malalignments Adversely Affect Patellofemoral Joint Kinematics in Patellofemoral Arthroplasty. Arthroplasty Today. 2023. PubMed
  10. Williamson TR et al. Robotic arm-assisted patellofemoral joint arthroplasty achieves good short-term survival and functional outcomes: A systematic review and meta-analysis. Knee Surgery, Sports Traumatology, Arthroscopy. 2026. PubMed
  11. Saidy E et al. Can Robotic-Assisted Arthroplasty Change the High Early Revision Rate After Patello-femoral Arthroplasty? An Analysis From the Australian Orthopaedic Association National Joint Replacement Registry. Journal of Arthroplasty. 2026. PubMed
  12. Blyth MJG et al. Robotic arm-assisted versus conventional unicompartmental knee arthroplasty: Exploratory secondary analysis of a randomised controlled trial. Bone & Joint Research. 2017. PubMed
  13. Banger M et al. Robotic arm-assisted versus conventional medial unicompartmental knee arthroplasty: five-year clinical outcomes of a randomized controlled trial. The Bone & Joint Journal. 2021. PubMed
  14. Kayani B et al. A prospective double-blinded randomized controlled trial comparing conventional jig-based versus robotic arm-assisted medial unicompartmental knee arthroplasty. The Bone & Joint Journal. 2026. PubMed
  15. Hong J et al. Lateral unicompartmental knee arthroplasty anatomy, indications, technique, and outcomes: a narrative review. Archives of Orthopaedic and Trauma Surgery. 2025. PubMed
  16. Maritan G et al. Similar survivorship at the 5-year follow-up comparing robotic-assisted and conventional lateral unicompartmental knee arthroplasty. Knee Surgery, Sports Traumatology, Arthroscopy. 2023. PubMed
  17. Ponzio DY, Lonner JH. Preoperative Mapping in Unicompartmental Knee Arthroplasty Using Computed Tomography Scans Is Associated with Radiation Exposure and Carries High Cost. Journal of Arthroplasty. 2015. PubMed
  18. UK Health Security Agency. Ionising radiation: dose comparisons. Updated 2026. GOV.UK
  19. Kanakala S et al. Robotic Assisted Versus Manual Unicompartmental Knee Arthroplasty: A Systematic Review and Meta-Analysis. Orthopaedic Surgery. 2026. PubMed
  20. Ade-Conde M et al. Pin-related complications in robotic-assisted total knee arthroplasty: a systematic review and meta-analysis of 15,874 knees. Journal of Robotic Surgery. 2026. PubMed
  21. Wilson HA et al. Patient relevant outcomes of unicompartmental versus total knee replacement: systematic review and meta-analysis. BMJ. 2019. PubMed
  22. Pincus D et al. Association Between Adoption of Robotic Total Knee Arthroplasty in Canada and Major Surgical Complications. Journal of Arthroplasty. 2026. PubMed
  23. Alakylä KJ et al. Early revision rate in robotic-assisted versus conventional total knee arthroplasty: a prospective cohort study 2021-2023 from the Finnish Arthroplasty Register. Acta Orthopaedica. 2026. PubMed
  24. Abdel Khalik H et al. The learning curve of robotic-assisted total knee arthroplasty: a systematic review and meta-analysis. Journal of Robotic Surgery. 2025. PubMed
  25. Blyth MJG et al. Robotic arm-assisted medial compartment knee arthroplasty is a cost-effective intervention at ten-year follow-up. The Bone & Joint Journal. 2025. PubMed

Related knee topics