Background: Pediatric elbow fractures are common but remain challenging to accurately diagnose. Digital tomosynthesis is a technique that has shown promise in difficult adult fracture patterns but has not been formally studied in the pediatric population. Objective: To assess the added value of digital tomosynthesis on the detection and diagnostic confidence of pediatric elbow fractures. Materials and methods: A retrospective study was performed between January 2016 and December 2017 in pediatric patients (≤18 years) to assess the ability of conventional elbow radiographs and digital tomosynthesis to detect elbow fractures. One hundred twenty-one pediatric patients with concern for pediatric elbow trauma (64 males, 57 females; mean age: 8.1 years, range: 1 year to 17 years) were imaged with both conventional elbow radiographs and digital tomosynthesis. Two blinded pediatric radiologists identified fractures and indicated their diagnostic confidence. Observer agreement was assessed with Cohen’s Kappa coefficient and a nonparametric Wilcoxon rank sum test was used to compare the degree of diagnostic confidence between standard radiographs alone and standard radiographs with digital tomosynthesis. McNemar’s test was used to assess the difference in the rate of fracture detection between the two methods and sensitivity, specificity, precision, accuracy and diagnostic odds ratios were calculated. Results: Compared with standard radiographs alone, standard radiographs with digital tomosynthesis improved inter-rater agreement, sensitivity, specificity, accuracy, precision and the diagnostic odds ratio for fracture detection and increased diagnostic confidence (Rater 1: P=0.01, Rater 2: P=0.003). Conclusion: The addition of digital tomosynthesis with conventional elbow radiographs improves diagnostic confidence and performance for the detection of pediatric elbow fractures.
- Diagnostic confidence
- Digital tomosynthesis
ASJC Scopus subject areas
- Pediatrics, Perinatology, and Child Health
- Radiology Nuclear Medicine and imaging