Lung ultrasound for the diagnosis of community-acquired pneumonia in infants and children

Abstract

Background Lung ultrasound (LUS) has been suggested as an effective tool in the evaluation of community-acquired pneumonia (CAP) in pediatrics. Objectives Evaluation of the validity of LUS in the diagnosis of CAP in children, compared with chest radiography (CXR). Patients and methods A prospective observational study was done. LUS and CXR were carried out in 90 children with median (range) age of 2.2 (0.5–15) years who presented with clinical features suggestive of CAP. Results One hundred and five patients were initially enrolled. Fifteen patients were secondarily excluded, resulting in 90 finally analyzed patients. CAP diagnosis was established in 84 patients, and six patients were finally diagnosed as having acute bronchiolitis. Of the 84 patients with CAP, CXR detected 73 (87%) cases and LUS spotted 82 (97.6%) cases. Moreover, CXR identified 11 false-negative cases and one false-positive case, whereas LUS detected two false-negative cases and one false-positive case. The sensitivity, specificity, positive predictive value, and negative predictive value of LUS were 97.6% [95% confidence interval (CI), 91.7–99.7], 83.3% (95% CI, 35.9–99.6), 98.8% (95% CI, 93.2–99.8), and 71.4% (95% CI, 37.8–91.2), respectively, whereas, the sensitivity, specificity, positive predictive value, and negative predictive value of CXR were 86.7% (95% CI, 78.5–92.5), 83.3% (95% CI, 43.6–96.9), 98.6% (95% CI, 92.4–99.7), and 31.2% (95% CI, 19.1–46.7), respectively. Conclusion LUS, done by an expert radiologist, can be considered a promising alternative tool to CXR in the diagnosis of children with suspected CAP.