Background: The overall incidence of cervical spine injury (CSI) has been estimated from small studies; the incidence of specific injury types is less well established. The approach to screening for CSI has not been well studied; variation may exist based on Trauma Center (TC) level and type (academic vs. nonacademic). We attempted to define the incidence of different types of CSI and determine whether a national standard for cervical spine clearance (CSC) could be identified. We hypothesized a significant variation in incidence of CSI and approach to CSC based on TC level and type.
Methods: In a survey of 615 TC, institutions were asked to describe themselves as academic/nonacademic and provide a Level I-IV. Questions concerned demographics, Injury Severity Score, incidence of CSI, clinical resources, and approach to CSC. Methods of CSC included protocols, use of flexion-extension films, computed tomography, magnetic resonance imaging, and cervical collars. Clinical scenarios examined indications and technique for CSC.
Results: A total of 637 surveys were sent to 615 TC (25 follow-ups), and 165 TC (25%) responded. A total of 156 TC provided data for type: academic 44 (28%), nonacademic 112 (72%). A total of 142 TC provided data for level: 49 (34%) Level I, 75 (53%), Level II, 18 (13%), Level III. A total of 111,219 patients were entered into the trauma registries of these TC. The overall incidence of all types of CSI was 4.3%, CSI without spinal cord injury was 3.0%, spinal cord injury without fracture was 0.70%, and delayed diagnosis of all types of CSI was 0.01%. There was no difference in the incidence of CSI overall or by subtype based on TC level or type. Injury Severity Score correlated with incidence of CSI without cord injury (r = 0.387, p < 0.01). Regarding approach to CSC, differences existed by TC level and type for responsibility for CSC and protocols for CSC (p < 0.05). Level II TC felt early flexion-extension views were potentially harmful (60%); Level I TC did not (39%) (p < 0.05). Regarding indications for CSC, there was agreement on 10 of 11 clinical scenarios. For three of five clinical scenarios examining radiographic approach to CSC there was a broad distribution of approaches to patients with normal radiographs and cervical pain, altered mental status, coma.
Conclusion: Incidence of CSI is uniform by TC level and type. Incidence of spinal cord injury without fracture is low: 0.7%. Reported rate of missed CSI is very low: 0.01%. There is good agreement (>78%) among TC on indications for CSC but less agreement on radiographic approach to CSC.