Robotic Knee Replacement Shows No Revision Advantage in Registry Study

Key Takeaways

• Among patients undergoing knee replacement for arthritis in UK public and private hospitals, matched robotic and conventional total-knee procedures showed no statistically significant difference in revision risk.
• Matched unicompartmental procedures also showed no statistically significant difference in revision risk between approaches.
• Cause-specific revision and intraoperative complication risks did not differ significantly between approaches for either operation.
• The proportion requiring a two-stage revision did not differ significantly for either operation.

For people undergoing knee replacement for arthritis, more precise implant positioning with robotic assistance raises a clinical question: does that precision translate into fewer failed implants? Total and unicompartmental replacements warrant separate consideration because their underlying revision risks differ. For each procedure, investigators compared robotic and conventional approaches to examine the risk of implant failure.

They retrospectively examined procedures from January 2018 through December 2024 in public and private hospitals across England, Wales, Northern Ireland, the Isle of Man, Guernsey, and Jersey. The cohort comprised 697,145 knee replacements: 675,034 conventional and 22,111 robotic. Separate total knee replacement (TKR) and unicompartmental knee replacement (UKR) analyses used propensity-score matching to estimate the average treatment effect among patients receiving robotic surgery. Matching addressed measured characteristics including age, sex, diagnosis, American Society of Anesthesiologists physical status, fixation, surgery year, thromboprophylaxis, and surgeon caseload. Kaplan–Meier analysis estimated implant survival, while Cox regression assessed revision risk; revision meant removing, exchanging, or adding an implant component. Mean follow-up was 2.5 years in each matched procedure-by-treatment group, although five-year survival was an estimate.

For matched TKR, estimated five-year implant survival was 98.5% with conventional surgery versus 98.6% with robotic surgery. The revision hazard ratio (HR) was 1.03 (95% confidence interval [CI] 0.86 to 1.26; P=0.71). No statistically significant difference in all-cause revision risk was detected.

For matched UKR, estimated five-year implant survival was 97.8% with conventional surgery versus 96.4% with robotic surgery. The revision HR was 1.03 (95% CI 0.75 to 1.42; P=0.86), with no statistically significant all-cause revision difference detected. Cause-specific revision and intraoperative complication risks did not differ significantly between approaches for either operation, nor did the proportion requiring two-stage revision. Weighting to assess a broader population and sensitivity analyses addressing missing body mass index (BMI) data and surgeons’ early robotic cases showed a similar pattern.

Matching measured characteristics cannot eliminate residual or unmeasured confounding in a retrospective comparison, including differences between centers, patient selection, surgeon preferences, robotic adoption, and care pathways. The projected five-year Kaplan–Meier estimates extend beyond the shorter mean observation period and do not settle longer-term implant survival. Fewer revision events constrained the unicompartmental and cause-specific comparisons; BMI missingness affected primary matching, robotic systems varied, and comparable patient-reported pain and function outcomes were unavailable. Investigators noted that using robotics for all knee replacements in the National Health Service (NHS) would add at least £100 million (roughly US$140 million) annually. A nonsignificant revision comparison does not establish equivalence.

Investigators found no statistically significant early difference in all-cause or cause-specific revision or intraoperative complications between robotic and conventional TKR or UKR. Observational confounding and limited follow-up leave longer-term implant survival unsettled.

Clinician Questions

What is a two-stage revision after robotic or conventional knee replacement?

A two-stage revision of a failed robotic or conventional knee replacement requires more than one operation to correct the failed implant; investigators used this measure to assess revision complexity.

Which robotic systems were represented in the UK knee replacement registry comparison?

Among UK patients undergoing robotic-assisted total or unicompartmental knee replacement for arthritis, Mako Total Knee, CORI Surgical System, and ROSA were among the systems represented; Mako was the most commonly used. The analysis pooled systems rather than estimating revision risk for individual platforms.

Recommended Reading

• For another study of total knee arthroplasty: Knee Arthroplasty Anesthesia: PACU Time, Waste, and Item Costs