The accuracy and learning curve of pedicle screw placement by using O-arm navigation in thoracic spine

English Keywords:Thoracic spine pedicle screw  O-arm intraoperative navigation  Learning curve

English Abstract:

  【Abstract】 Objectives: This study aimed to evaluate the accuracy andlearning curveof pedicle screw placement by using O-arm navigation versus C-arm guidance in thoracic spine. Methods: Retrospective analysis was performed on 109 cases of thoracic spine surgery performed by Peking University International Hospital from May 2015 to May 2018, including 49 males and 60 females. The age ranged from 27 to 77 years old, with an average age of 53.5±12.3 years. These cases were divided into navigation group(group A, n=66) and non-navigation group(group B, n=43) according to whether used of O-arm navigation. 35 cases of group A were completed by doctor 1 who had 3 years of experience in spinal surgery (group A1); 31 of group A were completed by doctor 2 who had 20 years of experience (group A2). On the other hand, 18 of group B were completed by doctor 1 (group B1) and 25 of group B were completed by doctor 2 (group B2). Postoperative pedicle screw position was graded according to Neo′s method, and satisfactory screw placement(grade 0 +1 screws) and screw penetration into the cortex (non-grade 0 screws) in each group were compared. The cortical perforation rate at different segment in the same group were compared. The screwsetting time of each group were recorded. Linear regression analysis was used to investigate the relationship between the screw setting time and operation numbers, and compare the difference of screw settingtime and cortical perforation rate in different stages, thus, to analyze the learning curves of O-arm navigation in thoracic spine. Results: A total of 668 thoracic pedicle screws were inserted in the 4 groups, of which 222 were inserted in group A1, with a satisfactory rate of 97.3% and a cortical perforation rate of 10.8%. In group A2, 188 screws were inserted, the satisfactory rate was 97.9%, and the cortical perforation rate was 10.1%. In group B1, 120 screws were inserted, with a satisfactory rate of 91.7%, and cortical perforation rate 20.0%. In group B2, 138 screws were inserted, with a satisfactory rate of 97.8%, and cortical perforation rate of 9.4%. The satisfactory rate of group B1 was significantly lower than the other groups, and the cortical perforation rate was also significantly higher than the other groups. Besides, there is significant difference of cortical perforation rates in different segments in thoracic pedicle in group B1(37.5%, 28.6%, 13.4%, P=0.02), and there is no significant difference in other 3 groups(A1/A2/B2). The setting time for each screw of group B1(4.87±0.34min) was significantly higher than the other respectively (4.38±0.97min, 4.40±1.00min, 4.18±0.22min). The setting time for each screw in the navigation group (A1, A2) decreased significantly with the increase of the number of surgical cases. Linear regression analysis showed that there was a significant negative linear correlation between the setting time and the number of surgical cases in group A1(F=123.3, P＜0.001), the setting time = 5.88-0.084×number of cases, R2=0.79). There was a same result in group A2(F=141.6, P＜0.001，the setting time = 6.01-0.10×number of cases, R2=0.83). Cortical perforationrates in the first 10, 11-20, and 21-30 cases of A1 group were not statistically significant(P=0.97), and the same results were found in the A2 group(P=0.96). Conclusions: The O-arm navigation assisted thoracicpedicle screw can significantly improve the accuracy of pedicle screw in young surgeon, which increased the safety of the operation. The screw setting time decreased significantly with the increase of the number of surgical cases,andearly application of O-arm by surgeonsat different seniority showed a high accuracy of screw placement.