Lois K. Lee, MD, MPH1Fahd A. Ahmad, MD, MPH2Lorin R. Browne, DO3
Key Points
Question Which of 3 prospectively derived clinical cervical spine injury (CSI) prediction rules for radiologic imaging (Pediatric Emergency Care Applied Research Network CSI prediction rule [PECARN CSI rule], National Emergency X-Radiography Utilization Study [NEXUS], Canadian Cervical Spine [c-spine] rule [CCR]) has the best test characteristics?
Findings In this comparative effectiveness study of the performance of CSI prediction rules in a cohort of 22 430 children, the PECARN CSI rule had the best test characteristics and lowest projected CT imaging rate.
Meaning The PECARN CSI rule had the best test characteristics for risk stratifying children with potential CSI after blunt trauma who require cervical spine imaging.
Abstract 
Importance Pediatric cervical spine injury (CSI) is a rare occurrence; however, CSI can result in significant disability and death. It is essential to determine the optimal CSI clinical prediction rule to risk stratify children with potential CSI after blunt trauma who require cervical spine imaging.
Objective To compare the test characteristics and projected imaging rates between 3 prospectively derived CSI clinical prediction rules: the Pediatric Emergency Care Applied Research Network CSI prediction rule (PECARN CSI rule), the National Emergency X-Radiography Utilization Study (NEXUS), and the Canadian Cervical Spine (c-spine) rule (CCR).
Design, Setting, and Participants This comparative effectiveness study was a planned secondary analysis of a prospective observational study enrolling from December 2018 to October 2021 in 18 PECARN emergency departments. Eligible participants were children up through age 17 years presenting after blunt trauma. Data were analyzed between March 2024 and January 2025.
Exposures Enrollment in the primary study to develop and validate the PECARN CSI prediction rule.
Main Outcome Measures Test characteristics with 95% CIs (sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV]) and area under the curve (AUC) for the receiver operator curves (ROC) for the detection of CSI using each of the 3 rules. We also estimated the projected c-spine imaging rate (radiography or computed tomography [CT]) based on criteria for each of the 3 rules.
Results There were 22 430 eligible children enrolled (median [IQR] age 8 [2.0-13.0] years; 13 068 male [58.3%]) and 433 (1.9%) had CSI. C-spine imaging was performed 12 768 children (56.9%): 8912 (39.7%) had radiography and 3856 (17.2%) were imaged with CT. The sensitivity of the 3 rules was: PECARN CSI rule, 93.3% (95% CI, 90.9%-95.7%); NEXUS, 85.7% (95% CI, 82.4%-89.0%); and CCR, 90.8% (95% CI, 88.0%-93.5%). The NPV of the 3 rules was: PECARN, 99.8% (95% CI, 99.7%-99.9%); NEXUS, 99.6% (95% CI, 99.5%-99.7%); and CCR, 99.7% (95% CI, 99.6%-99.8%). Strictly applying each rule resulted in projected CT imaging of 1549 children (6.9%) for PECARN, 2419 (10.8%) for NEXUS, and 2968 (13.2%) for CCR.
Conclusions and Relevance In this comparative effectiveness study of CSI prediction rules in children, the PECARN CSI rule had the highest sensitivity and NPV for identifying children at risk for CSI after blunt trauma, with the lowest projected CT imaging rate.
