The accuracy of screw placement related to robotic assistance has been reported. Notably, patient-related factors were the major contributors to successful and accurate screw placement. We did achieve a higher level of clinical satisfaction with robotic-assisted screwing, 98.2% clinically acceptable screwing in this study, compared to the accuracy of conventional freehand screwing as previously reported.4,15,17−19 Although clinically acceptable screwing included graded A and B, intraoperative planning was based on the most accurate and perfect position for simulated screwing (grade A), while only 84.8% of the screws were actually placed in the same path as planned. For example, risk factors for suboptimal robotic-assisted pedicle screw placement included osteoporosis, obesity, and congenital scoliosis.20 Similarly, osteoporosis and obesity as patient-related factors affected the accuracy of screw placement.21 Obese patients have thicker subcutaneous soft tissues and higher tone of back muscles. After the placement of trocar is completed, the stronger tissues may squeeze the sleeve, resulting in a bias that cannot be detected by the naked eye, which eventually leads to displacement of a screw. This characteristic may also play a role in the age and gender factors. Possibly, osteoporosis-induced displacement may be related to the inability of a screw to fit firmly to the surrounding bone.20,22 In practice, even in the presence of systemic factors such as obesity and osteoporosis, it is always possible to have both screws that are identical to those planned (grade A) and screws that are off course (grades B, C, D, and E) in the same patient's screwing procedure.
There was no significant difference in accuracy between robot-assisted pedicle screwing and CBT screwing. The CBT screw has a smaller length and width than the pedicle screw, but has better holding power due to its trajectory characteristics. Previous literature reported the same clinical results in fixation with CBT screws compared to pedicle screwing.23 In a study on transforaminal lumbar fusion, similar fusion rates and accuracy were obtained with CBT and pedicle screws.5 In our opinion, CBT screws have little soft tissue between the sleeve and the bone of vertebral plate due to midline exposure, resulting in less soft tissue traction and obstruction that can interfere with screw accuracy. In contrast, pedicle screws with midline exposure may require more soft tissue traction, which can increase screw deflection. However, there was a lack of descriptive analysis of the accuracy of robotic-assisted CBT versus pedicle screwing in a large sample size. In our study, there was no statistically significant difference between the two groups in terms of accuracy.
Sliding of the guide wire and the screw at the planning entrance may be an important reason for the accuracy of the screw. Ringel et al. and Ghasem et al. also found that most robot-assisted screw displacement problems were associated with slippage of the implanted cannula on the bone surface at the screw entrance. 24,25In our study, the plane in which the entry point is located is presented in a three-dimensional space, and this oblique plane is characterized by the angle of the vertebral plate in axial and sagittal positions. The greater the angle between the axial and sagittal positions, the more likely it is that the screw will be displaced. The large axial angle is an independent risk factor for screw inaccuracy. This slope effect may be more pronounced in the presence of spinal deformities, such as the presence of a rotating vertebral body, and the effect at the main curve in scoliosis patients. At the same time, due to the different morphological characteristics of the thoracic and lumbar vertebrae, the direction of the sliding offset during the screwing process also has its own characteristics. The size of the screw can also affect the accuracy of the screw placement. In particular, the diameter is an independent risk factor for screw deviation from the planned path. In the case of a certain width of the pedicle, the larger the diameter of the screw, the more likely it is to invade the cortex. Even if the depth of screw insertion does not match the planned depth, it is an acceptable result if the anterior vertebral cortex or important structures such as blood vessels and nerves are not damaged. (Fig. 6) We compared the differences in screw placement between the left and the right side. The right side position of a screw was an independent risk factor for unsatisfactory screw placement. In fact, during screw placement procedure, the patient was lying in a prone position on the operating table, and the surgeon usually stood on the left side of the patient to place the screw, while the technician adjusted the robotic arm on the right side of the patient. Although the range of motion of robotic arm is greater, there are still physical limitations for the operator in placing screws on the patient's right side.
The distance between the tracer and the screw placement segment correlates with the accuracy of the screw, the further the distance the less accurate the placement. As previously reported, the accuracy of pedicle screw placement depended on the distance between the tracer and the screw.26 Tian et al. and Jin et al. also found in their study that the distance between the screw and the tracer more than 3 segments was a risk factor for misalignment of the navigated pedicle screw.21,27 Although general anesthesia surgery is predominantly chest breathing, different tidal volumes can lead to different degrees of up and down movement of the entire torso, away from the center. Our results indicate the greater the movement, the lower the screw accuracy. Even though the anesthesiologist will adjust the tidal volume and respiratory rate to match the screw placement during surgery, this deviation cannot be avoided. As a key step in robotic operation, the impact of the tracer on surgical process can range from improper placement, which may result in incomplete exposure of the surgical area to the operating table, to inadvertent intraoperative touching, which may cause structural deflection of the entire surgical area. The position of the tracer is highly dependent on the operator's experience, and there is no definitive conclusion on the superiority vs. inferiority of the tracer position. Part of the reason for the long time to use robotic assisted surgery in the early days was due to improper installation of the tracer. Therefore, we recommend that the tracer should be positioned as close to the surgical segment as possible to ensure accurate screw placement during long-segment procedures.19,28 We can even take 2 or more tracer placements to ensure the accuracy of long segment robotic surgery
This study mainly evaluates the accuracy of robot-assisted screw placement technology based on the vertebral body’s characteristics. There are several limitations. First of all, this is a single-center retrospective study, and the sample size is relatively small. Thus, a larger sample size including patients from multiple centers is required to validate our conclusion in the current study. Second, our results for screw deviation were analyzed based on CT findings after postoperative removal of drainage tube for approximately 10 days, which lacked long-term accuracy. Thus, follow-up investigations at 3 months, 6 months, or even years are needed to demonstrate the causes of screw deviation from the planned route. For the next step, we will explore short- and long-term consequences related to robotic surgery-assisted screws by combining patient-related factors, surgical content, and anatomical features.