Craniocervical Dissociation in Pediatric Patients: Pearls and Pitfalls of Diagnosis and Management.

Abstract

The aims of this study were to document the injury pattern in pediatric traumatic craniocervical dissociation (CCD) and identify features of survivors. Pediatric traumatic CCDs, diagnosed between January 2004 and July 2016, were reviewed. Survivors and nonsurvivors were compared. Categorical and continuous variables were analyzed with Fisher exact and t tests, respectively. Twenty-seven children were identified; 10 died (37%). The median age was 60 months (ranges, 6-109 months [survivors], 2-98 months [nonsurvivors]). For survivors, the median follow-up was 13.4 months (range, 1-109 months). The median time to mortality was 1.5 days (range, 1-7 days). The injury modality was motor vehicle collision in 18 (67%), pedestrian struck in 8 (30%), and 1 shaken infant (3%). For nonsurvivors, CCD was equally diagnosed by plain radiograph and head/cervical spine computed tomography scan. For survivors, CCD was diagnosed by computed tomography in 7 (41%), magnetic resonance imaging in 10 (59%), and none by radiograph. Seven diagnosed by magnetic resonance imaging (41%) had nondiagnostic initial imaging but persistent neck pain. Magnetic resonance imaging was obtained and was diagnostic of CCD in all 7 (P < 0.01). Survivors required significantly less cardiopulmonary resuscitation (P < 0.01), had lower Injury Severity Scores (P < 0.01), higher Glasgow Coma Scale scores (P < 0.01), and shorter transport times (P < 0.01). Significantly more involved in motor vehicle collisions survived (P = 0.04). Nine (53%) had no disability at follow-up evaluation. In pediatric CCD, high-velocity mechanism, cardiac arrest, high Injury Severity Score, and low Glasgow Coma Scale score are associated with mortality. If CCD is correctly managed in the absence of cardiac arrest or traumatic brain or spinal cord injury, children may survive intact.