Am I a Candidate for Partial Knee Replacement Instead of a Total?

Overview

Robotic-assisted partial knee replacement, or unicompartmental knee arthroplasty (UKA), resurfaces only the damaged compartment, preserving the rest of the joint's bone and ligaments. Candidates return to sport more often than after total knee replacement (80.7% vs 69.0%) and report higher satisfaction (90.1% vs 81.3%), but registry data also show a real, higher revision rate (Vossen et al. 2025; Jansen et al. 2020).

Who this is for

Partial knee replacement is for patients whose arthritis is genuinely confined to one compartment of the knee — most often the medial side — with the rest of the joint's cartilage and ligaments intact. The traditional Kozinn and Scott selection criteria (age over 60, weight under 180 lb, minimal deformity, no heavy labor) have been substantially relaxed by modern evidence: a 1,000-knee series found 68% of patients would have been considered "contraindicated" by those classic rules, yet had no difference in 15-year implant survival and, in several measures, better function than the "ideal" group (Hamilton et al. 2017). What has not been relaxed are two things: inflammatory arthritis and significant ligament instability remain contraindications, and disease that has spread beyond one compartment — particularly meaningful patellofemoral bone loss — still points to a total knee replacement instead (OrthoInfo/AAOS; PMC6026888). An ACL-deficient knee is not an automatic exclusion, but it is a specialist-level decision rather than a routine one — one well-matched series found equivalent 5-year survival with or without an intact ACL, while other reviews report failure rates as high as 16% in ACL-deficient knees without reconstruction (Kikuchi et al. 2021; Joints review).

How I approach it

I think the pendulum on who qualifies for a partial knee has swung too far toward "almost anyone with one bad compartment," and I want to be candid about where I stand personally rather than just citing the literature. The Hamilton data is genuinely persuasive that age, weight, and heavy labor by themselves shouldn't disqualify a patient (Hamilton et al. 2017), and I use that evidence to broaden who I'll consider. But I still treat inflammatory arthritis and meaningful ligament instability as hard stops, and I look closely at the patellofemoral compartment on imaging before I commit to a partial rather than a total.

The harder thing I want to say directly: the biggest single factor in how well a partial knee replacement does over time isn't the implant or even the robot — it's how much of this specific operation the surgeon actually does. Registry data are unambiguous that survival at 5 years runs 96.0% for the highest-volume UKA surgeons and 90.1% for the lowest-volume, and that surgeons who use UKA as 40–60% of their knee replacement practice get materially better results than surgeons who use it in only a handful of cases a year (Baker et al. 2013; Optimal usage of UKA, Bone Joint J 2015). I do a moderate volume of arthroplasty overall, and partial knee replacement is a smaller share of that than total knee replacement is. I don't think that disqualifies me from doing this operation well, but I do think a patient deserves to hear it stated plainly rather than glossed over, and I take it seriously in how I select cases: I lean toward partial knee replacement in patients with a clean, single-compartment pattern and away from it when the anatomy is borderline, precisely because that is where surgeon experience matters most. If a patient's knee is anatomically borderline for a partial, I would rather have an honest conversation about that than push a smaller operation past its evidence base.

The operation

Robotic-assisted partial knee replacement begins with either preoperative imaging or intraoperative registration that builds a three-dimensional map of the diseased compartment. The robotic arm guides bone preparation within a pre-planned envelope and allows the surgeon to check and fine-tune ligament balance and component position before the implant is finally seated. Only the damaged compartment is resurfaced; the cruciate ligaments, the healthy cartilage in the other compartments, and the native bone stock elsewhere in the joint are left alone (OrthoInfo/AAOS). Robotic guidance measurably improves the precision of component placement: in a Level I randomized trial, 80% of tibial components landed within 2 degrees of the sagittal-plane target with robotic assistance, compared with 22% using conventional instrumentation, with similar gains across femoral positioning parameters (Bell et al. 2016). Even with the robot, roughly half to four-fifths of components land within that 2-degree window depending on which plane is measured — the technology narrows the spread of outliers, it doesn't eliminate them. Robotic assistance adds operative time, on the order of 15.6 minutes on average across pooled data (Bensa et al. 2024).

Risks and honest tradeoffs

Partial knee replacement carries a different risk profile than total knee replacement, and most of the differences favor the partial knee. Early (45-day) mortality is markedly lower, with a pooled relative risk of 0.27 in registry and database studies (Wilson et al. 2019). Venous thromboembolism is also less common, with an odds ratio of 0.49 in matched registry data (Liddle et al. 201460419-0/fulltext)). Length of hospital stay runs 1.2 to 1.7 days shorter across randomized, registry, and cohort data (Wilson et al. 2019). The tradeoff that runs the other way is revision: 8-year implant survival in the largest matched registry was 87.0% for partial knee versus 94.6% for total knee, a hazard ratio of 2.12 (Liddle et al. 201460419-0/fulltext)) — a gap that the section above explains is narrowed, but not eliminated, once revision threshold and surgical volume are accounted for. Robotic-specific risks include tracking-pin complications, reported at an overall rate of 1.4% across pooled series, most commonly superficial infection or pin dislodgement at the tibial site, with all reported complications resolving by follow-up (Thomas et al. 2022). One registry signal is worth stating plainly: revision for infection was higher for one specific robotic implant than for the best-performing conventional implant studied against it, a hazard ratio of 2.91 (St Mart et al. 2020).

Recovery and rehabilitation

Recovery from partial knee replacement is generally faster than from a total knee replacement, largely because less bone and soft tissue are disturbed during surgery.

• Day 0–1: Most patients are stood up and walking within 24 hours, and many go home the same day or within 24 hours of surgery (Oxford NDORMS patient booklet). A systematic review of nearly 9,700 patients found successful same-day discharge in 88% overall, rising to 91% in carefully selected patients (Bayoumi et al. 2022).
• 0–2 weeks: Walking aids are recommended for about two weeks; patients are advised not to discard crutches too early (Oxford NDORMS patient booklet).
• 6 weeks: Most patients are walking without aids and can drive once off crutches and in full control of the vehicle; leisure activities and gym work can typically begin in this window (Oxford NDORMS patient booklet).
• 3 months: Strenuous activity should still be avoided up to this point, and this is generally when return-to-work timing is finalized depending on job demands (Oxford NDORMS patient booklet).
• 6–12 months: High-impact activity and sport are typically cleared in this window, with 80.7% of partial knee patients returning to some level of sport, compared with 69.0% after total knee replacement, and full recovery continuing to progress up to a year (Vossen et al. 2025; Oxford NDORMS patient booklet).

1. Day 0–1

   Most patients are stood up and walking within 24 hours, and many go home the same day or within 24 hours of surgery (Oxford NDORMS patient booklet).

   A systematic review of nearly 9,700 patients found successful same-day discharge in 88% overall, rising to 91% in carefully selected patients (Bayoumi et al. 2022).

2. 0–2 weeks

   Walking aids are recommended for about two weeks; patients are advised not to discard crutches too early (Oxford NDORMS patient booklet).

3. 6 weeks

   Most patients are walking without aids and can drive once off crutches and in full control of the vehicle.

   leisure activities and gym work can typically begin in this window (Oxford NDORMS patient booklet).

4. 3 months

   Strenuous activity should still be avoided up to this point.

   and this is generally when return-to-work timing is finalized depending on job demands (Oxford NDORMS patient booklet).

5. 6–12 months

   High-impact activity and sport are typically cleared in this window.

   with 80.7% of partial knee patients returning to some level of sport, compared with 69.0% after total knee replacement, and full recovery continuing to progress up to a year (Vossen et al. 2025; Oxford NDORMS patient booklet).

Alternatives I considered

For a patient whose arthritis has already spread beyond one compartment, or who has inflammatory arthritis or real ligament instability, total knee replacement remains the more defensible choice, discussed in detail on the robotic total knee replacement page. For a younger patient with a focal cartilage defect rather than established arthritis, cartilage restoration is the earlier-stage option and is the right conversation before arthroplasty of any kind is on the table — covered on the cartilage repair hub.