Abstract: Objective To systematically evaluate the accuracy of dental implant surgery assisted by a dental implant robot system in both preclinical and clinical studies. Methods Studies of dental implant robots were searched on online platforms like PubMed, Cochrane Library, Web of Science, Embase, CNKI, CQVIP, and Wanfang data from the establishment of the database to August 2023. The studies included preclinical (model and animal experiments) and clinical research. The main outcome measures were three-dimensional deviations between the targeted position and the implanted position, including the coronal deviation, apical deviation, and angular deviation of implants. Stata 14.0 and RevMan software were used for meta-analysis of the data. Subgroup analysis was used to compare errors in four aspects: different study designs and subjects, preparation of implant cavities versus insertion of implants, dentition condition, and specific anatomical structures. Results Finally, 23 studies (18 preclinical studies and 5 clinical studies) from 20 papers were included, comprising a total of 840 implant sites. The results of the meta-analysis showed that in the preclinical study of dental implant robot-assisted implantation, the coronal deviation, apical deviation, and angular deviation were 0.60 mm (95% CI: 0.51~0.70, I²=0.0%), 0.73 mm (95% CI: 0.62~0.84, I²=24.9%), and 1.25°(95% CI: 1.02~1.48, I² = 47.9%), respectively. In clinical studies, the coronal deviation, apical deviation, and angular deviation were 0.62 mm (95% CI: 0.39~0.86, I²=0.0%), 0.63 mm (95% CI: 0.39~0.87, I²=0%) and 1.23°(95% CI: 0.71~1.76, I²=15.9%), respectively. Conclusion There was no statistical difference in the coronal deviation, apical deviation, and angular deviation of dental implant surgery assisted by the dental implant robot system in preclinical and clinical studies, providing a reference for subsequent experiments and clinical applications.

Key words: Dental implant robot, Digital, Computer-assisted operation, Accuracy