key: cord-1015072-mmestvus
authors: Orlandi, D; Battaglini, D; Robba, C; Viganò, M; Bergamaschi, G; Mignatti, T; ML, Radice; Lapolla, A; Turtulici, G; Pelosi, P
title: COVID-19 phenotypes, lung ultrasound, chest computed tomography, and clinical features in critically ill mechanically ventilated patients
date: 2021-07-24
journal: Ultrasound Med Biol
DOI: 10.1016/j.ultrasmedbio.2021.07.014
sha: 10ed562171108d11bc04bfcb6eaa4ba6af36ff28
doc_id: 1015072
cord_uid: mmestvus

Chest computed tomography (CT) may provide insights in the pathophysiology of COVID-19, although it is not suitable for a timely bedside dynamic assessment of patients admitted to intensive care unit (ICU); therefore, lung ultrasound (LUS) has been proposed as complementary diagnostic tool. The aims of this study were to investigate different lungs’ phenotypes in patients with COVID-19, assessing differences between ICU survivors and non-survivors in CT and LUS scores. We also explored the association between CT and LUS, and oxygenation (PaO(2)/FiO(2)) and clinical parameters. Thirty-nine COVID-19 patients were included. CT scan revealed type 1, 2, 3 phenotypes in 62%, 28%, and 10% respectively. Among survivors, pattern-1 was prevalent (p<0.005). Chest CT and LUS scores differed between survivors and non-survivors both at ICU admission and 10-days thereafter and were associated with ICU-mortality. Chest CT score positively correlated with LUS findings at ICU admission (r=0.953, p<0.0001), while inversely correlated with PaO(2)/FiO(2) (r=-0.375, p=0.019), and C-Reactive Protein (r=0.329, p=0.041). LUS score inversely correlated with PaO(2)/FiO(2) (r=-0.345, p=0.031). COVID-19 presents distinct phenotypes, with differences between survivors and non-survivors. LUS should be considered as valuable monitoring tool in ICU setting, since it may correlate with CT findings and mortality, although it cannot predict oxygenation changes.

A novel human coronavirus, the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), was identified in Wuhan, China, in late 2019 . The pandemic is still ongoing in the majority of countries with a significant impact on national healthcare systems and general activities (WHO 2020; Albano et al. 2020; Allam et al. 2020; Balestrino et al. 2020; Bandirali et al. 2020) . Coronavirus disease 2019 (COVID-19) is a new viral infection that commonly starts from the lower respiratory tract and further extends to other organs, making treatment options challenging Robba et al. 2020) . More than 80% of confirmed COVID-19 cases present as a mild febrile illness. However, a small proportion of patients will experience critical illness, with many of these requiring mechanical ventilation and multiorgan support (Katagiri et al. 2020) . To date, the respiratory management of COVID-19 has relied on the general principles of acute respiratory distress syndrome (ARDS) management, and chest computed tomography (CT) may provide interesting insights into the pathophysiology and individualization of mechanical ventilation in these patients Lu et al. 2020; Lanza et al. 2020) . However, chest CT is not suitable for a timely dynamic assessment of patients, particularly in critically ill cases. Chest X-ray can be performed at the bedside but has the drawback of low sensitivity (Bandirali et al. 2020) . In recent years, lung ultrasound (LUS) examination has become more widely used in the evaluation of lung diseases (Volpicelli et al. 2012) , as it is a convenient, fast, non-invasive, readily repeatable tool and does not involve ionizing radiation. Hence, LUS should be considered as a valid alternative to CT scan for characterizing COVID-19 pneumonia in ICU setting and difficulties to perform sequential CT scans (Allinovi et al. 2020; Vetrugno et al. 2020; .

The primary aim of this study was to investigate the presence at CT scan of different lungs' phenotypes in critically ill patients affected by COVID-19, and to assess differences between ICU survivors and non survivors in CT TSS and LUS scores. We also aimed to explore the possible association between chest CT scan and LUS; and between chest CT, LUS and oxygenation (arterial partial pressure of oxygen (PaO 2 )/fraction of inspired oxygen (FiO 2 ), ventilation and clinical parameters.

This study was conducted in accordance with the current version of the Declaration of Helsinki. Informed consent was obtained from each enrolled patient. Our institutional review board (IRB) of Liguria region, Italy approved this current study (registry number 312/2020) that evaluated deidentified data and involved no potential risk to patients. To avert any potential breach of confidentiality, no link between the patients and the researchers was made available.

This is an observational retrospective cohort study of patients with COVID-19 admitted to two COVID-designated ICUs at Ospedale Evangelico Internazionale and San Martino Policlinico Hospital in Genova, Italy, from March 6 th , 2020, to April 17 th , 2020. Figure 1 represents the consort flow diagram of this study.

We included critically ill patients with COVID-19, based upon a positive SARS-CoV-2 polymerase chain reaction assay, with available CT scan and/or LUS at ICU admission. All patients admitted to a COVID-19-ICUs were cared by a standard ICU care team (i.e., pre-COVID) and staffing ratios. No critical shortages in medications, ventilators, dialysis machines, or other critical care equipment were noticed.

Non-contrast chest CT scans were obtained using a 64 slice multi-detector CT (SOMATOM Sensation 64, Siemens Medical Solutions, Forchheim, Germany) and a 16 slice multi-detector CT (Lightspeed 16, GE Healthcare, Chicago (IL), USA), at ICU admission and around 10 days thereafter. Other spot CT examinations were performed in some patients depending on specific conditions. The overall number of included CT scans was 94.

COVID-19 chest CT in-hospital patterns were classified according to three main phenotypes recently proposed by Robba et al. (Robba et al. 2020 ) 1) multiple, focal, possibly overperfused ground-glass opacities with crazy paving appearance; 2) inhomogeneously distributed atelectasis and peribronchial opacities with predominant lung consolidations, and 3) a patchy, ARDS-like pattern". Then, lobar involvement of each lung lobe was assessed for percentage and classified as minimal (1-25%), mild (26-50%), moderate (51-75%), or severe (76-100%), with corresponding score as 1, 2, 3, or 4. The CT Total Severity Score (TSS) was obtained summing the five lobe scores (range from 0 to 20), . The classification and score of chest CT examinations was made by two radiologists with 5 and more than 20 years of experience in chest CT (DO and GB) blinded to patients past medical history, vital signs, symptoms, laboratory measurements and previous scan results. Disagreement was solved by consensus, and in case of persistent disagreement we considered the classification and score performed by the most experienced operator.

LUS examinations during ICU inpatient care were performed at ICU admission and at least every two days. The overall number of included LUS examinations was 362.

We followed a 12-zone protocol (Soummer et al. 2012; Mento et al. 2020) . Each intercostal space of upper and lower parts of the anterior, lateral, and posterior regions of the left and right chest wall were carefully examined, and findings (pleural effusion, confluent and isolated B-lines, irregular pleural line, consolidations) were recorded on 10-seconds video clips (Volpicelli et al. 2020a; Volpicelli et al. 2020b and Cantinotti et al. 2020; . For each of the 12zone a score from 0 to 3 was given according to the finding: irregular or isolated B-lines (1 point), confluent B-lines (2 points) and consolidations or pleural effusion (3 points). The total "Lung Ultrasonography (LUS) Score" was calculated by summing the scores of all 12 zones (range of possible scores, 0-36). Scores from 1 to 10 were rated as mild, 10-20 as moderate, and more than 20 as severe. The examinations were performed using a Mindray TE7 ultrasound system (Mindray biomedical electronic Co., Ltd, Shenzhen, China) and a Philips SparQ ® (Philips, Bothell, USA), equipped with a curvilinear array transducer (1.5-4.5 MHz) lung pre-set. During the examination, the physicians were blinded to the patient past medical history, vital signs, symptoms, laboratory measurements and previous scan results. The LUS examinations of ICU patients were performed by three radiologists and one intensivist, blinded for review, with 5, 2, 1 and 2 years of experience in LUS (GB, DO, TM and DB respectively) who rotated individually in the ICU (GB, DO, TM) or worked in the ICU (DB).

Patient data, including sociodemographic information, clinical data, and laboratory data, were obtained from the electronic medical record. Diagnostic imaging data were obtained from the RIS-PACS system. Data concerning LUS and chest CT scan at ICU admission and during the critical phase of the disease (around 10 days after ICU admission during the multi-systemic clinical syndrome with impaired/disproportionate and/or defective immunity) were collected (Turk et al. 2020) .

Data were collected until June 28 th , 2020. No sample size calculation was performed a priori for this exploratory and descriptive study. Baseline characteristics are presented as mean and standard deviation (SD) for continuous variables and count and proportions for categorical variables. Statistical analysis of the obtained data was performed using a Cohen's Kappa (κ) test to compare chest CT abnormal findings to LUS and chest X-ray abnormal findings. Data distribution of continuous variables was assessed by Shapiro-Wilk tests, and parametric and non-parametric tests were performed in according to the test results. Pearson's correlation tests were applied to test correlation between continuous variables and unpaired Student's t test (or Mann-Whitney U tests) were used to evaluate the passible differences in these variables between ICU-survivors and nonsurvivors. Fisher's exact test was used to compare the proportion of categorical variables between groups. Generalized linear regression models were used to assess the influence of each parameter and their combination on ICU-survival. Statistical significance was set at p value < 0.05.

Calculations were performed using R software v. 4.0.3 (R Core Team, Wien, Austria).

In the study period, 39 critically ill adults with SARS-CoV-2 infection were included (Fig. 1A) . The mean patient age was 63.2±13.4 (SD) years with 7 patients (18%) who were 75 years or older.

There were 9 females (23.1%) while the majority of patients were Caucasians (30 [76.9%]). Cardiovascular diseases were the most common comorbid condition (17 [43.6%]), followed by diabetes (15 [38.4%]). Nine patients (23.1%) had body mass index of 30 Kg/m 2 or greater. ICU admission occurred at 11.3 ± 8.9 days from symptoms onset; the PaO 2 /FiO 2 ratio was 204.8 ± 111.4 and the mean C-reactive protein (CPR) value was 139.5 ± 95.1 mg/L (upper limit of normal 10 mg/L). Demographics are fully reported in Table 1 .

All patients included in the study had recent chest X-rays, 83% of them had chest CT while 34% had LUS.

The subgroup analysis, performed in survivors and non-survivors, showed a nearly significant older age of non-survivors' group, with a significant higher incidence in the subgroup of patients over 75 years old (71% non survivors).

The mean ICU stay period was 18.3 ± 12.1 days, and patients who survived had significantly longer ICU-stay compared to non survivors (18.8 ± 9.0 vs 11.8 ± 5.6 days, p=0.001).

Thirty-two patients (82.0%) received invasive mechanical ventilation and 28 patients (72%) received at least one pronation cycle (Guèrin et al. 2020) .

Chest CT scans at ICU admission presented type 1 pattern in 24 patients (62%); type 2 pattern in 11 patients (28%); and type 3 pattern in 4 patients (10%). Among survivors, 20 patients (71%) presented with type 1 phenotype, 7 patients (25%) with type 2, and 1 patient (4%) with type 3; whereas among non-survivors, type 1, 2, 3 occurred in 4 (36%), 4 (36%), and 3 patients (27%), respectively, with a significant prevalence of type 1 pattern in the survivors' group (p<0.005).

At ICU admission, the mean CT TSS score was 11.6±3.7 and mean LUS score was 22.4±7.5. Initial ICU clinical findings, critical care interventions, and outcomes are summarized in Table 2 .

Chest CT scores and LUS scores significantly differed between survivors and non-survivors and in a multiple regression model considering all the elements at admission (CT, LUS, Age, BMI, PaO 2 /FiO 2 , c-reactive protein (CPR), interleukin-6 (IL6), Ferritin, gender, race and temperature) as 9 possible predictors of ICU-death, only chest CT and LUS demonstrated to influence significantly the outcome (p<0.0001).

The evolution over time of LUS score, CPR levels and PaO 2 /FiO 2 ratio were different between survivors and non-survivors. In fact, LUS and CPR showed an increasing trend among patients who died in ICU, while a decreasing trend was observed in those who survived (p=0.009 and p<0.0001, respectively). Conversely, during the ICU stay, the PaO 2 /FiO 2 ratio was reduced in non-survivors compared to survivors (p=0.003).

We found a significant direct correlation between chest CT and LUS scores at ICU admission (r=0.953, p<0.0001), a significant inverse correlation between admission CT TSS score, LUS score and PaO 2 /FiO 2 ratio (r=-0.375 and r=-0.345, with p=0.019 and p=0.031 respectively) and a significant direct correlation between chest CT at ICU admission with TSS score and CPR levels (r=0.329, p=0.041) but not between LUS score and CPR levels (r=0.266, p=0.102) (Figure 2,   Figure 3 ).

In our cohort of critically ill COVID-19 patients, we found that: 1) at ICU admission, the phenotype 1 was the most frequent at chest CT scan, followed by phenotype 2 and 3. Among survivors, we observed almost the same distribution of overall findings, with a significant prevalence of phenotype 1. No significant phenotype prevalence was seen among non-survivors; 2) chest CT and LUS scores differed between survivors and non-survivors both at ICU admission and at 10 days follow-up; 3) chest CT strongly correlated with LUS findings; 4) chest CT inversely correlated with PaO 2 /FiO 2 ratio, and CPR, while LUS PaO 2 /FiO 2 ratio, only; 5) in a multiple regression model, chest CT and LUS were the best parameters individually associated with clinical, laboratory, and ventilatory parameters.

One of the aims of the present study was to characterize and classify patients according to CT findings. In this setting, distinct phenotypes at chest CT scan were identified, as previously described (Robba et al. 2020; Gattinoni et al. 2020) . Phenotype 1/L with multiple focal possibly over perfused ground glass opacities was observed in 62% of cases; inhomogeneously distributed atelectasis and peri-bronchial opacities were observed in 28% (phenotype 2/L), while a patchy ARDS-like pattern was detected in 10% of cases (phenotype 3/H). In a previous report, Gattinoni et al. showed that phenotype 3/H was present in 20-30% of critically ill patients, which is slightly higher compared to our overall results . However, looking at the nonsurvivor's subgroup, we observed comparable results (27%).

Doubts remain regarding the existence of COVID-19 phenotypes (Tobin et al. 2020; Tonelli et al. 2021; Gattinoni et al. 2020b) . Despite variable respiratory system elastance, lungs recruitability, and clinical course, the current literature seems to agree that COVID-19 phenotypes are phases of a single disease, characterized by distinct host responses to the virus and different levels of lung damage Chiumello et al. 2020; Mauri et al. 2020) . However, postmortem findings have previously confirmed the existence of distinct chest CT phenotypes between COVID-19 survivors and non-survivors (Carsana et al. 2020; Robba et al. 2020; Pan et al. 2020; Jin et al. 2020) , in accordance with our findings.

To confirm the existence of distinct phenotypes as different phases of the same disease, in our cohort survivors showed residual lung alterations at chest CT and LUS at the time of ICU discharge. As previously described by Gaspardone et al. 2020 , these findings suggest a slow lung anatomic healing after ARDS due to severe viral pneumonia. This finding supports the concept that an appropriate follow-up of residual pulmonary lesions after ICU discharge should be pursued. In this setting, LUS could be of valuable help thanks to its easier application in a low-resource setting (Allinovi et al. 2020) .

Furthermore, our data confirm that the worst the CT pattern, the hardest the clinical course Turcato et al. 2020; Cartocci et al. 2020) , raising concerns about potential therapeutic strategies to be different in each phase of the disease (Battaglini et al. 2020; Robba et al. 2021 ).

Chest CT scan revealed a pivotal role in diagnosis and treatment of severe COVID-19 patients, possibly influencing individualized therapeutic strategies (Garg et al. 2021 ). However, chest CT scan may be difficult to be performed in critically ill patients due to possible clinical instability that makes contraindicated the transfer to the CT room. In this setting, others simpler tools like LUS have been proposed to better characterize COVID-19 patterns, with the limitation of less specificity and operator-dependency, but with the advantages of not exposing the patient to ionizing radiation and bedside availability Lu et al. 2020; Lanza et al. 2020) .

In our study, chest CT and LUS scores significantly differed between survivors and nonsurvivors confirming that both may be considered as essential tools for patients' outcome stratification at ICU admission and on follow-up (Feng et al 2020; Rojatti M et al. 2020; Pan F et al. 2020) . This has been previously confirmed by Pan et al. who found that persistent progression with predominant crazy-paving pattern was the major manifestation of COVID-19 in non-survivors. Several other studies investigated the role of chest CT and LUS for detecting patients at risk of death. A recent report found that higher LUS score could predict 30-days mortality (Borghesi et al. 2020) although this point is still debated (Colombi et al. 2020a ). On the contrary, qualitative and quantitative chest CT assessments associated with clinical data may be considered an optimal tool for the stratification of survival. In fact, pneumonia extent (greater than 40%) was detected as possible predictor of outcome in COVID-19 (Colombi et al. 2020b) .

In our cohort of COVID-19 critically ill patients, we observed a strong correlation between chest CT and LUS at ICU admission (r=0.953, p<0.0001), suggesting the possible utility of LUS technique in case of contraindications or impossibility to perform chest CT, and for daily clinical assessment of pneumonia progressions . Similarly, Volpicelli et al. found a strong correlation between LUS and CT imaging, thus confirming our results (Volpicelli et al. 2021 ).

However, other authors concluded that admission chest CT showed better performance than LUS for COVID-19 diagnosis, at varying disease prevalence (Colombi et al. 2020) . In fact, the main LUS limitation is related to its poor specificity, with findings overlapping with those from other pneumonia or lung pathological conditions (e.g., chronic heart failure or pulmonary fibrosis) (Tung-Chen et. al. 2020) .

Chest CT inversely correlated with admission PaO 2 /FiO 2 ratio (r=-0.375, p=0.019), and CPR (r=0.329, p=0.041), while LUS showed an inverse correlation with admission PaO 2 /FiO 2 ratio (r=-0.345, p=0.031) but did not correlate with CPR. These data suggest that the lower is the oxygenation, the worse is the chest CT and LUS profile confirming the association between clinical features and radiographic or ultrasonographic images Turcato et al. 2020; Cartocci et al 2020) . Similarly, in non-COVID-19 ARDS setting, LUS was adopted for detecting improvement in oxygenation, confirming that it could be promising for monitoring aeration at bedside, but no for predicting oxygenation response (Haddam et al. 2016) . Additionally, LUS can predict response to the prone position in awake non-intubated patients with COVID-19 associated ARDS (Avdeev et al. 2021) .

We acknowledge that we present a lower LUS -PaO 2 /FiO 2 ratio correlation if compared with other similar studies Soldati et al. 2020b ). However, this reflect our experience during the first strike of the pandemic and our results were obtained in critically ill mechanically ventilated COVID-19 patients with possible both pneumonia and ventilatory-related lung changes which are frequently not distinguishable. These findings are also consistent with other previously published series performed on comparable patients (Rojatti et al.; Millington et al. 2018; Dargent et al. 2020) .

This study has some limitations. Firstly, the retrospective design. Chest CT images were assessed basing on clinical necessity and patients' status, making impossible to perform regular imaging at the same time. Second, chest CT scans were performed at patient ICU admission and during ICU stay. However, being critically ill, many patients could not receive a chest CT scan during their acute phase due to clinical instability, limiting a possible transport to the CT room. In order to reduce this heterogeneity bias, most of the chest CT scans performed during ICU stay were assessed during a critical phase of the disease (during the multi-systemic clinical syndrome with impaired/disproportionate and/or defective immunity) corresponding to ten days after ICU admission, if deemed feasible. Third, chest CT scans were reviewed by experienced radiologists using the CT Total Severity Score (TSS). However, this method could be affected by interpretation bias; nonetheless, computer-aided quantitative analysis of the CT exam (Quantitative Computed Tomography [QCT]) was also recently used for this purpose, showing promising role in predicting COVID-19 clinical outcome (Lanza et al. 2020) . Fourth, others possible confounding factors were not considered in the multivariate analysis (i.e., multiple organ dysfunction) for data unavailability.

In critically ill mechanically ventilated COVID-19 patients, chest CT and LUS scores at ICU admission significantly differed between survivors and non-survivors. We also observed a significant correlation between admission clinical, oxygenation and imaging findings and between CT and LUS scores during patients' follow-up. In conclusion, although chest CT scan cannot be replaced by LUS in the diagnostic process, LUS should be considered as a valuable complement to diagnosis and follow-up in ICU setting, since it has many advantages such as convenient, fast, noninvasive, readily repeatable, and does not involve ionizing radiation. Tables   Table 1 Baseline demographic characteristics and comorbidities in intensive care unit (ICU) survivors and non-survivors. Data are expressed in percentages or in mean and standard deviation (SD).

[ and lung ultrasound (LUS) scores, divided between survivors and non-survivors at Intensive care unit (ICU) admission and 10 days follow-up. Data are expressed in mean and standard deviation (SD).

Overall ( 

The bio-mission of interleukin-6 in the pathogenesis of COVID-19: A brief look at potential therapeutic tactics

Impact of coronavirus disease 2019 (COVID-19) emergency on Italian radiologists: a national survey

Ultrasound-Guided Interventions During the COVID-19 Pandemic-A New Challenge

Lung Ultrasound May Support Diagnosis and Monitoring of COVID-19 Pneumonia

ICU and Ventilator Mortality Among Critically Ill Adults With Coronavirus Disease 2019. Emory COVID-19 Quality and Clinical Research Collaborative

In Reply: The Coronavirus Disease 2019 Global Pandemic: A Neurosurgical Treatment Algorithm

GECOVID (GEnoa COVID-19) group. Computed tomography assessment of PEEPinduced alveolar recruitment in patients with severe COVID-19 pneumonia

Chest Radiograph Findings in Asymptomatic and Minimally Symptomatic Quarantined Patients in Codogno, Italy during COVID-19 Pandemic

Chest X-ray severity index as a predictor of in-hospital mortality in coronavirus disease 2019: A study of 302 patients from Italy

Emerging therapies for COVID-19 pneumonia

Prognostic Value of a New Lung Ultrasound Score to Predict Intensive Care Unit Stay in Pediatric Cardiac Surgery

Pulmonary post-mortem findings in a series of COVID-19 cases from northern Italy: a two-centre descriptive study

Chest CT for early detection and management of coronavirus disease (COVID-19): a report of 314 patients admitted to Emergency Department with suspected pneumonia

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: A descriptive study

Physiological and quantitative CT-scan characterization of COVID-19 and typical ARDS: a matched cohort study

Comparison of admission chest computed tomography and lung ultrasound performance for diagnosis of COVID-19 pneumonia in populations with different disease prevalence

Qualitative and quantitative chest CT parameters as predictors of specific mortality in COVID-19 patients

Lung ultrasound score to monitor COVID-19 pneumonia progression in patients with ARDS

Acute respiratory distress syndrome advances in diagnosis and treatment

Early prediction of disease progression in COVID-19 pneumonia patients with chest CT and clinical characteristics

Diagnostic accuracy of CT and radiographic findings for novel coronavirus 2019 pneumonia: Systematic review and meta-analysis

Lung Ultrasound in COVID-19 A Role Beyond the Acute Phase?

COVID-19 pneumonia: different respiratory treatments for different phenotypes?

COVID-19 phenotypes: leading or misleading?

Prone Positioning in Severe Acute Respiratory Distress Syndrome

Lung ultrasonography for assessment of oxygenation response to prone position ventilation in ARDS

Chest CT findings related to mortality of patients with COVID-19: A retrospective case-series study

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

A Pattern Categorization of CT Findings to Predict Outcome of COVID-19 Pneumonia. Front Public Health

Continuous Renal Replacement Therapy for a Patient with Severe COVID-19

Quantitative chest CT analysis in COVID-19 to predict the need for oxygenation support and intubation

CT image visual quantitative evaluation and clinical classification of coronavirus disease (COVID-19)

A Clinical Study of Noninvasive Assessment of Lung Lesions in Patients with Coronavirus Disease-19 (COVID-19) by Bedside Ultrasound

Potential for Lung Recruitment and Ventilation-Perfusion Mismatch in Patients With the Acute Respiratory Distress Syndrome From Coronavirus Disease

On the Impact of Different Lung Ultrasound Imaging Protocols in the Evaluation of Patients Affected by Coronavirus Disease 2019: How Many Acquisitions Are Needed?

Expert Agreement in the Interpretation of Lung Ultrasound Studies Performed on Mechanically Ventilated Patients

Different computed tomography patterns of Coronavirus Disease 2019 (COVID-19) between survivors and non-survivors

A New Lung Ultrasound Protocol Able to Predict Worsening in Patients Affected by Severe Acute Respiratory Syndrome Coronavirus 2 Pneumonia

Distinct phenotypes require distinct respiratory management strategies in severe COVID-19

Ten things you need to know about intensive care unit management of mechanically ventilated patients with COVID-19

Lung Ultrasound and Respiratory Pathophysiology in Mechanically Ventilated COVID-19 Patients-an Observational Trial

Is There a Role for Lung Ultrasound During the COVID-19 Pandemic?

Proposal for International Standardization of the Use of Lung Ultrasound for Patients With COVID-19: A Simple, Quantitative, Reproducible Method

Correlation Between Chest CT Findings and Clinical Features of 211 COVID-19 Suspected Patients in Wuhan

Does Making a Diagnosis of ARDS in Patients With Coronavirus Disease

Spontaneous Breathing and Evolving Phenotypes of Lung Damage in Patients with COVID-19: Review of Current Evidence and Forecast of a New Scenario

Correlation between Chest Computed Tomography and Lung Ultrasonography in Patients with Coronavirus Disease 2019 (COVID-19)

Correlation between arterial blood gas and CT volumetry in patients with SARS-CoV-2 in the emergency department

Three critical clinicobiological phases of the human SARS-associated coronavirus infections

International evidence-based recommendations for point-of-care lung ultrasound

Sonographic signs and patterns of COVID-19 pneumonia

What's new in lung ultrasound during the COVID-19 pandemic

Descriptive analysis of a comparison between lung ultrasound and chest radiography in patients suspected of COVID-19

Clinical Characteristics of 138 Hospitalized Patients with

Novel Coronavirus-Infected Pneumonia in Wuhan, China

World Health Organization Director-General's remarks at the media briefing on

Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease

IL-6, interleukin-6

oxygen; FiO2, fraction of inspired oxygen; LUS, lung ultrasound; CT, computed tomography; SD, standard deviation; * t test] Characteristics Overall (n=39 ) ICU survivors (n=28) ICU non-survivors (n=11) P value ICU Admission

The authors received no specific funding for this work.

All authors have no conflict of interest to disclose.