179. Safety and accuracy of freehand pedicle screw placement and the role of intraoperative O-arm: a single institution experience

BACKGROUND CONTEXT Suboptimal placement of pedicle screws may lead to neurovascular complications, pseudarthrosis, postoperative pain and the need for revision surgery. PURPOSE The objective of this study was to investigatethe accuracy of pedicle screw placement by freehand technique and to compare revision surgery rates among three different imaging verification pathways. STUDY DESIGN/SETTING Retrospective observational study. PATIENT SAMPLE Adult patients who underwent freehand pedicle screw placement in the thoracic, lumbar or sacral levels in a single-institution between January 1, 2016, and June 30, 2020. OUTCOME MEASURES The proportion of malpositioned screws defined as grades C-E on the Gertzbein-Robbins grading system and the number of patients requiring an additional revision surgery for screw malposition. METHODS Patients were stratified into three cohorts based on the intraoperative imaging modality used to assess the accuracy of screw position: intraoperative x-rays (cohort 1); intraoperative O-arm (cohort 2); or intraoperative computed tomographic (CT) scan (cohort 3). Postoperative CT scans were performed on all patients in cohorts 1 and 2. Postoperative CT scan was not required in cohort 3. RESULTS A total of 9,179 pedicle screws were placed in the thoracic or lumbosacral spine in 1,311 patients. Of these, 210 (2.3%) screws were identified as Gertzbein-Robbins grades C-E on intra/postoperative CT scan; 137 thoracic screws and 73 lumbar screws (P< 0.001). A total of 409 patients underwent placement of 2,754 screws followed by intraoperative x-ray (cohort 1); 793 patients underwent placement of 5,587 screws followed by intraoperative O-arm (cohort 2); and 109 patients underwent placement of 838 screws followed by intraoperative CT scan (cohort 3). Postoperative CT scans identified 65 (2.4%) and 127 (2.3%) malpositioned screws in cohorts 1 and 2, respectively. Eleven screws (0.12%) were significantly malpositioned and required a second operation for screw revision. Nine patients (0.69%) required revision operations: 8 of these patients were from cohort 1 and one patient was from cohort 2. In cohort 3, 18 of 838 (2.1%) screws were identified as malpositioned by intraoperative CT-scan: 2 of these screws were revised intraoperatively during index procedure. No revision surgeries took place in this cohort. CONCLUSIONS Freehand pedicle screw placement is safe and effective with a 97.7% accuracy rate (Gertzbein-Robbins grades A or B). Nine patients (0.69%) required screw revision. When compared to intraoperative x-Ray, intraoperative o-arm verification decreased the revision surgery rate for malpositioned screws from 0.37% to 0.02%. In addition, our analysis suggests that the use of intraoperative O-arm can obviate the need for postoperative CT scans. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs. Suboptimal placement of pedicle screws may lead to neurovascular complications, pseudarthrosis, postoperative pain and the need for revision surgery. The objective of this study was to investigatethe accuracy of pedicle screw placement by freehand technique and to compare revision surgery rates among three different imaging verification pathways. Retrospective observational study. Adult patients who underwent freehand pedicle screw placement in the thoracic, lumbar or sacral levels in a single-institution between January 1, 2016, and June 30, 2020. The proportion of malpositioned screws defined as grades C-E on the Gertzbein-Robbins grading system and the number of patients requiring an additional revision surgery for screw malposition. Patients were stratified into three cohorts based on the intraoperative imaging modality used to assess the accuracy of screw position: intraoperative x-rays (cohort 1); intraoperative O-arm (cohort 2); or intraoperative computed tomographic (CT) scan (cohort 3). Postoperative CT scans were performed on all patients in cohorts 1 and 2. Postoperative CT scan was not required in cohort 3. A total of 9,179 pedicle screws were placed in the thoracic or lumbosacral spine in 1,311 patients. Of these, 210 (2.3%) screws were identified as Gertzbein-Robbins grades C-E on intra/postoperative CT scan; 137 thoracic screws and 73 lumbar screws (P< 0.001). A total of 409 patients underwent placement of 2,754 screws followed by intraoperative x-ray (cohort 1); 793 patients underwent placement of 5,587 screws followed by intraoperative O-arm (cohort 2); and 109 patients underwent placement of 838 screws followed by intraoperative CT scan (cohort 3). Postoperative CT scans identified 65 (2.4%) and 127 (2.3%) malpositioned screws in cohorts 1 and 2, respectively. Eleven screws (0.12%) were significantly malpositioned and required a second operation for screw revision. Nine patients (0.69%) required revision operations: 8 of these patients were from cohort 1 and one patient was from cohort 2. In cohort 3, 18 of 838 (2.1%) screws were identified as malpositioned by intraoperative CT-scan: 2 of these screws were revised intraoperatively during index procedure. No revision surgeries took place in this cohort. Freehand pedicle screw placement is safe and effective with a 97.7% accuracy rate (Gertzbein-Robbins grades A or B). Nine patients (0.69%) required screw revision. When compared to intraoperative x-Ray, intraoperative o-arm verification decreased the revision surgery rate for malpositioned screws from 0.37% to 0.02%. In addition, our analysis suggests that the use of intraoperative O-arm can obviate the need for postoperative CT scans.