Diagnostic value of lung ultrasound in COVID-19: systematic review and meta-analysis

• Main Authors: Natalia N. Vetsheva, Roman V. Reshetnikov, Denis V. Leonov, Nikolas S. Kulberg, Olesya A. Mokienko
• Format: Article
• Language: English
• Published: Eco-Vector 2020-12-01
• Series: Digital Diagnostics
• Subjects: covid-19; lung ultrasound; severity grade estimate; diagnostic value; sensitivity; specificity
• Online Access: https://journals.eco-vector.com/DD/article/viewFile/46834/pdf

BACKGROUND:Effective and safe tools assisting triage decisions for COVID-19 patients could optimize the pressure on the healthcare system. COVID-19 often has respiratory manifestations, and medical imaging techniques provide an opportunity to assess the diseases severity. AIMS:To estimate the sensitivity and specificity of lung ultrasound for different degrees of pulmonary involvement in COVID-19 patients by a systematic review of English articles using PubMed and Google Scholar databases. Search terms included lung ultrasound, chest ultrasound, thoracic ultrasound, ultrasonography, COVID-19, SARS-CoV-2, coronavirus, diagnosis, diagnostic value, specificity, and sensitivity. Only studies addressing lung ultrasound diagnostic accuracy for patients with suspected COVID-19 using thoracic computed tomography, reverse transcription polymerase chain reaction, or laboratory data as a reference standard were included. Independent extraction of articles was performed by two authors using predefined data fields with subsequent assessment of study quality indicators. The random-effect model was used to analyze and pool lung ultrasound sensitivity and specificity across the included studies. Sixteen studies met our inclusion criteria, but only three of them divided patients into distinct and defined groups depending on the disease severity. We used the remaining studies data to assess the secondary outcomes: the values of sensitivity and specificity of lung ultrasound for COVID-19 regardless of the patients clinical status. Heterogeneity for primary and secondary outcomes was observed that remained when pooling for different scenarios (screening, assessing severity) and cohorts of participants. Lung ultrasound had the highest accuracy for confirmed COVID-19 patients with severe disease (sensitivity 87.6% 12.3%, specificity 80.5% 7.1%), and the lowest accuracy for the patients with mild disease (sensitivity 72.8% 7.1%, specificity 74.3% 2.7%). CONCLUSIONS:Lung ultrasound can be used in patients with confirmed COVID-19 to detect serious damage to the lung tissue. The diagnostic value of the method for assessing mild and moderate lung lesions is relatively low.