| Summary: | Introduction: The value of thoracic injury rule out criteria (TIRC) as a tool for decreasing the number of unnecessary chest radiographs in children has not been evaluated yet. Therefore, the present study was designed as a multi-center study to assess the validity of TIRC model in detection of very low risk children for traumatic intrathoracic injuries.
Methods: In this diagnostic accuracy study, clinical data and chest radiographs of 974 children less than 18 years of age (72.0% boys) who had presented to 5 hospitals, in Iran in 2018 were assessed. Data gathering and interpretation of radiographs were done by two independent researchers in each hospital. In the end, discriminatory power and calibration of the model was assessed with a 95% confidence interval (95% Cl).
Results: In the present study, age was not a predicting factor of abnormal findings in radiographs of children and adolescents (p=0.75); therefore, it was omitted from TIRC model and pediatric TIRC (pTIRC) was designed. Area under the curve of pTIRC rule was 0.97 (95% CI: 0.96-0.98) for prediction of abnormal chest X-Ray in children and adolescents. The sensitivity and specificity of pTIRC was 100% and 90.1%, respectively. The calibration of this decision rule had great concordance with the perfect line with a slope of 0.99 and intercept of 0.001. There was a 90.1% reduction in the number of unnecessary chest radiographs when using pTIRC decision rule.
Conclusion: pTIRC decision rule was introduced in the present study. pTIRC has excellent performance in identification of traumatic intrathoracic injuries and decreasing the number of unnecessary chest radiographs.
|
|---|
