Accuracy and potential risk factors of pedicle screw placement using a noninvasive-registration robotic system in the thoracolumbar spine.

To evaluate the accuracy of screw placement using the ORTHBOT robotic system in degenerative spine surgeries at our institution and identifying independent risk factors influencing screw deviation. The clinical data of 75 patients who underwent Robotic-Assisted Spine Surgery (RASS) at our hospital from May 1, 2022 to June 31, 2024. Concurrently, clinical data from 75 robot-assisted spine surgery cases in our treatment group were collected, including patients' general information (age, gender, duration of disease, and length of hospital stay). Recorded parameters included the number of screws, operative time, BMI, bone density, degree of intervertebral disc degeneration (Pfirrmann grading), facet joint degeneration (Fuiiwara-MRI grading), vertebral rotation (Nash-Moe method), and screw grading (Gertzbein and Robbins scale). A total of 399 screws were evaluated. Statistical analysis was conducted using GraphPad software. Quantitative data were described according to their distribution, and categorical data were expressed as percentages. Initially, univariate analysis was performed, and variables with statistical significance were then incorporated into a multivariate logistic regression model to further assess the independent effects of each variable on screw placement accuracy. Screw placement accuracy was categorized as satisfactory or unsatisfactory, and the odds ratios (OR) with 95% confidence intervals (CI) for each influencing factor were calculated, with a P < 0.01 considered statistically significant. A total of 75 patients who underwent RASS surgery (all receiving percutaneous screw implantation) were included, comprising 40 males and 35 females, aged 25-84 years, with an average age of 63.81 ± 11.81 years. The patients had an average BMI of 25.74 ± 3.63, an average bone density T-score of - 0.19 ± 1.95, an average hospital stay of 13.15 ± 5.25 days, and an average disease duration of 50.88 ± 72.49 months. Among all patients, past medical histories included hypertension and diabetes, with 30 patients having hypertension and 7 having diabetes. All RASS procedures were performed under general anesthesia. A total of 329 Grade A screws (82.46%) and 32 Grade B screws (8.02%) were placed, with Grades A and B combined constituting satisfactory screws, totaling 361 (90.48%). Additionally, there were 26 Grade C screws (6.52%), 10 Grade D screws (2.51%), and 2 Grade E screws (0.50%), totaling 38 (9.52%), which were classified as unsatisfactory screws (Grades C, D, and E). Univariate analysis indicated that the risk of inaccurate screw placement was significantly higher in elderly patients (≥ 65 years) compared to the < 65 group, which is closely related to the common occurrence of osteoporosis, vertebral morphological variations, and unclear bony landmarks in the elderly. The risk was increased in patients with severe osteoporosis (T-score≤-3.5), suggesting that excessively low bone density reduces screw purchase, leading to screw trajectory deviation. BMI, disease duration, and operative time did not have a significant impact on the risk of unsatisfactory screw placement. Increased spinal rotation raised the risk (P = 0.009), reflecting the three-dimensional reconstruction challenge of pedicle spatial orientation posed by rotational deformities. An increased grade of facet joint degeneration had a significant impact (P < 0.05), possibly due to facet joint hypertrophy and sclerosis causing deviation of the K-wire from the planned trajectory. Multivariate logistic regression analysis indicated that severe vertebral rotation (Grade III-IV) (P < 0.01) significantly affected screw accuracy, as spinal rotation limited the range of motion of the robotic arm. In osteoporotic patients, decreased bone mineral density (BMD) significantly increased the risk of failure, with an odds ratio (OR) of 1.832 [95% CI: 1.212-2.741] when-3.5 < T≤-2.5, and an OR of 3.502 [95%CI: 1.923-6.384] when T≤-3.5. This may be related to lower screw purchase and screw displacement during decompression surgery. A facet joint degeneration grade higher than 2 also increased the risk of screw displacement (P < 0.01), as more severe degeneration, with increased osteophyte formation and higher cortical bone density, may increase the difficulty of K-wire insertion, leading to displacement on the facet joint surface. Age did not have a significant impact on screw accuracy (P = 0.028). This study conducted a retrospective analysis of clinical data from 75 patients undergoing robot-assisted spinal surgery (RASS), confirming the high precision of the domestically developed ORTHBOT system in pedicle screw placement, with a satisfactory screw (Grade A/B) rate of 90.48%. These findings provide reliable evidence for the clinical application of RASS technology in degenerative spinal diseases. Univariate analysis identified age, spinal rotation grade, bone mineral density (BMD), and facet joint degeneration grade as risk factors leading to reduced screw placement accuracy, while BMI, disease duration, and operative time showed no significant influence. Multivariate regression analysis further determined that bone density loss (T-score≤-2.5), grade III-IV spinal rotation, and grade 2-3 facet joint degeneration were independent risk factors for screw deviation.