Learning curve of robot-assisted L 4 and L 5 pedicle screw insertion for lumbar spondylolisthesis
10.3760/cma.j.cn121113-20211103-00635
- VernacularTitle:机器人辅助L 4、L 5椎弓根螺钉内固定术治疗腰椎滑脱症的学习曲线研究
- Author:
Mingxing FAN
1
;
Yanming FANG
;
Yan AN
;
Qi ZHANG
;
Wei TIAN
Author Information
1. 北京积水潭医院脊柱外科，北京 100035
- Keywords:
Lumbar vertebrae;
Spondylolysis;
Robotic surgical procedures;
Learning curve
- From:
Chinese Journal of Orthopaedics
2022;42(6):341-348
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To study the learning curve and inflection point of robot-assisted L 4 and L 5 pedicle screw insertion for lumbar spondylolisthesis. Methods:A retrospective study was conducted on the data of 43 patients with L 4 and L 5 pedicle screw insertion for lumbar spondylolisthesis from January 2016 to December 2020 using surgical robot, including 19 males and 24 females, aging 59 (48, 66) years old. According to Meyerding classification, there were 23 grade I slippage, 18 grade II slippage, and 2 grade III slippage. The screw deviation and screw accuracy grade were assessed. The operation time, intraoperative blood loss, the number of intraoperative fluoroscopies, postoperative complications, and postoperative hospital stay were recorded. Cumulative Sum (CUSUM) was used to analyze the learning curve, and the learning curve is divided into early and late learning stages according to morphology. T test and Wilcoxon rank-sum test were used for statistical analysis and comparison of indicators between early and late learning stages. Results:43 patients with lumbar spondylolisthesis successfully completed the operation, with 60 L 4 pedicle screws and 70 L 5 pedicle screws inserted. The accuracy of L 4 pedicle screw placement began to improve since the 23rd placement, and the accuracy of L 5 pedicle screw placement began to improve since the 20th placement. Using the 23rd pedicle screw (the 14th patient) to divide the learning curve as the early stage and the late stage. There was no statistically significant difference in the operation time (225.0±74.0 min vs. 207.0±81.2 min, t=0.65, P=0.521), intraoperative blood loss[200 (75, 500) ml vs. 100 (60, 200) ml, Z=1.30, P=0.195], the number of intraoperative fluoroscopies[10 (6, 10) vs. 10 (6, 10), Z=-0.37, P=0.712] and postoperative complications (8% vs. 0, P=0.302) between the early stage and late stage of learning curve. In the late stage of learning, the postoperative hospital stay was shorter[4.5 (3, 6) d vs. 6.0 (5, 9) d, Z=2.00, P=0.046]and the pedicle screw insertion accuracy was higher[L 4: 1.33 (1.06, 1.79) mm vs. 2.23 (1.12, 4.55) mm, Z=2.43, P=0.015; L 5: 1.47 (0.98, 1.87) mm vs. 3.21 (1.64, 4.87) mm, Z=3.90, P=0.001]. The accuracy of screw placement was similar between the early and late stages[L 4: 95%(21/22) vs. 97%(37/38), P=1.000; L 5: 91%(20/22) vs. 96%(46/48), P=0.585]. Conclusion:Robot-assisted L 4 and L 5 pedicle screw placement in the treatment of lumbar spondylolisthesis had a relatively obvious learning curve. Starting from the placement of the 23rd screws, the accuracy of screw placement gradually increased.
