Request an appointment
Reserve your spot with our team in just minutes.
Other Knee Treatments
Partial or Total Knee Replacement: Which Do You Need, and Does the Robot Matter?
If you have been told you need a knee replacement, the most useful question you can ask is: does the whole knee need replacing, or only the part that is worn out? Many people are never offered the choice. I think you should be, and so does the national guidance.
I perform total knee replacement, and Mako robotic partial knee replacement of the inner side, the outer side and the kneecap joint, at The Lister Hospital in Chelsea. This page sets out how I decide between them, what the large UK trials show, and what a robot does and does not add.
The short answer
Partial and total knee replacement at a glance
- Your knee has three compartments: inner, outer and behind the kneecap. A partial replacement resurfaces only the worn one and keeps your own ligaments.
- In the main UK trial, partial and total replacement gave similar knee scores at ten years, and partial replacement cost the NHS less.
- Partial replacement means a smaller operation, a faster recovery and a knee that tends to feel more natural.
- In national registry data, partial replacements are revised more often than totals, partly because they are easier to revise. That is the trade-off to weigh.
- A robot makes implant position more precise. For total knees, the largest trial found no difference in how the knee felt at one year. For partial knees, a ten-year trial favoured the robot.
What is the difference?
A total knee replacement resurfaces all three compartments of the knee and removes the anterior cruciate ligament, and sometimes the posterior one as well. It is a dependable operation for a knee that is worn throughout.
A partial knee replacement, also called a unicompartmental replacement or partial knee resurfacing, replaces only the compartment that has worn out. The healthy cartilage, both cruciate ligaments and most of the bone are left alone. Because the ligaments that guide the knee are still there, the knee moves much more like your own.
There are three kinds: the inner (medial) side, which is the most common; the outer (lateral) side; and the joint behind the kneecap (patellofemoral). Where two compartments are worn and the third is healthy, two partial replacements can be combined in one operation and the cruciate ligaments still kept.
Why I offer a partial replacement when the knee allows it
The advantages are practical. The operation is smaller, with less bone removed and less blood loss. Hospital stays are shorter and serious complications are less common. In a study of more than 100,000 patients in the National Joint Registry, for every 100 patients given a partial instead of a total there was about one fewer death in the first four years, alongside about three more reoperations (Liddle, 2014).
The knee also tends to feel better. In one comparison, patients with a partial replacement were more likely to forget they had an artificial joint in daily life, scoring about 74 out of 100 on the Forgotten Joint Score at one year compared with about 59 for total replacement (Zuiderbaan, 2017). That study was not randomised, so the groups may have differed, but it matches what patients tell me.
NICE guidance says patients with arthritis confined to the inner side of the knee should be offered the choice of partial or total replacement (NICE NG157 and QS206). If nobody has discussed a partial replacement with you, it is reasonable to ask why.
What the TOPKAT trial showed
TOPKAT is the trial that matters most here. It randomly allocated 528 patients across 27 NHS hospitals, all with arthritis on the inner side of the knee, to a partial or a total replacement.
At ten years there was no meaningful difference in knee scores between the two groups, and the same proportion in each group, 5 per cent, had needed a revision operation. Partial replacement cost less and was the more cost-effective choice (Beard, 2026). At five years, 91 per cent of patients given a partial said they would have the operation again, compared with 84 per cent given a total (Beard, 2019).
Two cautions. TOPKAT studied the inner side of the knee only, so it does not tell us about outer side or kneecap replacements. And it compared operations done by surgeons who were experienced in both.
The trade-off: revision
The main argument against a partial replacement is that, in national registry data, partials are revised about twice as often as totals (Liddle, 2014). That is real, and you should weigh it.
Part of the explanation is that a partial is easier to revise, so surgeons and patients choose to do so at a lower level of symptoms. A New Zealand registry study found that among patients with a poor result, partials were revised several times more often than totals with the same poor score (Tay, 2023). In other words, an unhappy total knee often stays as it is, while an unhappy partial gets converted. The revision rate in a randomised trial, where that bias is smaller, was the same in both groups.
Arthritis can also progress in the compartments that were left alone. Choosing the right knee for a partial replacement is what keeps that risk low, and it is where the surgeon's judgement matters most.
Kneecap (patellofemoral) replacement
When the wear is only behind the kneecap, replacing the whole knee means removing two healthy compartments and a healthy ligament. A blinded randomised trial of 100 patients with arthritis confined to the kneecap joint found faster recovery and better knee-specific quality of life over two years with a kneecap replacement, and the total knee group had lost about 10 degrees of movement at two years (Odgaard, 2018).
This is an operation where surgeon experience has been shown to matter. National Joint Registry data link surgeon volume and implant design to how long these replacements last (Vella-Baldacchino, 2025). It is an area of particular interest for me, and I perform it with the Mako robotic system.
Does the robot matter? The RACER trial
A robotic system does not operate by itself. I plan the operation on a three-dimensional model of your knee, and the robotic arm holds the cutting tool within the boundaries of that plan. The question is whether that precision changes your result.
For total knee replacement, the answer so far is not clearly. RACER-Knee, a UK trial of 339 patients across 10 hospitals, compared robotic with conventional total knee replacement. Implant position was more precise with the robot, but there was no difference in how natural the knee felt at one year, and the robotic operation cost more (Parsons, 2026). Follow-up to ten years is continuing. I will not tell you a robot guarantees a better total knee, because the best evidence we have does not show it.
For partial knee replacement, the picture is different. A partial implant is small and unforgiving of position, and a few degrees of error can overload it. A randomised trial in Glasgow followed 129 patients for ten years after an inner side partial replacement: none of the robotic knees had been revised, compared with about one in nine of the conventional ones (Blyth, 2025). It is one centre and a modest number of patients, so it is not the final word, but it points the same way as the logic. NICE has said robotic systems can be used in the NHS while more evidence is gathered (NICE HTE22, 2025).
That is why I use the Mako robotic system for partial knee replacement, where accuracy has the most to give. I perform these operations at The Lister Hospital in Chelsea. More detail is on my robotic knee replacement page.
Who is suited to which?
A partial replacement is usually the better choice when the wear is confined to one compartment, or two with the third healthy, the ligaments are working, the knee straightens and bends reasonably well, and the leg is not badly out of line.
A total replacement is usually the better choice when all three compartments are worn, the ligaments are not working, the knee is very stiff or deformed, or there is an inflammatory arthritis affecting the whole joint.
Age by itself does not decide it. Neither does weight. What decides it is the pattern of wear on your X-rays and scan, the examination, and what you want the knee to do.
Before any replacement
We don't stack treatments. We do the right things in the right order.
Some people are offered a replacement before strength, weight and simpler treatments have been given a fair chance. Others are kept on injections and physiotherapy long after the knee has gone past what those can do, and lose years of activity. Both are mistakes. I look at where your knee is on that road and tell you plainly. If you are not there yet, my knee injections and exercise pages explain what comes first. If you are, it is better to operate while you are still strong enough to recover well.
Fixed prices for each operation are on my knee replacement cost page.
How my team and I look after you
I will examine your knee, go through your X-rays and scan with you, and show you which compartments are worn. You will leave knowing whether a partial replacement is possible for your knee, and if it is not, why. My team prepares you before the operation, including strength and nutrition, and stays with you through the recovery with the physiotherapists I work with. My recovery kit page lists what is worth having at home, and you can book an appointment when you are ready.
Told you need a knee replacement?
Book a consultation. I will show you which parts of your knee are worn and whether a partial replacement is an option.
Partial and Total Knee Replacement Questions
For a knee with arthritis in one compartment, a partial replacement gives similar knee scores at ten years, a faster recovery and fewer serious complications, and the knee tends to feel more natural. It is revised more often in registry data. For a knee worn throughout, a total replacement is the right operation.
In the TOPKAT trial, 5 per cent of partial replacements and 5 per cent of total replacements had been revised at ten years. Registry figures for partials are higher, partly because a partial is easier to revise. If a partial does wear out or arthritis progresses elsewhere, it can be converted to a total replacement.
It is more precise. For total knee replacement, the RACER-Knee trial found no difference in how the knee felt at one year. For partial knee replacement, a ten-year randomised trial found fewer revisions with robotic assistance. Precision matters most where the implant is smallest.
Age by itself does not decide it. The pattern of wear, the state of the ligaments and the alignment of the leg matter far more. Partial replacement is used in active patients in their fifties and in patients in their eighties who would benefit from a smaller operation.
Sometimes. If two compartments are worn and the third is healthy with working ligaments, two partial replacements can be combined in one operation. This keeps both cruciate ligaments. Whether it suits you depends on the examination and imaging.
References
- Beard DJ, et al. TOPKAT trial, ten-year results. Lancet Rheumatol. 2026;8:e116-126. doi:10.1016/S2665-9913(25)00250-4
- Beard DJ, et al. The clinical and cost-effectiveness of total versus partial knee replacement in patients with medial compartment osteoarthritis (TOPKAT): 5-year outcomes of a randomised controlled trial. Lancet. 2019;394:746-756. doi:10.1016/S0140-6736(19)31281-4
- Liddle AD, Judge A, Pandit H, Murray DW. Adverse outcomes after total and unicompartmental knee replacement in 101,330 matched patients. Lancet. 2014;384:1437-1445. doi:10.1016/S0140-6736(14)60419-0
- Tay ML, et al. Bone Joint J. 2023;105-B:269-276. doi:10.1302/0301-620X.105B3.BJJ-2022-0872.R2
- Zuiderbaan HA, et al. Unicompartmental knee arthroplasty versus total knee arthroplasty: which type of artificial joint do patients forget? Knee Surg Sports Traumatol Arthrosc. 2017;25:681-686. doi:10.1007/s00167-015-3868-1
- Odgaard A, et al. Clin Orthop Relat Res. 2018;476:87-100. doi:10.1007/s11999.0000000000000017
- Vella-Baldacchino M, et al. J Bone Joint Surg Am. 2025;107(8):819-828. doi:10.2106/JBJS.24.00703
- Parsons N, et al. RACER-Knee trial. Lancet. 2026;408:924-934. doi:10.1016/S0140-6736(26)00986-4
- Blyth MJG, et al. Bone Joint J. 2025;107-B:72-80. doi:10.1302/0301-620X.107B1.BJJ-2024-0245.R2
- National Institute for Health and Care Excellence. Joint replacement (primary): hip, knee and shoulder (NG157), 2020; quality standard QS206; robot-assisted surgery for orthopaedic procedures, early value assessment (HTE22), 2025.
Patient Reviews
What Our Patients Say
Verified Doctify reviews from patients who have received treatment for knee problems under the care of Dr Arj Imbuldeniya at OrthoLongevity.