Accuracy and Determinants of Radiographic Diagnosis in Pediatric Hand Fractures

To investigate the accuracy of radiographic diagnosis of pediatric hand fractures and factors that contribute to diagnostic errors.

We retrospectively reviewed cases of pediatric hand injuries in Srinagarind Hospital, Thailand, from January 2019 to December 2021. A total of 350 patients with accessible radiographs were interpreted by junior/senior radiology trainees, a pediatric/a senior radiologist, and two senior orthopedists. Age, sex, hand side, fracture type, and injury mechanism were collected. Results: Among 167 patients with fractures, 55 (32.9%) had multiple fractures. For missed fractures (103 instances), junior radiology trainees missed more frequently than pediatric radiologists, with 35 (34%) vs. 8 (7.8%) instances (P = 0.004). Overall accuracy in the interpretation of different fracture types was 100% for greenstick, 77.4% for torus buckle, 76% for simple fractures. The pediatric radiologist showed the highest accuracy in the diagnosis of Salter-Harris type compared to others, 49 (28.7%) vs. 25-33 (14.6%-19.3%) (P = 0.001). On univariable analysis, vehicle injury mechanism was a borderline negative factor associated with missed fractures (odds ratio (OR), 0.38; 95% confidence interval (CI), 0.14-1.01; P = 0.05).

Our findings suggest that review by experienced pediatric radiologists is associated with higher diagnostic accuracy for pediatric hand fractures, particularly Salter-Harris injuries. Prospective studies are needed to determine whether incorporating senior review into routine workflows reduces missed diagnoses.