A combination of clinical presentation with positive reverse transcriptase polymerase chain reaction (PCR) testing is the current standard to diagnose Coronavirus Disease 2019 (COVID-19).^{1, 2} Challenges to this approach include lack of specificity of signs and symptoms, and limitations of PCR testing including test availability, delays in obtaining test results, and false negative test results early in the clinical course.^{1, 3}

Chest computed tomography (CT) scans have high sensitivity (98%) for detecting pulmonary infiltrates compared to PCR testing (78%) in COVID-19.¹ However, CT scans are costly, require extensive disinfection, lack portability, and expose patients to radiation. The American College of Radiology explicitly recommends against routine use of CT scans in COVID-19 patients. Lung ultrasound (LUS) has shown strong correlation with chest CT scans for diagnosing and monitoring COVID-19 lung disease.¹ Its portability, ease of disinfection, and immediate availability of results are major advantages in COVID-19.

This article describes common LUS findings, diagnostic accuracy of LUS compared to CT scans, different LUS protocols and scoring systems, and potential use for prognostication in COVID-19.

Diagnostic Accuracy

LUS has comparable diagnostic accuracy as chest CT scans for severe COVID-19 lung disease.³ In an observational study of suspected COVID-19 patients, LUS had a sensitivity of 92%, specificity of 71%, positive likelihood ratio of 3.1, and negative likelihood ratio of 0.1 compared to chest CT scans, and no significant difference was seen in sensitivity and specificity of LUS versus chest CT scan. Another study demonstrated similar sensitivity (89%) for LUS in patients suspected of COVID-19 presenting to an emergency department.⁵

LUS Findings in COVID-19

The posterior and lower lung zones are most often affected in COVID-19.² New or worsening infiltrates in the anterior zones may herald clinical deterioration.⁶ LUS findings in COVID-19 typically extend to the periphery, making them easily visualizable with ultrasound. LUS patterns have been progressively described as follows (see images):^{1, 2}

• mild to moderate (early): Irregular and thickened pleural line; discrete B-lines alternating with normal lung with A-lines (“skipped lesions”); small consolidations (~1 cm).
• severe (progressive): Confluent or fused B-lines; large consolidations.
• critical (advanced): Extensive confluent B-lines and consolidations in upper and anterior lung zones; bilateral interstitial pattern with consolidations ± air bronchograms in the posterobasal lung zones.

Pleural effusions and lymphadenopathy are only seen in 7-9% of COVID-19 patients.¹ A smooth pleural line with discrete B-lines in the upper lung lobes is suggestive of cardiogenic pulmonary edema, while an isolated lower lobe consolidation with dynamic air bronchograms is more likely bacterial pneumonia.^{2, 4}

Protocols

Multiple LUS protocols have been described for evaluating COVID-19 patients. A low-frequency phased-array^{3, 6} or curvilinear transducer⁵ is used to evaluate the lung parenchyma while a high-frequency linear-array transducer allows detailed assessment of the pleural line. A lung or abdominal exam preset with tissue harmonic imaging turned off is typically used, and the screen depth is set to 12-15cm.³

Data comparing various COVID-19 LUS scanning protocols are limited. Heldeweg, et al, found a 6- and 12-zone protocol may be equivalent when correlating findings to a CT scoring index to predict a composite outcome of death and prolonged ICU stay. Similar to the popular BLUE protocol (Bedside Lung Ultrasound in Emergency), a 6-zone protocol allows faster imaging and reduces clinician exposure time.²

Scoring System and Prognostication

Soldati, et al, proposed a standardized COVID-19 LUS scoring system that can be used for triage, severity classification, and prognostication.
^{1, 2, 5, 6} Each lung zone is scored 0 to 3 and the total score of all lung zones reflects the following degree of reduced lung aeration:²

• 0 = Normal aeration pattern with continuous pleural line and A-lines.
• 1 = ≥3 Discrete B-lines suggesting some loss of aeration. Pleural line may appear thickened and irregular.
• 2 = Confluent B-lines with or without small subpleural consolidations, suggesting severe loss of aeration.
• 3 = Large consolidation, signifying complete loss of aeration.

This scoring system was applied to hospitalized COVID-19 patients and the predictive ability of an abnormal LUS exam was superior to CXR. Patients with a high (19-36 points) versus low (0-18 points) LUS score had a 2.6-fold increased mortality and a 4.2-fold increased composite outcome of death or need for mechanical ventilation.⁶

Areas of Uncertainty

Several areas of uncertainty exist for future research of LUS in COVID-19. First, consensus on a standardized protocol and scoring system is needed. Second, even though LUS outperforms CXR for detection of pulmonary infiltrates due to COVID-19,^{4, 6} the effect of LUS-guided care versus routine care on patient outcomes, healthcare costs, and resource utilization are needed. Additionally, the role of LUS to guide decisions about mechanical ventilation and proning of critically ill COVID-19 patients with respiratory failure warrants further investigation.

Conclusion

LUS outperforms CXR for detection of pulmonary infiltrates and correlates well with chest CT findings in COVID-19.^1-3 A 6- or 12-zone LUS scanning protocol provides high diagnostic accuracy in COVID-19, and a LUS score can be used for prognostication.^{2, 5, 6} Future research and consensus are needed to develop standardized protocols and evaluate the impact of LUS on health outcomes of COVID-19 patients.