The clinical presentation of pediatric pelvic fractures

Abstract

Background: Few studies have addressed the presentation and clinical impact of pediatric pelvic fractures. We sought to describe pediatric blunt trauma patients with pelvic fracture (PF) and to evaluate the sensitivity and specificity of physical examination at presentation for diagnosis.

Methods: Retrospective analysis of all PF and control (NPF) patients from our pediatric institution over an 8-year period.

Results: A total of 174 patients (88 PF, 86 NPF) were included. Median patient age was 8 years (range, 3 months to 18 years), with 54% males. The most common mechanisms of injury for PF patients were automobile-related accidents (75%). There were 140 patients (87%) who were transported by air or ground medical services. At presentation, approximately 16% of PF patients had a Glasgow Coma score of <15, a mean Revised Trauma Score of 7.49, and a median Injury Severity Score (ISS) of 9. Thirty-one PF patients (35%) had an ISS of >15 indicating severe, multiple injuries. Sixty-eight PF patients (77%) had severe isolated injuries (Abbreviated Injury Scale 1990 value of >3); 11% of PF patients required transfusions, and 2% died. Fifteen PF patients (17% ) had no pelvic ring disruption; 39 (43%) had a single pelvic ring fracture, 22 (2%) had two pelvic ring fractures, 2 (2%) had acetabular fractures, and 10 (11%) had a combination of pelvic fractures. An abnormal physical examination of the pelvis was noted in 81 patients with PF (92% sensitivity, 95% confidence interval [CI] = 0.89-0.95), 15 NPF patients had an abnormal examination (79% specificity, 95% CI = 0.74-0.84). The positive predictive value of the pelvis examination was 0.84, and the negative predictive value was 0.89. The most common abnormal pelvis examination finding was pelvic tenderness in 65 PF patients (73%). A total of seven PF patients had a normal examination of the pelvis; four had a depressed level of consciousness (defined as GCS <15), and six patients had a distracting injury.

Conclusions: Pediatric blunt trauma patients with pelvic fracture represent a severely injured population but generally have lower transfusion rates and mortality than noted in adult studies. The pelvis examination appears to be sensitive and specific in this retrospective study. However, an altered level of consciousness and/or distracting injuries may affect examination sensitivity and specificity. Based on this retrospective study, we cannot advocate eliminating pelvic radiographs in the severely injured, blunt trauma patient. Prospective studies are recommended.