Pulmonary ultrasound in COVID-19 and non-COVID-19 pneumonia in South Africa: An observational study.

Background: Pulmonary ultrasound techniques have historically been applied to acute lung diseases to describe lung lesions, particularly in critical care. Objectives: To explore the role of lung ultrasound (LUS) in hospitalised patients with hypoxaemic pneumonia during the COVID-19 pandemic. Methods: This was a single-centre prospective, observational study of two groups of adult patients with hypoxaemic pneumonia: those with COVID-19 pneumonia, and those with non-COVID-19 community-acquired pneumonia (CAP). A pulmonologist performed bedside LUS using the Bedside Lung Ultrasound in Emergency (BLUE) protocol, and the findings were verified by an independent study-blinded radiologist. Results: We enrolled 48 patients with COVID-19 pneumonia and 24 with non-COVID CAP. The COVID-19 patients were significantly older than those with non-COVID CAP (median (interquartile range (IQR)) age 52 (42 - 62.5) years v. 42.5 (36 - 52.5) years, respectively; p=0.007), and had a lower prevalence of HIV infection (25% v. 54%, respectively; p=0.01) and higher prevalences of hypertension (54% v. 7%; p=0.002) and diabetes mellitus (19% v. 8%; p=0.04). In both groups, close to 30% of the patients had severe acute respiratory distress syndrome. A confluent B-line pattern in the right upper lobe was significantly associated with COVID-19 pneumonia compared with the C pattern (relative risk (RR) 3.8; 95% confidence interval (CI) 1.7 - 8.6). Bilateral changes on LUS rather than unilateral or no changes were associated with COVID-19 pneumonia (RR 1.55; 95% CI 1.004 - 2.387). There were no statistically significant differences in median (IQR) lung scores between patients with COVID-19 pneumonia and those with non-COVID CAP (8 (4 - 11.5) v. 7.5 (4.5 - 12.5), respectively). Patients with COVID-19 pneumonia had a higher than predicted mortality. Logistic regression analysis showed a higher Simplified Acute Physiology Score (SAPS II) (RR 1.11; 95% CI 1.02 - 1.21) and a lower total LUS score indicating B lines v. consolidation (RR 0.80; 95% CI 0.65 - 0.99) to be associated with mortality. Conclusion: Patients with right upper zone consolidation were more likely to have non-COVID CAP than COVID-19 pneumonia. Finding a B pattern as opposed to consolidation was associated with mortality. The admission LUS score was unable to discriminate between COVID-19 and non-COVID CAP, and did not correlate with the ratio of partial pressure of oxygen to fractional inspired oxygen, clinical severity or mortality. Study synopsis: