Accuracy of craniocervical measurements on CT for identifying partial or complete craniocervical ligament injuries in pediatric patients.

PURPOSE: To assess the accuracy of craniocervical measurements for identifying craniocervical injuries and the frequency of subjective findings of craniocervical injuries on CT in pediatric patients. METHODS: Case-controlled retrospective review of patients ≤ 16 years old with craniocervical junction injuries. Receiver operator curves were created for common craniocervical measurements on CT comparing patients with complete and partial craniocervical injuries to uninjured cohort. Frequency of subjective CT findings of craniocervical injury was assessed in the injured cohort. RESULTS: For complete disruption injuries (CD) (n = 27), C1-C2 distance (AUC = 0.90, 95%CI = 0.83-0.97), atlanto-occipital distance (AUC = 0.95-0.98, 95%CI = 0.90-1.00), and basion-dens distance (AUC = 0.90, 95%CI = 0.82-0.98) had excellent accuracy diagnosing injury. Powers ratio (AUC = 0.85, 95%CI = 0.76-0.94) had good, basion-posterior axial line (AUC = 0.74, 95%CI = 0.61-0.86) fair, and atlanto-dental distance (AUC = 0.69, 95%CI = 0.57-0.82) poor accuracy. For partial disruption injuries (PD) (n = 21), basion-dens distance (AUC = 0.75, 95%CI = 0.62-0.88) had fair accuracy diagnosing injury. Powers ratio (AUC = 0.63, 95%CI = 0.47-0.79), C1-C2 distance (AUC = 0.60, 95%CI = 0.45-0.75), atlanto-dental distance (AUC = 0.55, 95%CI = 0.39 = 0.71), atlanto-occipital distance (AUC = 0.63-0.65, 95%CI = 0.47-0.81), and basion-posterior axial line (AUC = 0.60, 95%CI = 0.44-0.76) all had poor accuracy. Eighty-one percent (n = 22) of CD and 38% (n = 8) of PD patients had non-concentric atlanto-occipital joints. One hundred percent of CD patients had ≥ 1 soft tissue finding and eighty-one percent (n = 22) had ≥ 2 findings. Seventy-three percent (n = 16) of PD patients had ≥ 1 soft tissue finding. Eighty-six percent (n = 18) of PD patients had non-concentric atlanto-occipital joints and/or soft tissue findings. CONCLUSION: Craniocervical measurements have poor accuracy for identifying craniocervical injuries in pediatric patients with incomplete craniocervical ligament disruption. Subjective findings of craniocervical injury are frequently present on CT in pediatric patients and can help increase sensitivity for identifying injury.

There has to be an easier way: facet fracture characteristics that reliably differentiate AOSpine F1 and F2 injuries.

PURPOSE: To identify morphologic features of isolated cervical spine facet fractures that can reliably differentiate AOSpine F1 and F2 injuries. MATERIALS AND METHODS: Retrospective review of cervical spine CTs on all patients who sustained isolated cervical fractures of the facets presenting to our level 1 trauma center from August 2012 through December 2015. CTs were reviewed for facet fracture characteristics and AOSpine facet fracture classification. Association between facet fracture characteristics and AOSpine classification was assessed through multivariable logistic regression models. RESULTS: Fifty-six patients with cervical spine fractures isolated to the facets were included in the study. The mean age was 36 (range 9-90) years with 55.4% (n = 31) males. A significant correlation was found between subtype F1 and subtype F2 in laterality (left- or right-sided) (p = 0.004), interfacetal fracture involvement (p < 0.0001), transverse process involvement (p < 0.001), displacement of fracture fragment (p < 0.001), comminution of fracture (p < 0.0001), and vertebral arch disruption (p = 0.001). After multivariable analysis, left side laterality (p = 0.03), transverse process involvement (p = 0.01), and fracture comminution (p = 0.003) were associated with F2 fractures. CONCLUSION: Facet fractures with transverse process involvement or comminution have a higher probability of being an F2 fracture. These characteristics may be helpful when categorizing facet fractures using the AOSpine classification.