Imaging assessment of lumbar pedicle screw placement: sensitivity and specificity of plain radiographs and computer axial tomography

Abstract

STUDY DESIGN: This is a diagnostic study in which cadavers were evaluated by imaging and dissection after pedicle screw placement. OBJECTIVE: This study was designed to evaluate the accuracy of plain radiographs and/or CT in determining pedicle instrumentation. SUMMARY OF BACKGROUND DATA: One biomechanical advantage provided by pedicle screws fixation is improved fusion rates. The accuracy of pedicle screw placement has relied on radiograph and CT. Our goal is to evaluate the accuracy of these methods. METHODS: Cadaveric lumbar spines were exposed by a posterior approach. Titanium pedicle screws were inserted. Radiographs and CTs were obtained. The spines were then dissected to examine screw location. A blinded examiner reviewed the imaging. Screws were considered misaligned if they were extrapedicular or had evidence of cortical perforation. RESULTS: The sensitivity of radiographs, CT, and combined is 93.9%, 94%, and 98.7%, respectively. The specificity of radiograph, CT, and combined is 12.5%, 36.7%, and 40.7%, respectively. The positive predictive value for radiograph, CT, and combined is 83.2%, 88.1%, and 90.4%, respectively. The negative predictive value for radiograph, CT, and combined is 30.8%, 55.9%, and 84.6%, respectively. The accuracy of radiograph, CT, and combined is equal to 79.4%, 84.4%, and 90%, respectively. CONCLUSION: Patients who present with significant complaints of pain or neurologic deficits, optimal determination of pedicle screw positions is with the combination of radiographs and CT. Furthermore, routine evaluation of patients' status post pedicle screw fixation can be reliable obtained with radiographs. The biomechanical advantage of pedicle screw fixation has improved fusion rates and diminished the number of levels required for stability. Despite the widespread clinical use of pedicle screw fixation, there is relatively little literature concerning accuracy. This study was designed to evaluate the accuracy of plain radiographs and/or CT in determining the proper position of pedicle screws.