Abstract: OBJECTIVE: Total knee arthroplasty has been a common treatment for single-compartment osteoarthritis for some time. Less invasive unicompartmental knee arthroplasty has not been widely available due to low survival rates, but with the development of robot-assisted unicompartmental knee arthroplasty, this goal has become more promising. This study observed the postoperative effects and differences between robot-assisted unicompartmental knee arthroplasty and traditional unicompartmental knee arthroplasty in the treatment of osteoarthritis by means of meta-analysis.
METHODS: Relevant databases including Wanfang, VIP, CNKI, EMBASE, PubMed, Cochrane Library, and Web of Science were searched by computer in English and Chinese, and the time was from database inception to June 2020. Quality of retrieved articles was assessed. The outcome indicators included operation time, postoperative complications, implant accuracy, posterior slope and revision rate. Meta-analysis of the collected data was carried out using the RevMan 5.3 software provided by the Cochrane Collaboration. 
RESULTS: (1) A total of 11 articles were included. There were three randomized controlled trials, one low risk bias, one medium risk bias, and one high risk bias. There were four cohort studies and four case-control studies with NOS score above 6, which were of high quality. (2) Meta-analysis results showed that robot-assisted unicompartmental knee arthroplasty was superior to traditional surgery in the accuracy of the tibial component placement (MD=-1.32, 95%CI: -1.74 to -0.91, P < 0.000 01), the accuracy of the femoral component placement (MD=-1.71, 95%CI: -2.82 to -0.61, P=0.002), the postoperative tibial posterior slope (MD=-2.40, 95%CI: -2.80 to -2.22, P < 0.000 01), and the revision rate (RR=0.34, 95%CI:0.16-0.72, P=0.005). (3) Although robot-assisted unicompartmental knee arthroplasty took longer time (MD=11.53, 95%CI:5.12-17.94, P=0.000 4), it did not increase the incidence of surgical complications (RR=0.86, 95%CI:0.45-1.64, P=0.65). 
CONCLUSION: When performing unicompartmental knee arthroplasty, robot-assisted unicompartmental knee arthroplasty has higher prosthesis positioning accuracy, better postoperative tibial inclination and lower revision rate without increasing the incidence of complications. It is a surgical method that is more ideal than traditional surgical methods, and is worthy of further discussion in clinical practice, but the above conclusions still need to be verified by large-sample and high-quality randomized controlled studies.

Key words: knee, joint, robot, joint arthroplasty, unicompartmental arthroplasty, arthritis, complications, prosthesis, meta-analysis