Variability in Treatment for Patients with Cervical Spine Fracture and Dislocation: An Analysis of 107,152 Patients.

Abstract

Cervical spine injuries are a common cause of morbidity and mortality; however, the optimal treatment of many of these injuries is debated, and previous studies have shown substantial variation in treatment. We sought to examined treatment variation in arthrodesis and halo/tong placement in cervical spine injury patients over a 12-year period. Data from the Healthcare Cost and Utilization Project National Inpatient Sample, from 2000 to 2011, were used for this study. Patients were identified with a cervical vertebral facture or dislocation based on the International Classification of Diseases, 9th Revision codes. Using χ2 analysis, spinal arthrodesis rates and halo/tong placement rates were compared between hospitals based on teaching status for patients with and without spinal cord injury (SCI). The records of 107,152 patients with cervical fractures were examined. From 2000 to 2011, the overall arthrodesis rates fell from 25.2% to 20.6% (P < 0.001), and halo/tong placement rates fell from 13.2% to 3.6% (P<0.001). In patients with cervical fracture without SCI, arthrodesis rates fell from 17.6% to 13.9% (P < 0.001), in cervical fracture patients with SCI, arthrodesis rates rose from 50.0% to 58.9% (P < 0.001), and in cervical dislocation patients, arthrodesis rates rose from 47.6% to 57.5% (P<0.001). During the 12-year period, teaching hospitals had higher arthrodesis rates compared with nonteaching hospitals for patients with cervical fractures with SCI (57.3% vs. 53.4%, P= 0.001) and higher halo/tong placement rates for patients with cervical dislocations (2.7% vs. 1.7%, P= 0.004). Individual hospital variation showed a 3.5-fold variation in arthrodesis rates in 2000 to 2002, which fell to 3.0-fold by 2009 to2011. Arthrodesis rates for cervical fracture patients significantly decreased, and arthrodesis rates for cervical dislocation and SCI patients increased from 2000 to 2011, with variability in treatment based on hospital teaching status. Rates of halo/tong placement rapidly decreased for cervical spine trauma at both teaching and nonteaching hospitals. Individual hospital treatment variation also decreased over the study period. Further clinical studies examining the optimal treatment for spine trauma may lead to continued decreases in treatment variability.