Lung Ultrasound Findings and Algorithms to Detect Pneumonia: A Systematic Review and Diagnostic Testing Meta-Analysis
Eduardo Messias Hirano Padrão, Bruno Caldeira Antônio, Tiffany Alexis Gardner, Isabele Ayumi Miyawaki, Cintia Gomes, José Eduardo Riceto Loyola Júnior, Maria João Andrade, Isabela Reis Marques, Isabela A F de Souza, Caroliny Hellen Silva, Vittoria Caporal S Moreira, Brian Bustos, Augusto Barreto do Amaral Neto, Jonah Rubin
Abstract
OBJECTIVE: Lung ultrasound is increasingly used for diagnosing pneumonia due to its accessibility, low cost, and lack of radiation exposure. This systematic review and meta-analysis aimed to evaluate the diagnostic accuracy of individual lung ultrasound findings and algorithms for pneumonia across various clinical settings compared with chest radiography and CT. DATA SOURCES: We systematically searched PubMed, Embase, and Cochrane databases. STUDY SELECTION AND DATA EXTRACTION: We searched for studies assessing the sensitivity and specificity of lung ultrasound findings and algorithms for pneumonia. Studies including adult patients with community-acquired, hospital-acquired, or ventilator-associated pneumonia (VAP) were eligible. Data on sensitivity, specificity, and likelihood ratios for ultrasonographic findings and algorithms were pooled using bivariate linear mixed models and Bayesian analyses. The risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. DATA SYNTHESIS: Twenty-six studies, totaling 3454 patients, were included. The Bed Lung Ultrasound in Emergency (BLUE) protocol demonstrated the highest sensitivity (0.88; 95% CI, 0.84-0.92) among all criteria studies, whereas dynamic air bronchograms had the highest specificity (0.96; 95% CI, 0.91-0.99). Focal B-lines had low sensitivity (0.24; 95% CI, 0.12-0.43) and high specificity (0.96; 95% CI, 0.86-0.99). Sensitivity analyses indicated reduced specificity for lung ultrasound in patients with VAP across all evaluated criteria. Bayesian analyses yielded consistent results across different prior assumptions. CONCLUSIONS: Lung ultrasound demonstrates good diagnostic performance for detecting community-acquired and hospital-acquired pneumonia. However, its utility in diagnosing VAP is limited, suggesting the need for complementary diagnostic tools in this patient group. This underscores the importance of lung ultrasound as a frontline diagnostic tool for pneumonia. To the best of our knowledge, this is the first meta-analysis to evaluate the specificity and sensitivity of each specific finding identified by lung ultrasound.