Background: Abdominal radiographies (AR) are a common part of the radiologic evaluation in the pediatric emergency department (ED). AR have a low efficacy in making a definitive diagnosis, leading to overuse, increased resource utilization and length of stay in the pediatric ED. The objective of this study is to evaluate the utilization of the AR by measuring its diagnostic utility in predicting intra-abdominal pathology in the ED. 

Methods: We conducted a retrospective, cross-sectional study of patients who visited a pediatric ED between January 1st, 2017 and December 31st, 2019. We included patients 0-18 years old with an abdominal radiography. The diagnostic testing accuracy of the AR was evaluated by analyzing the sensitivity, specificity, predictive values and likelihood ratio (LR). Comparisons among groups were calculated through Chi-square, or Fisher’s Exact, as appropriate. 

Results: A preliminary analysis of 310 (7.2%) visits were evaluated out of 4,288 identified cases. The top 3 chief complaints were abdominal pain (56.8%), vomiting (9%) and constipation (8%). An abnormal AR was found in 37% of cases. The prevalence of abnormal AR per chief complaint was 32%, 38% and 56%, respectively. There were 10% clinically significant diagnoses. The AR diagnostic testing accuracy showed 61% sensitivity, 65% specificity, 17% positive predictive value (PPV), 94% negative predictive value (NPV) and LR 8.2 (p=0.005). 

Conclusions: The top three chief complaints that received an AR were abdominal pain, vomiting, and constipation. Although the prevalence of an abnormal AR those cases was moderate, its clinical significance is low. This study shows a low prevalence of intra-abdominal pathologic processes that can be identified by an AR. Despite having a good NPV, the data suggests that an AR is not a useful diagnostic tool in the pediatric ED due to its limited ability to rule in or rule out clinically significant diagnoses.