Efficiency of OR-based sliding gantry CT
The present study’s analysis shows that the ORCT-based navigated placement of pedicle screws is more efficient than the CBCT-based with regard to duration of OR use and thereby resource implications (Table 3 and Table 4). The analysis of this very homogenous cohort as shown by baseline characteristics and surgical details (Table 1 and 2) proves the effect of an optimized OR equipped with a permanently installed ioCT with sliding gantry on standard spine surgery procedures in a high-volume tertiary care center. The significant difference of the patients’ presence in the OR (Interval 5 Entry—Exit) was significantly shorter for the ORCT group mainly base on the significantly shorter duration needed for preparing the positioned patient for the ioCT-based imaging and to a lesser extent on shorter times needed for the de-positioning of the patient (Fig. 3). As the detailed illustration for all intervals and the analysis of durations per level shows, durations of the CBCT group are more heterogeneously distributed in comparison with the ORCT group (Fig. 4 and Table 4). With regard to the validity of the present data, surgical procedures as recorded by Interval 3 Incision—Suture did not affect these differences. Of course, the time needed for imaging-based preparations in the CBCT group is included in Interval 2 Positioning—Incision, while preparations for ioCT imaging in the ORCT group count among Interval 3 Incision—Suture due to performing CT scout images for scan and radiation dose planning. The still missing difference for Interval 3 Incision—Suture should preliminary be reasoned by the higher number of scans which had to be performed for a higher number of levels in CBCT cases as compared to ORCT cases which can be done with one scan prior and one scan after all screws were placed. Additionally, it might also be reasoned by the higher number of levels decompressed in patients of the CBCT group. However, the number of cases which underwent decompression did not differ between groups, hence, the impact of this time-consuming intraoperative procedure included in Interval 3 Incision—Suture has less impact on the overall results (Table 2 and Table 4). Furthermore, the familiarity of handling the devices and particularly the preparation as well as de-positioning should have affected Interval 2 Positioning—Incision and Interval 4 Suture—Exit. However, we included patients to the two groups from the beginning of their availability at our department, respectively. Hence, the learning curves for the handling and preparation are included for both devices.
While the results could also be shown after the separation into instrumentations of cervicothoracic or thoracolumbar levels, we additionally found shorter durations for Interval 1 Entry—Positioning and Interval 3 Incision—Suture for the cervicothoracic subgroup. Furthermore, the analysis per level did not show significant differences (Table 4 and Fig. 4). This might be reasoned by the higher percentage of cervicothoracic cases in the ORCT group (Table 1).
At our department, we standardly perform at least two intraoperative scans for navigation and control of pedicle screws. This offers the possibility for an immediate intraoperative revision of screws based on a navigational intraoperative scan. It has to be mentioned that centers exist which only perform a navigation scan without a standard control scan. Of course, this is reasonable from a practical point of view. However, regarding the radiation exposure of patients and the surgical team, an earlier randomized controlled trial could show that the cumulative radiation dose of the procedure as performed in the present study is acceptable and even lower as compared to 2D-fluoroscopy-guided free-hand instrumentations .
Clinically, screws had to be revised less frequently in the ORCT group as compared to the CBCT group, while the accuracy of screw placement according to Gertzbein–Robbins classification showed no significant differences between the two groups (Table 3). Based on the results of the present study and personal experience, higher revision rates in the CBCT group are reasoned by an inferior visualization as compared to an ORCT-based spinal instrumentation (Fig. 1 and 2). Moreover, the revision of misplaced screws seems to be faster by the help of CBCT decreasing the threshold for revision in uncertain cases. Due to the retrospective character of the present study, this assumption cannot be supported by hard numbers though.
Apart from these comparisons of ORCT and CBCT focusing the durations of perioperative intervals, imaging quality, and accuracy of instrumentation, the two devices still have advantages and disadvantages. The CBCT device enables to perform the intraoperative imaging without dependence on radiologists and technical assistants as it is necessary for the ORCT device. This fact might be time-consuming and error-prone in case of not well-attuned teams. Otherwise, the ORCT device offers the opportunity to image multiple levels as defined by the surgeon. In contrast, the imaging of the CBCT device is limited to several vertebrae. This might be negligible in case of standard instrumentations including up to 5 levels but is a factor for longer constructs.
ioCT-based pedicle screw placement
Former publications have already shown a significantly lower risk of pedicle screw malpositioning as well as a higher accuracy and safety for the intraoperative CT (ioCT)-navigated insertion than for free-hand technique or 2D fluoroscopy [6, 8,9,10,11,12]. Several studies have investigated the use of mobile ioCT scanners [1, 19,20,21,22] In contrast to publications on CBCT-based pedicle screw placement, reports on the use of permanently installed ORCTs are scarce [7, 14,15,16]. While the accuracy of pedicle screw placement and revision rates of the present study are comparable to earlier publications on navigated pedicle screws, the differences of durations are difficult to compare since measurements, subgroups, and the complexity of surgeries vary across publications. With a special regard to instrumentation of the cervical spine and cervicothoracic junction, the present study’s results show one of the largest cohorts with CT-navigated cervical pedicle screws [23,24,25,26]. The present subanalysis showed less differences for the comparison of the ORCT and CBCT group. Apart from that we could prove the results of an earlier large cohort study showing now the safety of CT-navigated instrumentation of this region . In the present cohort, patients in the ORCT group underwent more often cervical pedicle screw placement (18.5% vs. 7.0%; p < 0.0001) than patients in the CBCT group due to the possibility of region matching (Brainlab AG, Munich, Germany; Table 1).
Our analysis also revealed that intraoperative screw evaluation does not take the Gertzbein–Robbins classification into account when it comes to the decision of intraoperative screw revision. That is because this decision requires other parameters such as haptic bone quality, screw firmness and anatomical difficulties during placement into direct account. As a result, the Gertzbein–Robbins classification was not used to describe the results since it was not used for intraoperative decision making at all. Important to note that there was no case in this consecutive series which had to be taken back to the OR due to required replacement of any pedicle screws.
The retrospective character of the present analysis might be a major limitation of our study. Hence, the analysis had to be limited to the presented parameters. For example, the learning curve for the two devices was not considered in the analysis. However, these data and especially the main results are based on digitally documented reports and are thereby highly reliable.