How the Mako system works inside the operating theatre

The short answer to the question most patients ask first: no, a robot does not perform the surgery. The Mako SmartRobotics system (Stryker) is a precision instrument the surgeon directs at every moment — it cannot make a decision, initiate a cut, or move independently.

The process runs in two stages. Several days before the operation, a CT scan of the patient's knee is taken and used to build a three-dimensional model of that individual's bone anatomy — not a standardised template. The surgeon uses this model to plan exactly where bone needs to be removed and where the implant components will sit, making those decisions before the patient enters theatre.

Intraoperatively, the plan is loaded into the robotic arm, which holds the cutting instrument. The system enforces a virtual boundary around the pre-approved resection zone: if the handpiece approaches healthy bone, cartilage, or soft tissue outside the planned area, the arm's haptic resistance physically prevents it from going further. Think of it as a GPS route combined with a physical guardrail — the destination and the safe corridor are set in advance, and straying outside them is mechanically blocked.

The surgeon guides the arm throughout, responds to what they see, and retains full control. The robot's role is to hold the cut precisely within the agreed plan, protecting the ligaments and periarticular tissue surrounding the knee from inadvertent damage.

Bone alignment precision — what the accuracy data shows

Precise bone cutting matters because a knee implant that sits even a few degrees outside the correct mechanical axis tends to load unevenly — accelerating wear, stressing the soft tissues, and contributing to the persistent pain that accounts for a significant share of TKR dissatisfaction. Getting the alignment right is not a surgical technicality; it is directly tied to how the knee feels years later.

Published accuracy data for Mako in TKR is encouraging. In a prospective real-world study of 155 patients, Wong et al. (2024) found that the hip–knee–ankle (HKA) axis was restored to within 0.76° of the intraoperative plan on average, and 98.1% of cases achieved a flexion–extension gap balance difference of 2 mm or less — the threshold associated with stable soft-tissue tension through the range of motion. Sires et al. (2021) reported that 94.29% of bone resections in 105 patients fell within 1 mm of the planned cut, with a mean coronal alignment deviation of 0.78°.

The outlier comparison is the clearest way to frame what this means in practice. Alignment outliers — typically defined as deviating more than 3° from the neutral HKA axis — occur in fewer than 10% of Mako cases, compared with rates exceeding 40% reported for conventional manual technique. Reducing that outlier pool matters because it is within the outlier group that the risk of dissatisfaction and early revision concentrates.

For unicompartmental knee replacement candidates, this precision evidence carries additional weight: in a procedure where the implant corrects only one compartment, even modest alignment errors can shift load onto the remaining cartilage and increase revision risk.

Accuracy is a surrogate measure, not an outcome guarantee. Achieving the planned position reliably does, however, remove a known and preventable source of problems.

TKR patient outcomes — what the evidence currently supports

For many patients considering surgery, the real question is simpler than the technical detail: will Mako TKR leave them with a better-functioning knee? Current evidence suggests a moderate advantage in patient-reported outcomes at medium- and longer-term follow-up — though that advantage has not yet been confirmed by a large randomised controlled trial.

The most comprehensive synthesis available is a 2025 Bone & Joint systematic review spanning 22 studies and 3,738 TKAs. Compared with manual TKR, Mako robotic-assisted TKA produced superior pooled patient-reported outcome scores at medium-term (SMD 0.46) and long-term follow-up (SMD 0.40). The Forgotten Joint Score — which captures how rarely patients are aware of their knee during daily life — showed a meaningful advantage for robotic TKA at medium-term follow-up (MD 5.50, 95% CI 2.19–8.81). Effect sizes are moderate rather than dramatic, and short-term differences are modest: the functional benefit appears to accumulate over time rather than emerge in the early weeks after surgery.

Uptake reflects growing surgical confidence in the system: by 2023, Mako robotic-assisted TKR had become the most performed primary TKR modality in Australia. For surgical teams, it also compares favourably in theatre efficiency — one single-centre study found Mako approximately 11 minutes faster than computer-navigated TKR, though comparisons with conventional manual technique depend considerably on surgeon experience and learning curve.

The critical evidence gap is a large randomised controlled trial. The NIHR-funded RACER-knee study — 332 participants randomised equally to robotic-assisted or conventional TKR — is under way, with the Forgotten Joint Score at 12 months as its primary outcome. Its findings will provide the clearest test yet of whether the alignment precision described in the previous section translates into measurable patient benefit under controlled conditions.

UKR outcomes — where the evidence is most compelling

Unicompartmental knee replacement has stricter candidacy criteria than TKR, and the margin for positioning error is narrower. Because a UKR implant resurfaces only one compartment, even modest malalignment shifts load onto the cartilage that remains — making accurate component placement a direct determinant of revision risk. This is why the precision argument for robotic assistance is particularly relevant here, and why the UKR evidence is, at present, more definitive than for TKR.

The most informative study is a five-year randomised controlled trial by Banger et al. (2021, Bone & Joint), involving 130 patients randomised to robotic arm-assisted or manual medial UKA. The finding was clear: the reintervention rate in the robotic group was 0%, against 9% in the manual group (p<0.001). Patient-reported outcome scores at five years were equivalent between the two groups — robotic UKA did not produce meaningfully better questionnaire results. What it produced was fewer patients requiring a further operation. Avoiding a second procedure is itself a substantial clinical benefit, even when function scores look similar on paper.

Broader data support this direction. A 2023 meta-analysis of 1,060 UKAs found robotic-assisted UKA significantly improved patient satisfaction compared with manual technique (OR 1.72, 95% CI 1.25–2.37). A 2025 systematic review found CT-planned, image-based robotic UKA generates fewer radiologic outliers in hip–knee–ankle angle and joint-line height than imageless systems — though revision rates between robotic types were not significantly different in that analysis. Registry figures suggest 10-year survivorship of 91–96% and a possible revision risk up to 45% lower than for conventional UKA; registry comparisons carry inherent confounders and warrant appropriate caution.

For well-selected patients, robotic UKR also supports rapid discharge. A feasibility study of 19 patients demonstrated safe discharge within 24 hours (mean stay 19.5 hours), a median Oxford Knee Score of 44/48 at six months, and no complications or readmissions — relevant where day-surgery pathways are being considered.

What the evidence does not yet confirm

Three gaps in the current evidence deserve honest acknowledgement, distinct from the caveats already noted in the TKR and UKR sections above.

Long-term TKR revision rates. Precise implant positioning is mechanistically likely to protect against early wear and instability, but whether the alignment gains translate into measurably lower revision rates at 10 or 15 years has not been established for total knee replacement. The UKR evidence — including the five-year randomised trial showing 0% reintervention in the robotic group — is more advanced on this specific question. For TKR, RACER-knee is the study most likely to close that gap, though its primary outcome is patient-reported function at 12 months rather than long-term revision.

Cost-effectiveness. Mako carries higher system acquisition and per-case costs than conventional manual TKR. Whether those costs are offset by clinical gains — fewer reinterventions, shorter inpatient stay, reduced long-term complications — has not been demonstrated in published economic analysis. That evidence will depend substantially on trial and registry outputs not yet available.

Which patients benefit most. Outcome data currently pool patients broadly by procedure type. Which individuals — differentiated by deformity severity, body mass, activity level, or age — derive the greatest advantage from robotic assistance has not been systematically characterised. The precision rationale applies across the patient group; how much it matters for a given individual is a question that clinical assessment, rather than population-level data, needs to answer.

None of these gaps calls the technology into question. They reflect a maturing evidence base, and individual suitability remains a matter for specialist evaluation.

Who Mako is used for and how to find out if it applies to you

The decision starts not with whether Mako is available, but whether knee replacement is the right operation at all. Mako is used for two procedures — total knee replacement for end-stage tri-compartmental osteoarthritis, and unicompartmental knee replacement for isolated single-compartment disease in well-selected patients. It is the delivery mechanism for the chosen procedure, not a route around establishing which procedure is appropriate.

Candidacy turns on the pattern of cartilage loss, limb alignment, BMI, activity level, and general health. The starting point is a consultant assessment: weight-bearing X-ray and, in most cases, MRI to map cartilage and soft tissue before any surgical pathway is discussed. For more active or younger patients, objective gait analysis — Lincolnshire Knee offers MAI Motion® biomechanical assessment — can quantify load distribution and movement asymmetry as part of the pre-surgical picture. Where robotic-assisted surgery is planned, a preoperative CT scan is required for the planning stage; patients should expect this as part of the standard workup, not as an additional burden.

Lincolnshire Knee is part of the MSK Doctors group and accepts patients without a GP referral. Book a consultant assessment at lincolnshireknee.co.uk.

Across the evidence reviewed, the consistent finding is that robotic assistance improves the mechanical precision of bone cutting — and for UKR that precision measurably reduces reinterventions. For TKR, functional advantages are emerging at medium-term follow-up but await high-quality RCT confirmation. The more useful question for any individual patient is not whether robotic surgery works in principle, but whether the underlying procedure is indicated. That is a clinical judgement, not a technology preference.

1. [1] A Comparison of the Short-Term Outcomes Following Robotic-Assisted and Computer-Navigated Total Knee Replacements. (2024). https://doi.org/10.1093/bjs/znae163.049 https://doi.org/10.1093/bjs/znae163.049
2. [2] Achieving discharge within 24 h of robotic unicompartmental knee arthroplasty may be possible with appropriate patient selection and a multi-disciplinary team approach. (2020). https://doi.org/10.1016/j.jor.2020.01.051 https://doi.org/10.1016/j.jor.2020.01.051
3. [3] Operative Time Differences in Primary TKA Between Navigated And Robotic Technology Assistance. (2025). https://doi.org/10.1177/2325967126S00014 https://doi.org/10.1177/2325967126S00014