Lung ultrasound versus chest X-ray in pediatric lower respiratory tract infections at a tertiary center: a prospective observational study.

BACKGROUND

Lower respiratory tract infections (LRTIs) are a major cause of morbidity in children. While chest X-ray (CXR) is widely used for diagnosis, it has limitations including radiation exposure and suboptimal sensitivity. Lung ultrasound (LUS) is an emerging rapid, non-invasive alternative. This study aimed to evaluate the diagnostic performance of LUS versus CXR in pediatric LRTIs, with a focus on subgroup variability.

METHODS

This prospective observational study included 338 children (aged 1 month-18 years) with suspected LRTIs admitted to a tertiary care hospital. All patients underwent clinical assessment, CXR, and LUS at admission and follow-up. LUS findings were compared with CXR using sensitivity, specificity, predictive values, accuracy, and Cohen's kappa (κ) for agreement.

RESULTS

LUS demonstrated high sensitivity for pulmonary pathology at admission (95.6%) and follow-up (96.7%), with specificity improving from 29.4% at admission to 62.9% at follow-up. Overall agreement between LUS and CXR progressed from fair at admission (Cohen's κ = 0.311, 95% CI: 0.198-0.424) to substantial at follow-up (Cohen's κ = 0.607, 95% CI: 0.501-0.713). Subgroup analysis demonstrated meaningful variation in LUS-CXR agreement across diagnostic categories. The highest concordance was observed in pneumonia (κ = 0.438-0.470, moderate agreement), consistent with LUS's established strength in detecting parenchymal consolidation. Agreement was notably lower in recurrent wheezing (κ = 0.208-0.283, fair agreement) and was poor to negligible in acute bronchiolitis (κ = -0.029-0.000) and the heterogeneous "others" group (κ = -0.053-0.182). LUS demonstrated high accuracy in detecting consolidations, subpleural lesions, and pleural abnormalities, and facilitated serial monitoring of disease progression during hospitalization.

CONCLUSION

LUS is a promising, radiation-free imaging modality for the diagnosis and follow-up of pediatric LRTIs, particularly pneumonia, demonstrating moderate to substantial agreement with CXR. Its lower specificity and reduced performance in airway-predominant conditions highlight the importance of cautious interpretation alongside clinical findings.