original un-edited accepted proof€¦ · keywords: sars-cov-2, chest computed tomography, lung...

Intensive Care Medicine ORIGINAL Un-edited accepted proof

Zieleskiewicz et al. Comparative study of lung ultrasound and chest computed tomography scan in the assessment of severity of confirmed COVID-19 pneumonia.

Intensive Care Medicine (2020). DOI: 10.1007/s00134-020-0

1

DOI: 10.1007/s00134-020-0

Title Comparative study of lung ultrasound and chest computed tomography scan in the assessment of severity of confirmed COVID-19 pneumonia

Authors’ names Laurent Zieleskiewicz,1,2 Thibaut Markarian,3 Alexandre Lopez,1 Chloé Taguet,3 Neyla

Mohammedi,1 Mohamed Boucekine,4 Karine Baumstarck,4 Guillaume Besch,5 Gautier

Mathon,6 Gary Duclos,1 Lionel Bouvet,7,8,9 Pierre Michelet,3 Bernard Allaouchiche,6,8,9

Kathia Chaumoître,10 Mathieu Di Bisceglie,10 Marc Leone1, for the AZUREA Network

Address for each

author

1. Aix Marseille University, Assistance Publique Hôpitaux de Marseille, Department of

Anesthesiology and Intensive Care Medicine, Hôpital Nord, Marseille, 13015, France

2. Center for Cardiovascular and Nutrition Research (C2VN), Aix Marseille Université,

INSERM, INRA, Marseille, 13005, France

3. Department of Emergency Medicine, Timone University Hospital, Marseille, France

4. Centre d’Etudes et de Recherches sur les Services de Santé et Qualité, Faculté de

Médecine, Aix-Marseille Université, Marseille, 13005, France

5. Department of Anesthesiology and Intensive Care Medicine, University Hospital of

Besancon, Besancon, France 2. EA3920, University of Franche-Comte, Besancon,

France

6. Hospices Civils de Lyon, Centre Hospitalier Lyon-Sud, Service de Réanimation, F-

69310, Pierre-Bénite, France

7. Service Anesthésie Réanimation, Groupement Hospitalier Est, Hospices Civils de

Lyon, Lyon, France

8. Université Claude Bernard, Lyon1

9. Université de Lyon, VetAgro Sup, Campus Vétérinaire de Lyon, UPSP 2016.A101,

Pulmonary and Cardiovascular Aggresion in Sepsis, F-69280, Marcy l’Étoile, France

10. Aix Marseille University, Assistance Publique Hôpitaux de Marseille, Service

d’Imagerie Médicale, Hôpital Nord, Marseille, 13015, France

Authors’ name

and positions

Laurent Zieleskiewicz,* Thibaut Markarian,* Alexandre Lopez, Chloé Taguet, Neyla

Mohammedi, Mohamed Boucekine, Karine Baumstarck, Guillaume Besch, Gautier

Mathon, Gary Duclos, Lionel Bouvet, Pierre Michelet, Bernard Allaouchiche, Kathia

Chaumoître, Mathieu Di Bisceglie, Marc Leone

* LZ and TM are joint first authors and contributed equally.

Corresponding

author

Zieleskiewicz Laurent, MD, Aix Marseille University, Assistance Publique Hôpitaux de

Marseille, Department of Anaesthesiology and Intensive Care, Hôpital Nord, Marseille

France ([email protected])




2

Authors’ contact

Laurent Zieleskiewicz, MD ([email protected]); Thibaut Markarian, MD (thibaut.markarian@ap-

hm.fr); Alexandre Lopez, MD ([email protected]); Chloé Taguet, MD ([email protected]); Neyla

Mohammedi, MD ([email protected]); Mohamed Boucekine, PhD ([email protected]);

Karine Baumstarck, MD, PhD ([email protected]); Guillaume Besch, MD, PhD (gbesch@chu-

besancon.fr); Gautier Mathon, MD ([email protected]); Gary Duclos, MD ([email protected]);

Lionel Bouvet, MD, PhD ([email protected]); Pierre Michelet, MD, PhD ([email protected]);

Bernard Allaouchiche, MD, PhD ([email protected]); Kathia Chaumoître, MD, PhD

([email protected]); Mathieu Di Bisceglie, MD ([email protected]); Marc Leone, MD,

PhD ([email protected])

Author contributions

The corresponding author (Laurent Zieleskiewicz) attests that all listed authors meet the authorship criteria and

that no others meeting the criteria have been omitted.

Name: Laurent Zieleskiewicz

Contributions: This author helped design and conduct the study, analyze the data and write the manuscript.

Name: Thibaut Markarian

Contributions: This author helped design and conduct the study, analyze the data and write the manuscript.

Name: Alexandre Lopez

Contributions: This author helped conduct the study, analyze the data and write the manuscript.

Name: Chloé Taguet

Contributions: This author helped conduct the study.

Name: Neyla Mohammedi


Name: Mohamed Boucekine

Contributions: This author helped analyze the data and write the manuscript.

Name: Karine Baumstarck

Contributions: This author helped analyze the data and write the manuscript.

Name: Guillaume Besch

Contributions: This author helped conduct the study and write the manuscript.

Name: Gautier Mathon


Name: Gary Duclos


Name: Lionel Bouvet

Contributions: This author helped write the manuscript.




3

Name: Pierre Michelet


Name: Bernard Allaouchiche


Name: Kathia Chaumoître


Name: Mathieu Di Bisceglie


Name: Marc Leone

Contributions: This author helped design the study and write the manuscript.

Conflicts of interest

LZ and TM received fees for teaching ultrasound to GE healthcare customers.

License for publication

The corresponding author (Laurent Zieleskiewicz) has the right to grant on behalf of all authors and does grant

on behalf of all authors a worldwide license to the publishers and its licensees in perpetuity and in all forms,

formats and media (whether known now or created in the future) to i) publish, reproduce, distribute, display and

store the contribution; ii) translate the contribution into other languages, create adaptations, reprints, include

within collections and create summaries, extracts and/or abstracts of the contribution; iii) create any other

derivative work(s) based on the contribution; iv) exploit all subsidiary rights in the contribution; v) allow the

inclusion of electronic links from the contribution in third party material wherever it may be located; and vi)

license any third party to do any or all of the above.

Transparency declaration

Laurent Zieleskiewicz affirms 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 (and, if relevant, registered) have been explained.




4

Abstract

Purpose

The relationship between lung ultrasound (LUS) and chest computed tomography (CT) scans in patients with

severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pneumonia is not clearly defined. The primary

objective of our study was to assess the performance of LUS in determining severity of SARS-CoV-2

pneumonia compared with chest CT scan. Secondary objectives were to test the association between LUS score

and location of the patient, use of mechanical ventilation, and the pulse oximetry (SpO2)/fractional inspired

oxygen (FiO2) ratio.

Methods

A multicentre observational study was performed between 15 March and 20 April 2020. Patients in the

Emergency Department (ED) or Intensive Care Unit (ICU) with acute dyspnoea who were PCR-positive for

SARS-CoV-2, and who had LUS and chest CT performed within a 24-hour period, were included.

Results

One hundred patients were included. LUS score was significantly associated with pneumonia severity assessed

by chest CT and clinical features. The AUC of the ROC curve of the relationship of LUS versus chest CT for the

assessment of severe SARS-CoV-2 pneumonia was 0.78 (CI 95% 0.68–0.87; p < 0.0001). A high LUS score was

associated with the use of mechanical ventilation, and with a SpO2/FiO2 ratio below 357.

Conclusion

In known SARS-CoV-2 pneumonia patients, the LUS score was predictive of pneumonia severity as assessed by

a chest CT scan and clinical features. Within the limitations inherent to our study design, LUS can be used to

assess SARS-CoV-2 pneumonia severity.

Keywords: SARS-CoV-2, chest Computed Tomography, lung ultrasound, diagnostic accuracy

Take home message

Lung ultrasound is a possible alternative to chest CT-scan, especially in resource-constrained environment, for

the diagnosis of SARS-CoV-2 pneumonia severity. Given its high availability at the bedside, it may be widely

used during the pandemic.

Abbreviations

AUC: Area under curve

COVID-19: Coronavirus disease 2019

CI: Confidence interval

CT: Computed tomography

DA: Diagnostic accuracy

ED: Emergency department

ESM: Electronical Supplemental Material

FiO2: Fractional inspired oxygen




5

ICU: Intensive care unit

LUS: Lung ultrasound

NPV: Negative predictive value

PaO2: Arterial oxygen partial

PCR: Polymerase chain reaction

PPV: Positive predictive value

ROC: Receiver operating characteristic

SARS-CoV-2: Severe acute respiratory syndrome coronavirus 2

Se: Sensitivity

Sp: Specificity

SpO2: Pulse oximetry




6

Introduction

Severe acute respiratory syndrome from coronavirus 2 (SARS-CoV-2) can be identified by a chest computed

tomography (CT) scan [1-2], typically showing ground-glass opacities, consolidation and interlobular septal

thickening [3]. CT scanning has recently been shown to predict prognosis in these patients [4]. However, the use

of chest CT scan is associated with adverse events and increased resource consumption [5-6], and especially

during this pandemic the safety of transferring patients for CT scan remains uncertain at both individual and

collective levels [7].

Lung ultrasound (LUS) has a high diagnostic accuracy for interstitial syndrome and alveolar consolidation,

which is superior to chest radiograph [8-9]. LUS has been recommended for the diagnosis and management of

pneumonia [10], even during a previous viral pandemic [11]. During the current coronavirus disease 2019

(COVID-19) pandemic, LUS might reduce in-hospital transfers, the exposure of healthcare workers, and the risk

of contamination of medical devices [12].

The main pattern of SARS-CoV-2 pneumonia using LUS is interstitial syndrome [13]. However, pneumonia

may show different features using different diagnostic methods [3,14]. To date, there has been case report [15],

but few studies, comparing LUS with chest CT scan in patients with SARS-CoV-2 pneumonia.

The primary objective of our study was to assess the performance of LUS in determining COVID-19 pneumonia

severity as assessed by chest CT scan. The secondary objectives were to test the association between LUS score

and clinical features including the location of the patient, mechanical ventilation and pulse oximetry

(SpO2)/fractional inspired oxygen (FiO2) ratio.

Method

Design

This multicentre observational study was conducted in four university hospitals. We performed a retrospective

analysis comparing the result of LUS examinations and chest CT, performed as part of routine care, in a

convenience sample of patients in either the ED or ICU. The ED and ICUs were located in different institutions.

Ethical considerations

The study was approved by the Committee for Research Ethics of the French Society of Anesthesia and

Intensive Care Medicine (CERAR IRB00010254-2020-062). In accordance with French law, patients were

informed regarding the use of their data for publication [16].

Population

We included adult patients admitted to the ED or ICU with confirmed SARS-CoV-2 infection, diagnosed by

acute dyspnoea (SpO2 < 94% and/or breathlessness [17]) together with a positive polymerase chain reaction




7

(PCR) test in a nasopharyngeal or bronchoalveolar sample, who had a LUS exam at admission as well as a chest

CT within the 24 h following the LUS.

Clinical features

At inclusion, each patient had a standard medical history and examination, monitoring of heart and respiratory

rate, blood pressure and SpO2, and arterial blood gas analysis.

For patients who were breathing spontaneously, FiO2 was calculated as follows: FiO2 = (21+ 3 * oxygen flow

(L/min)/100) [18].

For the purposes of analysis, we defined a low SpO2/FiO2 ratio as < 357 and high SpO2/FiO2 ratio as ≥ 357.

This cut-off was chosen as it is equivalent to a PaO2/FiO2 ratio of 300 mmHg [19].

LUS and chest CT examination

LUS was performed within the first 2 hours after admission. LUS was performed by imaging 12 lung regions,

modified for critical illness. Thus, we imaged the posterior areas behind the posterior axillary line rather than in

the paravertebral areas to avoid turning completely the patient [20]. LUS examinations were performed by

emergency physicians or intensivists in charge of the patient [21-22]. The skill of the operators was rated as

followed [21]:

- level 3 operator; LUS academic teacher with several publications in the field

- level 2 operator; more than 25 supervised procedures and 200 non-supervised procedures

- level 1 operator; at least 25 supervised procedures and less than 200 non-supervised procedures

Chest CT was performed at an appropriate time during the clinical course using a 128-slice CT (OPTIMA

CT660, GE Healthcare, Chicago, Illinois, US) in the supine position, with the patient instructed to hold their

breath after a deep inspiration. Most CT scans were non-contrast, low-dose chest CT.

Further details of the diagnostic procedures are available in the Electronic Supplemental Material (ESM 1).

Sample size considerations

The sample size calculation was performed to determine whether an area under the curve (AUC) of ≥ 0.80 was

achieved for a receiver operator characteristic (ROC) plot of LUS versus chest CT scan. Based on unpublished

data, with a precision of 10% and an expected proportion of severe pneumonia on chest CT scan of 40%, the

sample size required was 87. Taking into account the potential for incomplete data from LUS or chest CT scans,

we included 100 patients.

Statistical analysis

The characteristics of the patients are summarized as medians and interquartile ranges for continuous variables,

and as numbers and percentages for qualitative variables. Comparisons of patients’ characteristics, CT and




8

ultrasound parameters were performed between patients managed in the ED versus the ICU. The LUS score was

compared between the three severity grades (mild, moderate, and severe pneumonia) on CT scan using an

ANOVA test. The receiver operator characteristic (ROC) curve and AUC estimates were determined for the

relationship of LUS score and CT scan to diagnose severity of pneumonia. The optimal threshold for best

discrimination between non-severe and severe pneumonia was calculated using the Youden index. Sensitivity

(Se), specificity (Sp), negative predictive value (NPV), positive predictive value (PPV) and DA are provided

with their 95% confidence intervals (CIs). A grey zone represents a predictive test of low accuracy, that is, the

Se and Sp are both < 90% [23]. Se and Sp curves were constructed to calculate the grey zone for an LUS score

that was inconclusive for predicting severe pneumonia [24]. ROC and AUC, the optimal thresholds, Se, Sp,

NPV, PPV and DA were determined for LUS score to diagnose consolidation, interstitial syndrome, pleural

effusion and pleural irregularity according to the chest CT findings and use of mechanical ventilation. Mean

LUS scores were compared between low and high SpO2/FiO2 ratios. For all calculations, R software (R

Development Core Team) and SPSS Statistics for Windows, Version 20.0 (IBM, Armonk, NY) were used. The

significance level was set at p < 0.05.

Results

Patient features

Figure 1 presents the flowchart of the study. Among the 412 patients presenting with suspected SARS-CoV-2 in

the ED, 207 were symptomatic with dyspnea. Of these symptomatic patients, 77 underwent LUS exam followed

by a chest CT scan within the next 24 hours and were included in the ED group. Among the 87 patients admitted

to the ICU during this period, 23 fulfilled the inclusion criteria and were included in the ICU group. Thus 100

patients (200 hemi-thoraces) were analysed. The demographic and clinical data of the patients are summarized in

Table 1.

Descriptive analysis

LUS was performed by Level 3, Level 2 and Level 1 operators in 37%, 53% and 10% of cases, respectively. A

descriptive analysis of chest CT scan and LUS findings is reported in Table 2.

Primary outcome

The LUS score was significantly associated with chest CT scan severity (p < 0.0001; Table 3). The AUC of the

ROC curve of the relationship of LUS versus chest CT for the assessment of severe SARS-CoV-2 pneumonia

was 0.78 (CI 95% 0.68–0.87; p < 0.0001; Figure 2a). The maximum Youden index was 0.50. The LUS score

grey zone (score of 13–23) with inconclusive values included 38 patients (38%) (Figure 2b). An LUS score > 23

predicted severe SARS-CoV-2 pneumonia diagnosed by chest CT scan with a Sp > 90% and a PPV of 70% in 23

patients. An LUS score < 13 excluded severe SARS-CoV-2 pneumonia diagnosed by chest CT scan with a Se >

90% and an NPV of 92% in 39 patients. Thirty-eight patients (38%) were in the grey zone.

Secondary outcomes




9

The Se, Sp, PPV, NPV and DA of LUS for the chest CT scan diagnosis of alveolar consolidation, interstitial

syndrome, pneumothorax, pleural effusion and pleural irregularity are reported in ESM Table S1. LUS score was

associated with clinical status at the time of examination. The LUS score was significantly higher in the

mechanically ventilated patients (28 ± 5 vs 14 ± 8; p < 0.0001). All the mechanically ventilated patients had an

LUS score > 19. The ROC curve of LUS for mechanical ventilation was 0.92. The LUS score was significantly

higher in the patients with an SpO2/FiO2 ratio < 357 (19 ± 8 vs 11 ± 8; p < 0.0001).

Discussion

Our results show that the severity of SARS-CoV-2 pneumonia assessed by LUS is highly associated with

severity as assessed by chest CT scan. Thus, LUS could replace chest CT scan for the initial assessment of lung

involvement in most of confirmed symptomatic SARS-CoV-2 patients.

A chest CT scan is the gold standard to assess the severity of pneumonia in patients with SARS-CoV-2 [25]. An

international consensus statement concluded that ‘In a resource-constrained environment, imaging is indicated

for medical triage of patients with suspected SARS-CoV-2 who present with moderate-severe clinical features

and a high pre-test probability of disease’ [26]. However, there are risks and logistical limitations to the use of

CT. The transfer of unwell patients risks adverse events [5-6], there is increased potential exposure to the virus

of all the healthcare providers involved in the transfer, and there may be a lack of scanning capacity if there is a

high number of patients presenting. Unlike CT scan, LUS has the advantages that it is available at the point of

care, it may be performed by suitably skilled physicians in the ED or ICU, and there is negligible cost to each

individual examination [27]. The same trends were previously reported for the H1N1 pandemic [11], and our

results confirm similar utility of LUS to assess the severity of SARS-CoV-2 pneumonia. In addition, due to the

high incidence of thromboembolic events [28-29], LUS should be incorporated into multiorgan ultrasound

assessment to detect both venous thrombosis and [30-31] signs of acute right heart failure [32].

The discrepancy between LUS and CT scan results lies in the middle region, or grey zone, of LUS scores. From

these results, we suggest that chest CT scan would not be required if an initial LUS examination found a score <

13 (mild disease) or > 23 (severe disease), but only in the 38% of cases with an intermediate score between 13 –

23. Furthermore, LUS exams were not performed exclusively by expert operators, and it might be assumed that

improving the level of expertise would have improved the accuracy. In contrast, chest CT scans were all

interpreted by senior radiologists with a high level of expertise.

LUS scores differed according to clinical status of our highly select population. Patients in the ICU had a higher

SpO2/FiO2 ratio ≥ 357. Furthermore, all the patients having mechanical ventilation had an LUS score > 19.

Our study has several limitations. First, we included a convenience sample of patients with dyspnoea and

confirmed PCR, which represents only a small percentage of the many cases of pneumonia observed during the

pandemic. The chest CT scan and LUS were not performed simultaneously, but the median delay between chest




10

CT scan and LUS was four hours. As previously described [20], we used a modified LUS technique that did not

scan the paravertebral areas. If this introduced any alteration in the diagnostic accuracy, it would be expected to

have reduced identification of pulmonary consolidation, and hence the assessed degree of severity.

We wished to ensure that patients who were studied had both significant and proven SARS-CoV-2 infection. The

delay in getting test results has led to situations where CT scan is being used as the primary investigation to

diagnose and triage patients. As a positive PCR test was one of the inclusion criteria for our analysis, we cannot

comment on whether our findings would apply to patients who had an expedited scan on this basis.

We suggest that further research should assess the generalisability of the low and high thresholds suggested

above for replacement of CT scan by LUS as a tool to assess pneumonia severity, and the use of LUS as a

predictor of the clinical course and a guide to ongoing management [33-34].

Conclusion

In patients with proven SARS-CoV-2 pneumonia, LUS score is associated with severity as assessed by chest CT

scan and clinical features.

Acknowledgements

We wish to thank Mike Kinsella for his help in editing the manuscript. We thank Eloïse Cercueil for her help in

patient recruitment.

Role of the funder/sponsor: None.

References

1. Li Y, Xia L (2020) Coronavirus Disease 2019 (COVID-19): Role of Chest CT in Diagnosis and

Management. AJR Am J Roentgenol 1–7. Https://doi.org/10.2214/AJR.20.22954

2. Ai T, Yang Z, Hou H, et al (2020) Correlation of Chest CT and RT-PCR Testing in Coronavirus

Disease 2019 (COVID-19) in China: A Report of 1014 Cases. Radiology 200642.

Https://doi.org/10.1148/radiol.2020200642

3. Wu J, Wu X, Zeng W, et al (2020) Chest CT Findings in Patients With Coronavirus Disease 2019 and

Its Relationship With Clinical Features. Invest Radiol 55:257–261.

Https://doi.org/10.1097/RLI.0000000000000670

4. Mahdjoub E, Mohammad W, Lefevre T, et al (2020) Admission chest CT score predicts 5-day outcome

in patients with COVID-19. Intensive Care Med. Https://doi.org/10.1007/s00134-020-06118-y

5. Aliaga M, Forel J-M, De Bourmont S, et al (2015) Diagnostic yield and safety of CT scans in ICU.

Intensive Care Med 41:436–443. Https://doi.org/10.1007/s00134-014-3592-1

6. Jia L, Wang H, Gao Y, et al (2016) High incidence of adverse events during intra-hospital transport of

critically ill patients and new related risk factors: a prospective, multicenter study in China. Crit Care 20:12.

Https://doi.org/10.1186/s13054-016-1183-y




11

7. Hope MD, Raptis CA, Shah A, et al (2020) A role for CT in COVID-19? What data really tell us so far.

Lancet 395:1189–1190. Https://doi.org/10.1016/S0140-6736(20)30728-5

8. Lichtenstein D, Goldstein I, Mourgeon E, et al (2004) Comparative diagnostic performances of

auscultation, chest radiography, and lung ultrasonography in acute respiratory distress syndrome. Anesthesiology

100:9–15. Https://doi.org/10.1097/00000542-200401000-00006

9. Xirouchaki N, Magkanas E, Vaporidi K, et al (2011) Lung ultrasound in critically ill patients:

comparison with bedside chest radiography. Intensive Care Med 37:1488–1493. Https://doi.org/10.1007/s00134-

011-2317-y

10. Staub LJ, Mazzali Biscaro RR, Kaszubowski E, Maurici R (2019) Lung Ultrasound for the Emergency

Diagnosis of Pneumonia, Acute Heart Failure, and Exacerbations of Chronic Obstructive Pulmonary

Disease/Asthma in Adults: A Systematic Review and Meta-analysis. J Emerg Med 56:53–69.

Https://doi.org/10.1016/j.jemermed.2018.09.009

11. Testa A, Soldati G, Copetti R, et al (2012) Early recognition of the 2009 pandemic influenza A (H1N1)

pneumonia by chest ultrasound. Crit Care 16:R30. Https://doi.org/10.1186/cc11201

12. Buonsenso D, Pata D, Chiaretti A (2020) COVID-19 outbreak: less stethoscope, more ultrasound.

Lancet Respir Med. Https://doi.org/10.1016/S2213-2600(20)30120-X

13. Lomoro P, Verde F, Zerboni F, et al (2020) COVID-19 pneumonia manifestations at the admission on

chest ultrasound, radiographs, and CT: single-center study and comprehensive radiologic literature review. Eur J

Radiol Open 7:100231. Https://doi.org/10.1016/j.ejro.2020.100231

14. Peng Q-Y, Wang X-T, Zhang L-N, Chinese Critical Care Ultrasound Study Group (CCUSG) (2020)

Findings of lung ultrasonography of novel corona virus pneumonia during the 2019–2020 epidemic. Intensive

Care Med. Https://doi.org/10.1007/s00134-020-05996-6

15. Duclos G, Lopez A, Leone M, Zieleskiewicz L (2020) “No dose” lung ultrasound correlation with “low

dose” CT scan for early diagnosis of SARS-cov-2 pneumonia. Intensive Care Med.

Https://doi.org/10.1007/s00134-020-06058-7

16. Toulouse E, Masseguin C, Lafont B, et al (2018) French legal approach to clinical research. Anaesth

Crit Care Pain Med 37:607–614. Https://doi.org/10.1016/j.accpm.2018.10.013

17. Malas O, Cağlayan B, Fidan A, et al (2003) Cardiac or pulmonary dyspnea in patients admitted to the

emergency department. Respir Med 97:1277–1281. Https://doi.org/10.1016/j.rmed.2003.07.002

18. Frat J-P, Thille AW, Mercat A, et al (2015) High-flow oxygen through nasal cannula in acute

hypoxemic respiratory failure. N Engl J Med 372:2185–2196. Https://doi.org/10.1056/nejmoa1503326

19. Pandharipande PP, Shintani AK, Hagerman HE, et al (2009) Derivation and validation of spo2/fio2

ratio to impute for pao2/fio2 ratio in the respiratory component of the Sequential Organ Failure Assessment

(SOFA) Score. Crit Care Med 37:1317–1321. Https://doi.org/10.1097/CCM.0b013e31819cefa9

20. Mojoli F, Bouhemad B, Mongodi S, Lichtenstein D (2018) Lung Ultrasound for Critically Ill Patients.

Am J Respir Crit Care Med 199:701–714. Https://doi.org/10.1164/rccm.201802-0236CI

21. Arbelot C, Neto FLD, Gao Y, et al (2020) Lung Ultrasound in Emergency and Critically Ill

patientsnumber of Supervised Exams to Reach Basic Competence. Anesthesiology 132:899–907.




12

Https://doi.org/10.1097/ALN.0000000000003096

22. Volpicelli G, Elbarbary M, Blaivas M, et al (2012) International evidence-based recommendations for

point-of-care lung ultrasound. Intensive Care Med 38:577–591. Https://doi.org/10.1007/s00134-012-2513-4

23. Ray P, Le Manach Y, Riou B, Houle TT (2010) Statistical evaluation of a biomarker. Anesthesiology

112:1023–1040. Https://doi.org/10.1097/ALN.0b013e3181d47604

24. Coste J, Pouchot J (2003) A grey zone for quantitative diagnostic and screening tests. Int J Epidemiol

32:304–313. Https://doi.org/10.1093/ije/dyg054

25. Zhao W, Zhong Z, Xie X, et al (2020) Relation Between Chest CT Findings and Clinical Conditions of

Coronavirus Disease (COVID-19) Pneumonia: A Multicenter Study. American Journal of Roentgenology

214:1072–1077. Https://doi.org/10.2214/AJR.20.22976

26. Rubin GD, Ryerson CJ, Haramati LB, et al (2020) The Role of Chest Imaging in Patient Management

During the COVID-19 Pandemic: A Multinational Consensus Statement From the Fleischner Society. Chest.

Https://doi.org/10.1016/j.chest.2020.04.003

27. Mayo PH, Copetti R, Feller-Kopman D, et al (2019) Thoracic ultrasonography: a narrative review.

Intensive Care Med 45:1200–1211. Https://doi.org/10.1007/s00134-019-05725-8

28. Helms J, Tacquard C, Severac F, et al (2020) High risk of thrombosis in patients with severe SARS-

cov-2 infection: a multicenter prospective cohort study. Intensive Care Med 46:1089–1098.

Https://doi.org/10.1007/s00134-020-06062-x

29. Tavazzi G, Civardi L, Caneva L, et al (2020) Thrombotic events in SARS-cov-2 patients: an urgent call

for ultrasound screening. Intensive Care Med. Https://doi.org/10.1007/s00134-020-06040-3

30. Koenig S, Chandra S, Alaverdian A, et al (2014) Ultrasound assessment of pulmonary embolism in

patients receiving CT pulmonary angiography. Chest 145:818–823. Https://doi.org/10.1378/chest.13-0797

31. Nazerian P, Volpicelli G, Gigli C, et al (2017) Diagnostic Performance of Wells Score Combined With

Point-of-care Lung and Venous Ultrasound in Suspected Pulmonary Embolism. Acad Emerg Med 24:270–280.

Https://doi.org/10.1111/acem.13130

32. Creel-Bulos C, Hockstein M, Amin N, et al (2020) Acute Cor Pulmonale in Critically Ill Patients with

Covid-19. New England Journal of Medicine 0:e70. Https://doi.org/10.1056/nejmc2010459

33. Volpicelli G, Lamorte A, Villén T (2020) What’s new in lung ultrasound during the COVID-19

pandemic. Intensive Care Med. Https://doi.org/10.1007/s00134-020-06048-9

34. Markarian T, Zieleskiewicz L, Perrin G, et al (2019) A lung ultrasound score for early triage of elderly

patients with acute dyspnea. CJEM 21:399–405. Https://doi.org/10.1017/cem.2018.483




13

Table 1: Baseline clinical data

Patient Information All

n = 100

ED

n = 77

ICU

n = 23 P-value

Characteristics

Age, median (IQR), years 61 (54–77) 60 (51–78) 69 (55–73) 0.49

Sex

Men 65 (65) 45 (58) 20 (87) 0.01

Women 35 (35) 32 (42) 3 (13)

BMI > 30 kg/m 17 (17) 11 (14) 6 (26) 0.21

Co-morbidities

Hypertension 24 (24) 12 (16) 12 (52) 0.0003

Coronary disease 11 (11) 5 (7) 6 (26) 0.02

Heart failure 16 (16) 13 (17) 3 (13) 0.66

COPD 10 (10) 8 (10) 2 (9) 0.81

Chronic kidney disease 2 (2) 1 (1) 1 (4) 0.41

Liver disease 1 (1) 0 1 (4) 0.23

Diabetes 7 (9) 9 (39) 0.002

Immunodepression 1 (1) 1 (1) 0 0.58

Cancer 7 (7) 4 (5) 3 (13) 0.2

Clinical features

MAP, median (IQR), mmHg 90 (81–99) 92 (83–100) 84 (75–98) 0.16

Heart rate, median (IQR), bpm1 91 (82–104) 91 (82–104) 91 (82–105) 0.82

Respiratory rate, median (IQR),

bpm2 23 (18–28) 22 (18–28) 26 (20–30) 0.20

SpO2, median (IQR), % 95 (93–97) 96 (93–98) 93 (91–94) 0.0001

Oxygen rate flow, median (IQR),

L/min 2 (0–6) 0 (0–4) 9 (6–30) 0.002

Mechanical ventilation, n (%) 7 (7) 0 7 (30) < 0.0001

PaO2, median (IQR), mmHg 76 (66–88) 77 (65–90) 76 (67–82) 0.66

SpO2/FiO2 ratio, median (IQR) 335 (233–452) 428 (292–457) 184 (133–244) < 0.0001

Delay between ultrasound exam

and chest CT scan, median (IQR),

hours

4 (3–7) 4 (3–6) 3 (1–10) 0.5

Data are expressed as n (%) of participants, unless otherwise indicated.

Abbreviations: BMI: body mass index; COPD: chronic obstructive pulmonary disease; MAP: mean arterial

pressure; SpO2: pulse oximetry; PaO2: arterial oxygen partial; FiO2: fractional inspired oxygen; IQR:




14

interquartile range; CT: computed tomography, ED: emergency department; ICU: intensive care unit. 1bpm: beats per minute.

2bpm: breath per minute




15

Table 2: CT and LUS findings

Details of CT and LUS findings in each group were reported in ESM Table S2.

Findings CT

n = 100

LUS

n = 100 P-value

Interstitial syndrome

Absent 8 (8) 4 (4)

0.134 Unilateral 5 (5) 11 (11)

Bilateral 87 (87) 85 (85)

Consolidation

Absent 48 (48) 68 (68)

0.002 Unilateral 16 (16) 17 (17)

Bilateral 36 (36) 15 (15)

Pleural effusion

Absent 89 (89) 94 (94)

0.122 Unilateral 7 (7) 6 (6)

Bilateral 4 (4) 0

Pneumothorax 0 0 -

Pleural irregularity 15 (15) 32 (32) 0.005

Data are expressed as n (%) of participants unless otherwise indicated.

Abbreviations: CT: computed tomography; ED: emergency department; ICU: intensive care unit.

Table 3. Correlation: LUS score according to chest CT scan severity

Chest CT Scan

Severity Patients

LUS score

Mean ± SD

95% CI for Mean LUS Score

Min

LUS Score

Max P-value Lower

Bound

Upper

Bound

Minimal damage 18 8 ± 7 4 11 0 22

< 0.0001 Moderate damage 43 14 ± 8 11 16 0 30

Severe damage 39 20 ± 8 18 23 5 34

Abbreviations: CT: computed tomography; LUS: lung ultrasound; CI: confidence interval, SD: standard

deviation.




16

(A) ROC curve of LUS score versus chest CT scan.

AUC = 0.78 (CI 95% 0.68–0.87 ; p < 0.0001).

(B) Grey zone of LUS score versus chest CT scan.

Solid line: Se. Dotted line: Sp.

Figure 1: Flow chart

Emergency Department412 patients screened

Intensive care unit87 patients screened

207 eligible patients

77 patients included

205 patients not eligible(asymptomatic or no dyspnea)

130 patients excluded(LUS exam – chest CT > 24 h)

23 patients included

64 patients excluded(LUS exam – chest CT > 24 h)

Sens

itivi

ty

1 - Specificity

Figure 2: Diagnostic accuracy of LUS score for the detection of severe SARS-CoV-2 pneumonia assessed by to chest CT scan

(B) Grey zone of LUS score versus chest CT scan(A) ROC curve of LUS score versus chest CT scan




17

Electronic supplement material ESM 1: Methods LUS exam The LUS exam was performed within two hours of admission, using an abdominal convex probe and

longitudinal scans. The use of the abdominal probe allowed a visualisation of several intercostal spaces. For each

region, we scored the most severe finding. Twelve lung regions were investigated, imaging the posterior axillary

line rather than paravertebral [1], seeking evidence of alveolar consolidation, interstitial syndrome,

pneumothorax, pleural effusion or pleural irregularity.

Evidence of pneumothorax, as demonstrated by a lack of pleural sliding in the anterior chest area, rendered any

assessment of lung changes invalid and that investigation was excluded from further analysis [2].

Interstitial syndrome was defined as two or more positive regions within the four antero-lateral regions in each

lung (8-region technique). Lung consolidation was defined according to international guidelines as a subpleural

echo-poor region, or one with tissue-like echotexture. Small consolidations giving the aspect of an irregular

pleural line were recorded as pleural line irregularity, and not as consolidation [2].

The LUS score was calculated as the sum of point values from each scanning site (0 = normal scan, 1 = moderate

interstitial syndrome, 2 = severe interstitial syndrome [multiple or coalescent B lines] and 3 = alveolar

consolidation) [3]. Pleural irregularity was qualitatively defined by the investigator and was not included in the

calculation of the LUS score. A score from 0 to 36 was then calculated. Each lung was rated as positive or

negative for each lesion. For each hemithorax, positivity was noted if an abnormal image was found in at least

one positive region. Chest CT exam The main scanning parameters were as follows: tube voltage: 80–100 kVp, automatic tube current modulation

(30–80 mAs), pitch=1; matrix: 512x512, slice thickness of 0.625–1.250 mm. If pulmonary embolism was

suspected, pulmonary CT angiography with contrast medium (75mL, OMNIPAQUETM 350 mg I/mL, GE; tube

voltage: 100–120 kV, 300 mA tube current) was performed. The scan was performed from the apex to the bases

of the lung. The images were interpreted on a PACS workstation (GE) with multi-planar reconstruction tools,

using mediastinal and parenchymal windows.

Chest CT examinations were analyzed by a resident radiologist and an experienced senior radiologist.

The pulmonary pathological manifestations were described separately for each lung in each patient. Lung

features were assessed according to the Nomenclature Committee of the Fleischner Society [4] and the first

articles describing SARS-CoV-2-associated damages [4,5]: consolidation, ground-glass opacities (GGO), crazy-

paving (a combination of GGO, thickened polygonal septal lines and intralobular reticulations), irregularity of

the bronchial wall, pleural effusion, pneumothorax, pleural irregularity and architectural distortion. Based on

these observations, the overall severity evaluation was assessed according to the type and extent of the lesions

described [6,7]. Minimal alteration corresponded to simple and focused GGO lesions, with a total extent of less

than 10%. Moderate damage corresponding to different types of lesions were observed (GGO, crazy paving,




18

consolidation) on a pulmonary area of between 10 and 50%. Severe damage corresponded to severe

consolidation or architectural distortion, and / or more than 50% of the pulmonary parenchyma was affected. For

the requirements of the statistical analysis, we decided to differentiate two groups according to the severity of the

chest CT scan: not severe (including stages of minimal and moderate involvement) and severe.

The predominant distribution (peripheral, central or both) of the lesions was also described. Anomalies that were

secondary or not directly linked to an infectious pulmonary disease were also assessed: atelectasis, diffuse

airway wall irregularity, emphysema, lymphadenopathy, the presence or absence of pulmonary embolism (if a

chest CT scan with contrast had been performed).

Results:

CT and LUS results, descriptive analysis:

Chest CT scan did not show a difference in interstitial syndrome (p = 0.46), alveolar consolidation (p = 0.06),

bronchial wall thickening (p = 0.31) or pleural thickening (p = 0.74) between the two groups. Bilateral crazy

paving was found in 23 (30%) and 13 (76%) cases in the ED group and the ICU group, respectively (p =

0.0004). The lesions were primarily classified as peripheral in both groups (p = 0.40). (ESM table S1 A).

LUS showed bilateral interstitial syndrome in 63 (82%) and 22 (96%) cases in the ED group and the ICU group,

respectively (p = 0.25). Bilateral consolidations were found in five (6%) and 10 (43%) cases in the ED group and

the ICU group, respectively (p < 0.0001). Median LUS score in the full cohort was 15 (IQR = 8–22). Compared

with the ED patients, the ICU patients had a higher median LUS score of 21 (IQR = 18–23) vs 13 (IQR = 6–20)

(p < 0.001) (ESM Table S1 B).

Diagnosis performance of LUS for the diagnosis of an interstitial syndrome, consolidation, pleural

effusion and pleural irregularity for each hemithorax (CT scan as a gold standard).

See ESM Table S2.

ESM Table S1: CT and LUS findings

(A) CT findings

CT findings All

n = 100

ED

n = 77

ICU

n = 23 P-value


Absent 8 (8) 6 (8) 2 (9)

0.46 Unilateral 5 (5) 5 (7) 0

Bilateral 87 (87) 66 (85) 21 (91)

Consolidation

Absent 48 (48) 42 (54) 6 (26)

0.06 Unilateral 16 (16) 11 (14) 5 (21)

Bilateral 36 (36) 24 (32) 12 (53)




19

Pneumothorax 0 0 0

Pleural effusion

Absent 89 (89) 73 (94) 16 (70)

0.002 Unilateral 7 (7) 2 (3) 5 (21)

Bilateral 4 (4) 2 (3) 2 (9)

Crazy paving

Absent 51 (51) 49 (63) 2/17 (12)

0.0004 Unilateral 7 (7) 5 (7) 2/17 (12)

Bilateral 36 (36) 23 (30) 13/17 (76)

Missing data 6 (6)

Bronchial wall thickening

Absent 54 (54) 7 (41) 47 (61)

0.31 Unilateral 13 (13) 3 (18) 10 (13)

Bilateral 27 (27) 7 (41) 20 (26)

Missing data 6 (6)

Pleural thickening 15 (15) 11 (14) 4 (17) 0.74

Lung architecture distortion

Presence 15 (15) 14 (18) 1/12 (8) 0.68

Absence 74 (74) 63 (82) 11/12 (92)

Missing data 11 (11)

Pulmonary distribution

Subpleural 48 (48) 42/72 (58) 6/14 (43)

0.4 Central 2 (2) 2/72 (3) 0

Heterogeneous 36 (36) 28/72 (39) 8 (57)

Missing data 14 (14)


Abbreviations: CT: computed tomography; ED: emergency department; ICU: intensive care unit.

(B) LUS findings

LUS findings All

n = 100

ED

n = 77

ICU

n = 23 P-value


Absent 4 (4) 4 (5) 0

0.25 Unilateral 11 (11) 10 (13) 1 (4)

Bilateral 85 (85) 63 (82) 22 (96)

Consolidation




20

Absent 68 (68) 63 (82) 5 (22)

< 0.0001 Unilateral 17 (17) 9 (12) 8 (35)

Bilateral 15 (15) 5 (6) 10 (43)

Pleural effusion

Absent 94 (94) 19 (83) 75 (97)

0.02 Unilateral 6 (6) 4 (17) 2 (3)

Bilateral 0 0

Pneumothorax 0 0 0

Pleural irregularity 32 (32) 25 (32) 7 (30) 0.85

LUS score, median (IQR) 15 (8–22) 13 (6–20) 21 (18–23) 0.001


Abbreviations: LUS: lung ultrasound, IQR: interquartile range, ED: emergency department; ICU: intensive

care unit.




21

ESM Table S2: Sensitivity, specificity, positive and negative predictive values, and diagnostic accuracy of

LUS for the diagnosis of an interstitial syndrome, consolidation, pleural effusion and pleural irregularity

for each hemithorax (CT scan as a gold standard)

Syndrome LUS CT + CT- Sensitivity

(%)a

Specificity

(%)b

PPV

(%)c

NPV

(%)d

DA

(%)e

Interstitial

syndrome

LU +

LU -

165

14

16

5

92.2 23.8 91.2 26.3 85

Consolidation LU +

LU -

30

58

17

93

34.1 84.8 63.8 62.1 63

Pleural effusion LU +

LU -

5

10

3

182

33.3 98.4 62.5 94.8 94

Pleural

irregularity

(n = 100)

LU +

LU -

7

8

25

60

46.7 70.6 21.9 88.2 67

Each hemithorax ( n = 200) was characterized as positive (+) or negative (-) for abnormality by the presence

or absence of a single positive region, respectively.

TP true positive, TN true negative, FP false positive, FN false negative

a Sensitivity = [TP / (TP + FN)] x 100 b Specificity = [TN / (TN + FP)] x 100 c Positive predictive value (PPV) = [TP / (TP + FP)] x 100 d Negative predictive value (NPV) = [TN / (TN + FN)] x 100 e Diagnostic accuracy (DA) = [(TP + TN) / (TP + TN + FP + FN)] x 100




22

References Electronic supplement material

1. Mojoli F, Bouhemad B, Mongodi S, Lichtenstein D (2018) Lung ultrasound for critically ill patients.

Am J Respir Crit Care Med 199:701–714. Https://doi.org/10.1164/rccm.201802-0236CI

2. Volpicelli G, Elbarbary M, Blaivas M, et al (2012) International evidence-based recommendations for

point-of-care lung ultrasound. Intensive Care Med 38:577–591. Https://doi.org/10.1007/s00134-012-2513-4

3. Volpicelli G (2013) Lung sonography. J Ultrasound Med 32:165–171.

Https://doi.org/10.7863/jum.2013.32.1.165

4. Austin JH, Müller NL, Friedman PJ, et al (1996) Glossary of terms for CT of the lungs:

recommendations of the Nomenclature Committee of the Fleischner Society. Radiology 200:327–331.

Https://doi.org/10.1148/radiology.200.2.8685321

5. Fan N, Fan W, Li Z, et al (2020) Imaging characteristics of initial chest computed tomography and

clinical manifestations of patients with COVID-19 pneumonia. Jpn J Radiol 1–6.

Https://doi.org/10.1007/s11604-020-00973-x

6. Yuan M, Yin W, Tao Z, et al (2020) Association of radiologic findings with mortality of patients

infected with 2019 novel coronavirus in Wuhan, China. PLOS ONE 15:e0230548.

Https://doi.org/10.1371/journal.pone.0230548

7. Zhao W, Zhong Z, Xie X, et al (2020) Relation between chest CT findings and clinicalcConditions of

Coronavirus Disease (COVID-19) pneumonia: a multicenter study. American Journal of Roentgenology

214:1072–1077. Https://doi.org/10.2214/AJR.20.22976

original un-edited accepted proof€¦ · keywords: sars-cov-2, chest computed tomography, lung...

Documents