STUDY DESIGN: Prospective clinical trial. OBJECTIVE: The goal of this prospective study was to assess the reliability of plain radiographs, computed tomography (CT), and magnetic resonance imaging (MRI) to detect a pseudarthrosis after an anterior cervical fusion compared with intraoperative exploration. SUMMARY OF BACKGROUND DATA: Although anterior cervical fusions have been performed for >50 years, diagnosis of pseudarthrosis remains controversial, and even though various radiologic methods have been described to detect a pseudarthosis, no reports have compared these studies against the gold standard of intraoperative exploration to determine which is most accurate. METHODS: All explorations were performed >6 months after the index procedure and studies were obtained within a month of surgery. Indications for reoperation were adjacent level disease and/or pseudarthrosis repair. All patients were thoroughly explored using a microscope, uncinate-to-uncinate exploration, and if necessary, removal of part of the corticalized graft. Radiologic studies were blindly and independently reviewed by 3 spine surgeons uninvolved in the care of the patients. RESULTS: Of the 14 patients included, 8 had pseudarthroses. Assessment of the agreement between intraoperative and radiographic findings revealed a mean Kappa statistic of 0.67 (range, 0.51-0.75, P < 0.05) for plain radiographs, 0.81 (range, 0.71-0.87, P < 0.05) for CT, and 0.48 (range, 0.32-0.71, P < 0.05) for MRI. When all studies were taken into consideration, the mean Kappa statistic increased to 0.85 (range, 0.71-1.00, P < 0.05). Assessment of paired interobserver reliability revealed a mean Kappa statistic of 0.46 (range, 0.31-0.55, P < 0.05) for plain radiographs, 0.82 (range, 0.73-0.87, P < 0.05) for CT, and 0.32 (range, 0.21-0.40, P < 0.05) for MRI. When all radiographic studies were taken into consideration, paired interobserver reliability had a mean Kappa statistic of 0.70 (range, 0.55-0.85, P < 0.05). CONCLUSION: Our results indicate that CT most closely agrees with intraoperative findings. We therefore recommend that CT be used in any study that relies on fusion status assessment. It is important to remember, however, that even CT is not 100% accurate.
