Appropriateness of lung ultrasound for the diagnosis of COVID‐19 pneumonia

Abstract Background Chest radiography (CXR) and computerized tomography (CT) are the standard methods for lung imaging in diagnosing COVID‐19 pneumonia in the intensive care unit (ICU), despite their limitations. This study aimed to assess the performance of bedside lung ultrasound examination by a critical care physician for the diagnosis of COVID‐19 pneumonia during acute admission to the ICU. Method This was an observational, prospective, single‐center study conducted in the intensive care unit of Adan General Hospital from April 10, 2020, to May 26, 2020. The study included adults with suspicion of COVID‐19 Infection who were transferred to the ICU. Patients were admitted to the ICU directly from the ED after reverse transcriptase‐polymerase chain reaction (RT‐PCR) swabs were sent to the central virology laboratory in Kuwait, and the results were released 16 to 24 hours after the time of admission. A certified intensivist in critical care ultrasound performed the lung ultrasound within 12 hours of the patient's admission to the ICU. The treating physician confirmed the diagnosis of COVID‐19 pneumonia based on a set of clinical features, inflammatory markers, biochemical profile studies, RT‐PCR test results, and CXR. Results Of 77 patients with suspected COVID‐19 pneumonia, 65 (84.4%) were confirmed. The median age of the patients was 48 (31‐68) years, and 51 (71%) were men. In the group of patients with confirmed COVID‐19 pneumonia, LUS revealed four signs suggestive of COVID‐19 pneumonia in 63 patients (96.9%) (sensitivity 96.9%, CI 85%‐99.5%). Two patients presented with unilateral lobar pneumonia without other ultrasonic signs of COVID‐19 pneumonia but with positive RT‐PCR results. Among patients in the group without COVID‐19 pneumonia who had negative RT‐PCR results, 11 (91.7%) were LUS negative for COVID‐19 pneumonia (specificity 91.7%, 95% CI 58.72%‐99.77%). Conclusions During the COVID‐19 outbreak, LUS allows the identification of early signs of interstitial pneumonia. LUS patterns that show a combination of the four major signs offer high sensitivity and specificity compared to nasopharyngeal RT‐PCR.


| INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-  infects cells in the lower respiratory tract, 1 causing pneumonia as the primary complication of the disease. Chest radiography (CXR) and computerized tomography (CT) are the standard methods for lung imaging in diagnosing COVID-19 pneumonia in the intensive care unit.
The radiological picture of COVID-19 pneumonia typically shows interstitial diffuse bilateral pneumonia with lesions exhibiting an asymmetric and patchy distribution in the lung periphery, a suitable site for ultrasound investigation. 2 During the pandemic era of SARS-CoV-2, lung ultrasound has the advantage of being noninvasive and can be performed quickly at the bedside. Lung ultrasound can help diagnose cases with suspicion of COVID-19 pneumonia in patients who have respiratory symptoms that necessitate urgent and early admission to the critical care area.
Looking for early sonographic signs of pneumonia upon admission can be helpful in the early diagnosis of COVID-19 pneumonia. Herein, we tested the sensitivity and specificity of these signs in correlation with the standard test for SARS-CoV-2 disease with respiratory involvement.

| Study population
This was an observational, prospective, single-center study conducted in the intensive care unit of Adan General Hospital from April 10, 2020, to May 10, 2020. The Ethical Committee of the Ministry of Health in Kuwait approved the study protocol, and informed consent was obtained from all patients or their next of kin.
Consecutive patients were included if they were >18 years of age with suspicion of COVID-19 infection and had been transferred to the ICU with fever or suspected respiratory infection plus one of the following: respiratory rate > 30 breaths/min, severe respiratory distress, and SpO 2 <93% on room air. 3 Patients were admitted to the ICU directly from the ED after nasopharyngeal samples were collected for reverse transcriptase-polymerase chain reaction (RT-PCR) and sent to the central virology laboratory in Kuwait, and the results were released 16 to 24 hours after swab performance. Clinical data were entered on a separate standardized data collection form at the time of patient enrollment by the treating critical care physician. Clinical data included the patient's age and sex, presenting symptoms, medical history, oxygen saturation from pulse oximetry, and chest radiograph. A level 4 operator (entrusted to act unsupervised) in critical care ultrasound who was blinded to the RT-PCR results, 4 if available at the time of examination, performed the lung ultrasound within 12 hours of the patient's admission to the ICU ( Figure 1).
All included patients were prospectively evaluated until discharge.
The final diagnosis was made by the physician in charge based on RT-PCR, radiological reports (chest CT and CXR), clinical progress, inflammatory markers, and microbiology studies. We compared the ultrasound and RT-PCR results with the final diagnosis made by physicians in charge.
If the initial upper respiratory sample result was negative and the suspicion for disease remained high, repeat testing generally occurred 24 hours twice after the initial testing, or a lower respiratory tract sample was collected (eg, sputum, bronchoalveolar lavage fluid, tracheal aspirate) if accessible. We also compared the results of chest CT and LUS with the final diagnosis made by the treating physician who was blinded to the results of the LUS.

| Protocol
We performed lung ultrasonography for all patients admitted to the ICU with suspicion of COVID-19 infection using a 12-zone method. 5,6 F I G U R E 1 Flow diagram. LUS, lung ultrasound; RT-PCR, reverse transcriptase-polymerase chain reaction; positive (+) or negative (À) for abnormality; CT, computerized tomography; CXR, chest-X-ray; TP, true positive; TN, true negative; TN, true negative; FN, false negative There were six zones in each hemithorax: two anterior, two axillary,  3. Absence of significant pleural effusion.

Presence of multiple subpleural consolidations of various sizes
(videos S1 and S2).

| Data analysis
Statistical analyses were performed using SPSS 19. Sensitivity, specificity, and the positive and negative likelihood ratios of lung ultrasound and RT-PCR for the diagnosis of COVID-19 pneumonia were calculated. The McNamara test was used for dichotomous variables when appropriate P < .05 indicated significant differences. Our assumption of a result with 95% specificity and an approximate 1% prevalence and .05 confidence interval yielded an approximate optimal sample size of 90. 9 Table 1.

| RESULTS
In the group of patients with confirmed COVID-19 pneumonia (  Small studies have evaluated the sensitivity and specificity of CT compared to nasopharyngeal RT-PCR, and CT currently shows the highest sensitivity of any test for COVID-19. 10 Thoracic CT imaging has been proposed as a primary screening tool for COVID-19 detection as it performs better than PCR. 10 The radiological lung abnormalities found on CT may antedate the physical symptoms of COVID-19; however, CT is a finite resource, exposes additional healthcare personnel to infected patients, and may not be available in some healthcare set- LUS, when performed with trained clinicians, metanalysis, and review articles, can detect pneumonia with similar accuracy and reliability to chest radiographs. 11,12 In the ICU setting, LUS was superior to chest X-ray in detecting pneumonia. 13 Although CXR has poor sensitivity and specificity compared to chest CT and LUS, it remains the standard protocol for diagnosing the disease. Plain radiographs can miss up to 40% of confirmed COVID-19 cases due to the nature of the disease, in which lesions are peripherally distributed, and pathology is evident primarily in the terminal alveoli and close to the pleural interface. These areas are well visualized on CT and LUS but are more difficult to see on plain imaging. 14,15 We used a previously validated scheme of LUS for other diseases, as there is no validated scheme for COVID-19. 5 We included scanning of the posterior and lateral zones, where lung lesions are more commonly seen in patients with COVID-19. 16  to some authors could be due to small pulmonary infarcts. 18 We had two elderly patients with multiple comorbidities who presented with isolated large lobar consolidation without effusion and with dynamic air bronchograms. Despite their being negative for COVID-19, their RT-PCR results appeared positive. Whether lobar pneumonia was secondary (bacterial) or due to viral infection was difficult to determine with certainty, although lobar pneumonia with dynamic air bronchogram is most often bacterial. 16 The specificity of the RT-PCR tests is 100% because the primer design is specific to the genome sequence of SARS-CoV-2; however, the incidence of false negatives from nasopharyngeal swab sampling is high. 19 False negatives result primarily from deficient sampling techniques and inappropriate timing of sample collection in relation to illness onset. We had three cases of false negatives, and LUS was more in keeping with the clinical picture of COVID-19 infection and other lab tests, such as ferritin level, procalcitonin, D-dimer, and fibrinogen, and we proceeded to collect a lower respiratory tract sample, which turned out to be positive.
This study's limitation is that it is a single-center study involving a relatively small number of patients with a sample size that was less than optimal. Although we had a control group of patients with disorders that may be confused with COVID-19 (non-COVID-19 group), we did not include a sufficient number of patients. Zouheir Bitar had full access to all the data in this study and acts as guarantor for the integrity of the data and the accuracy of the data analysis.

TRANSPARENCY STATEMENT
The lead author confirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE
This prospective observational study was approved by the Research Ethics Committee of the Ministry of Health in Kuwait (approval number: 2020/1471). Informed consent from the patients was obtained from their legally authorized representatives before enrolment in the study. Images are entirely unidentifiable, and consent for publication is not required.

DATA AVAILABILITY STATEMENT
The datasets used and analyzed during the current study are available from the corresponding author on request.