Comparison of accuracy, revision, and perioperative outcomes in robot-assisted spine surgeries: systematic review and meta-analysis
Luke MacLean,Andrew M Hersh,Meghana Bhimreddy,Kelly Jiang,A Daniel Davidar,Carly Weber-Levine,Safwan Alomari,Brendan F Judy,Daniel Lubelski,Nicholas Theodore
DOI: https://doi.org/10.3171/2024.4.SPINE23917
2024-07-05
Abstract:Objective: Pedicle screw placement guidance is critical in spinal fusions, and spinal surgery robots aim to improve accuracy and reduce complications. Current literature has yet to compare the relative merits of available robotic systems. In this review, the authors aimed to 1) assess the current state of spinal robotics literature; 2) conduct a meta-analysis of robotic performance based on accuracy, speed, and safety; and 3) offer recommendations for robotic system selection. Methods: Following PRISMA guidelines, the authors conducted a systematic literature review across PubMed, Embase, Cochrane Library, Web of Science, and Scopus as of April 28, 2022, for studies on approved robots for placing lumbar pedicle screws. Three reviewers screened and extracted data relating to the study characteristics, accuracy rate, intraoperative revisions, and reoperations. Secondary performance metrics included operative time, blood loss, and radiation exposure. The authors statistically compared the performance of the robots using a random-effects model to account for variation within and between the studies. Each robot was also compared with performance benchmarks of traditional techniques including freehand, fluoroscopic, and CT-navigated insertion. Finally, we performed a Duval and Tweedie trim-and-fill test to assess for the presence of publication bias. Results: The authors identified 46 studies, describing 4670 patients and 25,054 screws, that evaluated 4 different robotic systems: Mazor X, ROSA, ExcelsiusGPS, and Cirq. The weighted accuracy rates of Gertzbein-Robbins classification grade A or B screws were as follows: ExcelsiusGPS, 98.0%; ROSA, 98.0%; Mazor, 98.2%; and Cirq, 94.2%. No robot was significantly more accurate than the others. However, the accuracy of the ExcelsiusGPS was significantly higher than that of traditional methods, and the accuracies of the Mazor and ROSA were significantly higher than that of fluoroscopy. The intraoperative revision rates were Cirq, 0.55%; ROSA, 0.91%; Mazor, 0.98%; and ExcelsiusGPS, 1.08%. The reoperation rates were Cirq, 0.28%; ExcelsiusGPS, 0.32%; and Mazor, 0.76% (no reoperations were reported for ROSA). Operative times were similar for all robots. Both the ExcelsiusGPS and Mazor were associated with significantly less blood loss than the ROSA. The Cirq had the lowest radiation exposure. Robots tended to be more accurate and generally their use was associated with fewer reoperations and less blood loss than freehand, fluoroscopic, or CT-navigated techniques. Conclusions: Robotic platforms perform comparably based on key metrics, with high accuracy rates and low intraoperative revision and reoperation rates. The spinal robotics publication rate will continue to accelerate, and choosing a robot will depend on the context of the practice.