O1020: Comparison of C-spine CT Decision Rules for Children After Blunt Trauma

Background: Several clinical decision rules, derived primarily in adults, were developed to assist in guiding imaging decisions for cervical spine injury (CSI) evaluation. Comparing the diagnostic accuracy for these rules in identifying children at high risk for CSI will inform best practices for imaging after blunt trauma. The objectives of this study were to compare the test characteristics and projected imaging rates between three CSI clinical decision rules: Pediatric Emergency Care Applied Research Network CSI decision rule (PECARN CSI rule), Canadian C-spine Rule (CCR), and the National Emergency X-Radiography Utilization Study (NEXUS).

Methods: We conducted a planned secondary analysis of a multi-center, prospective observational study of children (0-17 years old) presenting after blunt trauma to 18 PECARN emergency departments from December 2018-October 2021. We included all children enrolled in the primary study (to develop and validate the PECARN CSI rule). Study data on injury mechanisms, predisposing conditions, neck and neurological symptoms, and physical examination findings were applied to the three different CSI rules. We calculated the test characteristics with 95% confidence intervals (sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV)) and created Receiver Operator Curves for the detection of CSI for the three rules. We also estimated the projected c-spine imaging rate (x-ray or CT) based on criteria for each of the three rules.

Results: There were 22,430 eligible children enrolled and 1.9% (433) had CSI. C-spine imaging was performed in 56.9% (12,768): 39.7% (8912) with only x-rays and 17.2% (3856) with CT. The sensitivity of the three rules was: PECARN CSI rule 93.3% (95% CI 90.9, 95.7); CCR 75.3% (95% CI 69.9, 80.7); and NEXUS 76.7% (95% CI 72.7, 80.7) (Table). The NPV of the three rules was: PECARN 99.8% (95% CI 99.7, 99.9); CCR 97.8% (95% CI 97.3, 98.4); and NEXUS 99.4% (95% CI 99.3, 99.5). The Area Under the Curve for the Receiver Operator Curves was: PECARN 0.77, CCR 0.56, and NEXUS 0.76 (Figure). Strictly applying each rule resulted in projected imaging frequencies of: PECARN 41.1%, CCR 79.1%, and NEXUS 25.2%.

Conclusion: The PECARN CSI rule had the best sensitivity and NPV for identifying children at low risk for CSI after blunt trauma. Application of the PECARN CSI rule in our study population could have reduced c-spine imaging in these patients. The PECARN CSI rule has the best diagnostic accuracy for the evaluation of those at risk for CSI among pediatric trauma patients.