Defining the cervical spine clearance algorithm: A single-institution prospective study of more than 9,000 patients.

Abstract

Variability exists in the approach to cervical spine (c-spine) clearance after significant trauma. Using concurrently gathered data on more than 9,000 such patients, the current study develops an evidence-based and readily adoptable algorithm for c-spine clearance aimed at timely removal of collar, optimal use of imaging, and appropriate spine consultations. Prospective study of adult blunt trauma team alert (TTA) patients presenting at a Level I trauma center who underwent screening computed tomography (CT) to diagnose/rule out c-spine injury (January 2008 to May 2014). Regression analysis comparing patients with and without c-spine injury-fracture and/or ligament-was used to identify significant predictors of injury. The predictors with the highest odds ratio were used to develop the algorithm. Among 9,227 patients meeting inclusion criteria, c-spine injury was identified in 553 patients (5.99%). All 553 patients had a c-spine fracture, and of these, 57 patients (0.6% of entire population and 10.31% of patients with injury) also had a ligamentous injury. No patient with a normal CT result was found to have an injury. The five greatest predictors of ligament injury that follow were used to develop the algorithm: (1) CT evidence of ligament injury; (2) fracture pattern "not" isolated transverse/spinous process; (3) neurologic symptoms; (4) midline tenderness; and (5) Glasgow Coma Scale score <15. TTA patients should undergo screening c-spine CT to rule out injury. Most patients will have a negative CT and can have their collars removed. A select group of patients will require collars and spine consultation and a smaller subset of magnetic resonance imaging to rule out ligament injury. Therapeutic study, level III.