Point-of-Care Ultrasonography for Evaluation of Acute Dyspnea in the ED

(1)

Point-of-Care Ultrasonography for

Evaluation of Acute Dyspnea in the ED

Maurizio Zanobetti, MD; Margherita Scorpiniti, MD; Chiara Gigli, MD; Peiman Nazerian, MD; Simone Vanni, MD; Francesca Innocenti, MD; Valerio T. Stefanone, MD; Caterina Savinelli, MD; Alessandro Coppa, MD; So_{ﬁa Bigiarini, MD;} Francesca Caldi, MD; Irene Tassinari, MD; Alberto Conti, MD; Stefano Grifoni, MD; and Riccardo Pini, MD

BACKGROUND:Acute dyspnea is a common symptom in the ED. The standard approach to dyspnea often relies on radiologic and laboratory results, causing excessive delay before adequate therapy is started. Use of an integrated point-of-care ultrasonography (PoCUS) approach can shorten the time needed to formulate a diagnosis, while maintaining an

acceptable safety proﬁle.

METHODS:Consecutive adult patients presenting with dyspnea and admitted after ED

eval-uation were prospectively enrolled. The gold standard was theﬁnal diagnosis assessed by two

expert reviewers. Two physicians independently evaluated the patient; a sonographer per-formed an ultrasound evaluation of the lung, heart, and inferior vena cava, while the treating physician requested traditional tests as needed. Time needed to formulate the ultrasound and the ED diagnoses was recorded and compared. Accuracy and concordance of the ultrasound and the ED diagnoses were calculated.

RESULTS:A total of 2,683 patients were enrolled. The average time needed to formulate the ultrasound diagnosis was signiﬁcantly lower than that required for ED diagnosis (24

10 min vs 186 72 min; P ¼ .025). The ultrasound and the ED diagnoses showed good

overall concordance (

k

¼ 0.71). There were no statistically signiﬁcant differences in the

accuracy of PoCUS and the standard ED evaluation for the diagnosis of acute coronary syndrome, pneumonia, pleural effusion, pericardial effusion, pneumothorax, and dyspnea

from other causes. PoCUS was signiﬁcantly more sensitive for the diagnosis of heart failure,

whereas a standard ED evaluation performed better in the diagnosis of COPD/asthma and pulmonary embolism.

CONCLUSIONS:PoCUS represents a feasible and reliable diagnostic approach to the patient with dyspnea, allowing a reduction in time to diagnosis. This protocol could help to stratify

patients who should undergo a more detailed evaluation. CHEST 2017; 151(6):1295-1301

KEY WORDS:dyspnea; emergency medicine; ultrasound

ABBREVIATIONS:ALI = acute lung injury; CCT = chest CT; CXR = chest radiograph; ECHO = echocardiography; IVC = inferior vena cava; LUS = lung ultrasonography; PoCUS = point-of-care ultrasonography

AFFILIATIONS: From the Emergency Department, Careggi University Hospital, Florence, Italy.

FUNDING/SUPPORT:The authors have reported to CHEST that no funding was received for this study.

CORRESPONDENCE TO: Maurizio Zanobetti, MD, Largo Brambilla 3, 50134 Firenze, Italy; e-mail:zanomau@libero.it

DOI:http://dx.doi.org/10.1016/j.chest.2017.02.003

(2)

Acute dyspnea is a frequent presenting symptom in the ED. Emergency physicians need to rapidly and

accurately formulate a diagnosis to guide an early and appropriate therapeutic intervention, especially in critically ill patients in whom rapid stabilization is crucial to reduce morbidity and mortality.

Dyspnea is the leading symptom of many diseases, and rapid discrimination of the underlying pathologic

condition is often difﬁcult for the emergency physician.

A careful anamnesis and a thorough physical

examination can direct diagnostic suspicion, but these methods need to be integrated with laboratory and radiologic testing, whose results are often also

delayed.1-3

Patients with dyspnea are often evaluated by using chest radiograph (CXR) or chest CT (CCT) scans; these techniques, however, have some disadvantages, including radiation exposure and contraindication in

pregnant patients.4 Moreover, it can be difﬁcult to

perform CXR in critically ill patients, and results may be

misinterpreted.5,6Conversely, CCT scanning, which is

considered the gold standard for the differential

diagnosis of dyspnea,7has high costs, is not feasible in

patients in unstable condition, and is not available 24 h a

day in all ED settings.8

Point-of-care ultrasonography (PoCUS) has become a widely used diagnostic tool in the ED. Because it can be rapidly performed at the bedside by the emergency physician, it should be considered an extension of the

physical examination, adding anatomic and functional

information to the clinical data.9-11This approach allows

identiﬁcation of the most likely diagnosis and prompt initiation of urgent therapies.

Many studies have conﬁrmed the high accuracy of lung ultrasonography (LUS) in the differential diagnosis of acute respiratory failure. LUS seems to be at least as accurate as CXR, with a higher sensitivity for pulmonary edema, pneumothorax, pneumonia, and free pleural

effusion.12-15

Emergency echocardiography (ECHO) performed by emergency physicians can also add some important information regarding patients with acute dyspnea (ie, dilation of heart chambers, visual estimation of left

ventricular ejection fraction,16-18 presence of pericardial

ﬂuid). Integrating LUS and ECHO helps to differentiate

cardiogenic from noncardiogenic dyspnea.19-21 Finally,

the inferior vena cava (IVC) diameter and its variations during respiratory activity allow an indirect estimation of right atrial pressure and an assessment of volume

status, providing useful information to guideﬂuid

therapy.22,23

Based on this information, integration of lung, heart, and IVC ultrasound with PoCUS could lead to a rapid diagnosis and therefore to targeted, early treatment of patients with acute dyspnea in the ED. The aim of the present study was to evaluate the feasibility and diagnostic accuracy of PoCUS for the management of patients with acute dyspnea in the ED.

Patients and Methods

Design, Setting, Protocol, and Population

This prospective, blinded, observational study was performed in the ED of a university-afﬁliated teaching hospital. The study, which is consistent with the principles of the Declaration of Helsinki on clinical research involving human subjects, was approved by the institutional review board committee of the ED of Careggi University Hospital (project approval no. 05/2012). Written informed consent was obtained from each patient included in the study.

Consecutive adult patients (aged > 18 years) presenting with acute dyspnea of every degree were considered for the study; patients with dyspnea of traumatic origin or discharged after ED evaluation were excluded. The study was performed by 10 emergency physician sonographers who had previously attended a professional 80-h course of theoretical lessons and a training program comprising 150 LUS and 150 ECHO in the ED (as required by the American College of Emergency Physicians’ guidelines24_{) and with at least 2 years’}

experience.

Primary assessment of enrolled patients consisted of a routine evaluation with detection of vital signs, a medical history, physical examination, and 12-lead ECG; the treating physician then requested

CXR, CCT scans, ECHO performed by a cardiologist, and blood sampling or arterial blood gas analysis as needed. The treating physician gave notiﬁcation to one of the emergency physician sonographers, who performed the PoCUS before the results of these additional tests were received. After the completion of the ultrasound evaluation, the sonographer investigator, only aware of the results of the primary assessment, completed a standardized form, specifying which diagnosis among heart failure, acute coronary syndrome, pneumonia, pleural effusion, pericardial effusion, COPD/asthma, pulmonary embolism, pneumothorax, ARDS/acute lung injury (ALI), or other (cancer, interstitial lung disease, psychogenic dyspnea, metabolic disorder, neurologic disorder, or musculoskeletal chest pain) was the most likely (ultrasound diagnosis). For each patient, up to two concomitant diagnoses could be present. Finally, the treating physician, blinded to the PoCUSﬁndings, after evaluating results of the previously requested additional tests, completed a standardized form specifying which one or two of the aforementioned diagnoses were most likely (ED diagnosis).

For each patient, the time of entry into the visit room, the time needed to perform the ultrasound scan and relative diagnosis, and the time of formulation of ED diagnosis were recorded. Ultimately, the ultrasound and ED diagnoses were compared with theﬁnal diagnosis, which was considered the gold standard. The ﬁnal diagnosis was formulated

(3)

by two emergency medicine experts, who reviewed the clinical chart and were aware of the results of every laboratory, ultrasound, and radiologic examination performed on the patient both during the ED stay and the hospital stay. For each patient, up to two concomitant diagnoses could be present.

PoCUS Evaluation

PoCUS was performed with a multiprobe machine (MyLab 30 Gold; Esaote SpA) with a 4- to 8-MHz linear probe and a 2.5- to 3.5-MHz curved array probe. Each ultrasound examination was performed by following a systematic and standardized sequence (LUS, ECHO, IVC evaluation) and according to predeﬁned ultrasound protocols. The lungs were examined by using longitudinal and oblique scans on anterolateral and posterior thoracic areas. Anterolateral examination was performed with the patient in the supine or near-to-supine position; whenever possible, dorsal areas were scanned in the sitting position or by turning the patient in the lateral decubitus on both sides in case of forced supine position.

Lung examination was targeted to the detection of speciﬁc ultrasound patterns identiﬁed according to international recommendations on point-of-care lung ultrasound25_{: pulmonary edema as bilateral diffuse}

interstitial syndrome; pneumonia as the presence of lung consolidation and air bronchogram(s) with or without focal interstitial syndrome; pleural effusion as the presence of an anechoic space between the parietal and visceral pleura; pulmonary embolism was considered in the presence of two or more triangular or rounded pleural-based lesions indicating a pulmonary infarction,26 or, as COPD/asthma, in the absence of any aforementioned pattern in a patient with suggestive medical history; pneumothorax was deﬁned as the absence of lung sliding, B-lines, and lung pulse with the presence of lung point; and ARDS as the presence of subpleural anterior consolidations with the absence or reduction of lung sliding, spared areas of normal parenchyma, pleural line

abnormalities such as irregularly thickened or fragmented pleural line and nonhomogeneous distribution of B-lines.

Afterwards ECHO was then performed in an apical four-chamber view to allow for qualitative evaluation of left ventricular ejection fraction (which was considered reduced if< 50%) and the presence of right ventricular dilation (right/left ventricular end-diastolic diameter ratio > 1 at the level of atrioventricular valve annulus). A subcostal long-axis view was performed to assess the presence of pericardial effusion and to evaluate left ventricular ejection fraction.

Finally, IVC maximum and minimum diameters and the IVC collapsibility index were measured in the subcostal view in M-mode at 2 cm from the right atrial junction.27,28 The IVC collapsibility index was considered reduced if < 50%, normal, or increased if$ 50%.29

Findings of the ultrasound examination were collected in a standardized form.

Statistical Analysis

Data points are expressed as mean SD. The unpaired Student t test was used to compare normally distributed data. The Fisher exact test was used for the comparison of noncontinuous variables expressed as proportions. P values< .05 indicate statistical significance; all P values are two-sided. The diagnostic performance of PoCUS and of a traditional ED evaluation was assessed by calculating sensitivity, specificity, positive predictive value, negative predictive value, and likelihood ratios. The k statistic was used to compare the concordance between different diagnostic methods.30The McNemar test was used to compare the sensitivities and specificities of PoCUS and traditional ED evaluation for each diagnosis.31Calculations were performed by using SPSS version 20.0 (IBM SPSS Statistics, IBM Corporation).

Results

From January 2013 to December 2013, a total of 3,487 patients with dyspnea as their presenting symptom were evaluated at the ED. Among them, 804 patients were excluded from the study (734 were discharged from the ED, 57 had dyspnea of traumatic origin, 7 did not consent, and 6 died in the ED); 2,683 patients were thus

included. Enrolled patients had a mean age of 71.2

18.6 years (range, 18-107 years), and 1,316 were women. The main characteristics of enrolled patients are shown inTable 1.

During the ED evaluation or the hospital stay, 2,629 patients underwent CXR (98%), 295 underwent CCT scanning (11%), and 402 underwent ECHO performed by a cardiologist (15%). PoCUS was performed in every patient enrolled in the study with an average time of

7 2 min (3 1 min for LUS and 4 1 min for

ECHO). The average time needed to formulate the

ultrasound diagnosis was 24 10 min, whereas the

formulation of the ED diagnosis required a signiﬁcantly

longer time (186 72 min; P ¼ .025).

Causes of dyspnea according to the ultrasound

diagnosis, the ED diagnosis, and theﬁnal diagnosis are

reported inTable 2. The total number of ultrasound

diagnoses, ED diagnoses, andﬁnal diagnoses was 2,791,

2,798, and 2,890, respectively; patients with a double

TABLE 1] General Characteristics of the Study Population Characteristic Value Age, mean SD, y 71.2 18.6 Women, No. (%) 1,316 (49) SBP, mm Hg 134.2 DBP, mm Hg 75.2

Heart rate, beats/min 88.2

Respiratory rate, breaths/min 22.6

Body temperature,C 36.8

SaO2, % 93

Patients with sinus rhythm, No. 2,120

DBP¼ diastolic blood pressure; SaO2¼ oxygen saturation; SBP ¼ systolic blood pressure.

(4)

diagnosis comprised 108 (3.9%), 115 (4.1%), and 207 (7.2%). The most frequent double diagnosis was between heart failure and pneumonia (32 cases for ultrasound diagnosis, corresponding to 29.6% of double the ultrasound diagnosis; 49 cases for ED diagnosis, corresponding to 42.6% of double the ED diagnosis; and

75 cases for ﬁnal diagnosis, corresponding to 36.2% of

double ﬁnal diagnosis).

Concordance between the ultrasound and ED diagnoses

for each disease wasﬁrst calculated, and the results

are reported in Table 3. Concordance was optimal

(0.8<

k

< 1) for the diagnosis of heart failure,

pericardial effusion, COPD/asthma, and pneumothorax;

good (0.6 <

k

< 0.8) for acute coronary syndrome,

pneumonia, isolated pleural effusion, and other causes;

moderate (0.4<

k

< 0.6) for pulmonary embolism; and

poor for ARDS/ALI (0<

k

< 0.4). Overall concordance

was good (

k

¼ 0.71). Subsequently, ultrasound and ED

diagnostic performances for each speciﬁc diagnosis were

calculated comparing the ultrasound and ED diagnoses

with the ﬁnal diagnosis; results are shown inTables 4

and 5, respectively. For heart failure, the ultrasound

diagnosis was signiﬁcantly more sensitive than the ED

diagnosis (88% vs 77%; P < .001), despite a minor

speciﬁcity (96% vs 98%; P < .001). A statistically

signiﬁcant difference in sensitivity was also found in the

diagnosis of COPD/asthma and pulmonary embolism, in which the ultrasound diagnosis was inferior to the ED

diagnosis (87% vs 92% [P < .001] for COPD/asthma;

40% vs 91% [P< .001] for pulmonary embolism). In

addition, speciﬁcity of the ultrasound diagnosis for

ARDS/ALI was signiﬁcantly inferior to that of the ED diagnosis, even if almost optimal (99.5% vs 99.9%;

P¼ .002). There was no statistically signiﬁcant

difference between sensitivities and speciﬁcities of the

ultrasound and ED diagnoses in terms of acute coronary syndrome, pneumonia, pleural effusion, pericardial effusion, pneumothorax, and other causes of dyspnea. The ultrasound diagnosis was highly sensitive and speciﬁc for the diagnosis of pneumothorax, pneumonia, and pericardial effusion; moreover, it was also very speciﬁc for the diagnosis of acute coronary syndrome, pleural effusion, pulmonary embolism, and dyspnea from other causes.

Discussion

This prospective study evaluated the feasibility and usefulness of a PoCUS approach compared with a standard ED evaluation. From these results, we can afﬁrm that PoCUS is a reliable and accurate tool that could be used in the initial approach to patients in the ED with dyspnea.

Ourﬁrst goal was to show that the ultrasound diagnosis

was concordant with the ED diagnosis in enrolled patients; the high obtained value of concordance

(

k

¼ 0.71), which is associated with a signiﬁcant reduction

in diagnostic time in favor of PoCUS, conﬁrms that

PoCUS provides a reliable early diagnosis in patients with dyspnea and, subsequently, allows use of an early targeted therapy.

TABLE 2] Causes of Dyspnea According to Ultrasound Diagnosis, ED Diagnosis, and Final

Diagnosis Variable Ultrasound Diagnosis ED Diagnosis Final Diagnosis Heart failure 600 503 585 Acute coronary syndrome 32 30 42 Pneumonia 1,096 1,091 1,086 Pleural effusion 97 111 98 Pericardial effusion 45 48 44 COPD/asthma 735 782 759 Pulmonary embolism 41 95 95 Pneumothorax 39 45 44 ARDS/ALI 20 7 16 Other causes 86 86 121 Total 2,791 2,798 2,890

ALI¼ acute lung injury.

TABLE 3] Concordance Between Ultrasound Diagnosis and ED Diagnosis

Diagnosis k

Heart failure 0.810

Acute coronary syndrome 0.706

Pneumonia 0.788 Pleural effusion 0.730 Pericardial effusion 0.858 COPD/asthma 0.845 Pulmonary embolism 0.549 Pneumothorax 0.903 ARDS/ALI 0.294 Other 0.628 Total 0.711

(5)

The time needed to formulate an ultrasound diagnosis

was signiﬁcantly less than that required for an ED

diagnosis. Although we are aware that this difference could be partially due to co-existing clinical duties that distracted the treating physician from evaluation of the patient with dyspnea, the difference in the average time strongly favors the ultrasound evaluation.

Analyzing diagnostic performance, we found that PoCUS was statistically superior to the standard ED evaluation for a heart failure diagnosis, which

represented theﬁnal diagnosis in 20% of the study

patients; because the negative predictive value of the ultrasound diagnosis for heart failure was very high (97%), we concluded that the absence of the ultrasound patterns associated with heart failure (ie, diffuse interstitial syndrome, reduced ejection fraction) allowed us to rule out with high reliability this condition as the main cause of dyspnea.

The most common cause of dyspnea in this study sample was pneumonia (38% of diagnoses); the sensitivity and speciﬁcity of the ultrasound diagnosis were similar to what have been reported in the literature (sensitivity 88.5% vs 82.8% and speciﬁcity

91.6% vs 95.5%)15and comparable to those of the ED

diagnosis (neither P value signiﬁcant), but the diagnosis was reached earlier with PoCUS.

Moreover, no signiﬁcant difference in terms of

sensitivity and speciﬁcity was found for the diagnosis of

acute coronary syndrome, pleural effusion, pericardial effusion, pneumothorax, or dyspnea due to other causes,

representing 12% of theﬁnal diagnoses.

However, the sensitivity of the standard ED evaluation was statistically superior to that of PoCUS for the diagnosis of COPD/asthma and pulmonary embolism.

In our opinion, thisﬁnding is due to a prevalent clinical

diagnosis of COPD/asthma, mostly derived from the

patient’s medical history and from improvement after

speciﬁc therapy, often without radiologic or

ultrasound-speciﬁc ﬁndings. Concerning pulmonary embolism, the lower sensitivity of the ultrasound diagnosis compared with that of the ED diagnosis (40% vs 91%) is likely due to the essential information provided by the CT pulmonary angiography, which is frequently included in the standard ED evaluation of patients with suspected dyspnea. However, in the present study, of 41

ultrasound diagnoses of pulmonary embolism conﬁrmed

atﬁnal diagnosis, 40 (97%) were associated with right

ventricular dilation and 27 (66%) with the detection of pulmonary infarction. In 54 cases, pulmonary embolism

TABLE 4 ] Diagnostic Accur acy of Ultr asound Diagnoses Variable Sensitivity, % (95% CI) Speci ﬁ city, % (95% CI) PPV, % (95% CI) NPV, % (95% CI) þ LR (95% CI) –LR (95% CI) Heart failure 88 (85.1-90.6) 96 (95-96.8) 85.8 (82.8-88.5) 96.6 (95.8-97.4) 21.73 (17.61-26.82) 0.12 (0.10-0.16) Acute coronary syndrome 47.6 (32-63.6) 99.6 (99.2-99.8) 62.5 (43.7-78.9) 99.2 (98.8-99.5) 104.8 (54.85-200.26) 0.53 (0.39-0.70) Pneumonia 88.5 (86.4-90.3) 91.6 (90.1-92.9) 87.7 (85.6-89.6) 92.1 (90.7-93.4) 10.47 (8.90-12.32) 0.13 (0.11-0.15) Pleural effusion 77.6 (68-85.4) 99.2 (98.8-99.5) 78.4 (68.8-86.1) 99.2 (98.7-99.5) 95.46 (61.54-148.09) 0.23 (0.16-0.33) Pericardial effusion 86.4 (72.7-94.8) 99.7 (99.5-99.9) 84.4 (70.5-93.5) 99.8 (99.5-99.9) 325.59 (153.94-688.65) 0.14 (0.06-0.29) COPD/asthma 86.8 (84.2-89.2) 96.1 (95.1-96.9) 89.7 (87.2-91.8) 94.9 (93.8-95.8) 21.98 (17.60-27.45) 0.14 (0.11-0.16) Pulmonary embolism 40 (30.1-50.6) 99.9 (99.7-100) 92.7 (80.1-98.5) 97.8 (97.2-98.4) 345.07 (108.45-1097.94) 0.60 (0.51-0.71) Pneumothorax 87.8 (75.2-94.5) 100 (99.8-100) 98.8 (89.1-99.9) 99.8 (99.5-99.9) 4634.67 (289.35-74236.28) 0.12 (0.06-0.27) ARDS/ALI 43.8 (19.8-70.1) 99.5 (99.2-99.7) 35 (15.4-59.2) 99.7 (99.4-99.9) 89.75 (41.29-195.09) 0.57 (0.37-0.87) Other causes 45.5 (36.4-54.8) 98.8 (98.3-99.2) 64 (52.9-74) 97.5 (96.8-98) 37.57 (25.16-56.08) 0.55 (0.47-0.65) –LR ¼ negative likelihood ratio; NPV ¼ negative predictive value; PPV ¼ positive predictive value; þ LR ¼ positive likelihood ratio.

(6)

was not diagnosed according to PoCUS; none of these patients had right ventricular dilation, but 3 cases (5%)

hadﬁndings suggestive of pulmonary infarction. From

these results, it seems that right ventricular dilation is essential for the suspicion of pulmonary embolism; the presence of characteristic triangular or rounded pleural-based consolidation is seemingly hardly interpreted as a pulmonary infarction in the absence of right ventricular dilation.

This study did have some limitations. PoCUS examination was performed by emergency physicians

with at least 2 years’ experience in ultrasound;

application of the same methodology by physicians with less experience may lower accuracy and safety.

Another limitation is that patients who were not admitted after ED evaluation were not included in the study. The sensitivity of PoCUS for pulmonary embolism could have been increased by performing a compression ultrasound searching for DVT. This compression ultrasound was not included because when we conceived the study, this examination was considered too time-consuming to be included in an integrated ultrasound evaluation of patients with dyspnea compared with the expected prevalence of pulmonary embolism in this setting. The prevalence of ARDS diagnosis in this study population was so limited (16 patients) that we could not reach deﬁnitive conclusions about the clinical utility of PoCUS in these patients.

Conclusions

The present study showed a good concordance between the diagnoses reached after PoCUS evaluation and after standard ED evaluation in the differential diagnosis of patients with dyspnea. There was a signiﬁcant reduction

in diagnostic time in favor of PoCUS; thisﬁnding is

extremely important in the emergency setting because a rapid diagnosis and therapy are often critical for the

patient’s prognosis.

PoCUS may represent theﬁrst feasible and accurate

diagnostic approach to the patient with dyspnea in the ED, helping stratifying patients who should undergo a second-level diagnostic test (ie, cases with nonconclusive ultrasound evaluation or needing more detailed studies).

If these data are conﬁrmed by further studies, this

integrated ultrasound method could be the basis of a

new semeiotics and replace the currentﬁrst diagnostic

approach to patients presenting with dyspnea, allowing a drastic reduction in costs and diagnostic times.

TABLE 5 ] Diagnostic Accur acy of ED Diagnosis Variable Sensitivity, % (95% CI) Speci ﬁ city, % (95% CI) PPV, % (95% CI) NPV, % (95% CI) þ LR (95% CI) –LR (95% CI) Heart failure 77.3 (73.7-80.6) 97.6 (96.8-98.2) 89.9 (86.9-92.4) 93.9 (92.8-94.9) 31.8 (24.2-41.8) 0.23 (0.20-0.27) Acute coronary syndrome 52.4 (36.4-68) 99.7 (99.4-99.9) 73.3 (54.1-87.7) 99.3 (98.84-99.5) 172.92 (81.72-365.93) 0.48 (0.35-0.66) Pneumonia 89.8 (87.8-91.5) 92.7 (91.4-94) 89.4 (87.4-91.1) 93 (91.7-94.2) 12.36 (10.36-14.74) 0.11 (0.09-0.13) Pleural effusion 86.7 (78.4-92.7) 99 (98.5-99.3) 76.6 (67.6-84.1) 99.5 (99.1-99.7) 86.23 (58.37-127.39) 0.13 (0.08-0.22) Pericardial effusion 93.2 (81.3-98.6) 99.7 (99.5-99.9) 85.4 (72.2-93.9) 99.9 (99.7-99.9) 351.30 (166.92-739.34) 0.07 (0.02-0.20) COPD/asthma 92.2 (90.1-94) 95.7 (94.7-96.6) 89.5 (87.2-91.6) 96.9 (96-97.6) 21.64 (17.49-26.77) 0.08 (0.06-0.10) Pulmonary embolism 90.5 (82.8-95.6) 99.7 (99.3-99.8) 90.5 (82.8-95.6) 99.7 (99.3-99.8) 260.31 (135.16-501.36) 0.10 (0.05-0.18) Pneumothorax 95.5 (84.5-99.4) 99.9 (99.7-100) 93.3 (81.7-98.6) 99.9 (99.7-100) 839.68 (270.49-2606.65) 0.05 (0.01-0.18) ARDS 37.5 (15.2-64.6) 99.9 (99.8-100) 85.7 (42.1-99.6) 99.6 (99.3-99.8) 1000.12 (127.57-7840.72) 0.63 (0.43-0.91) Other diagnoses 54.6 (45.2-63.6) 99.2 (98.8-99.5) 76.7 (66.4-85.2) 97.9 (97.3-98.4) 69.87 (43.85-111.34) 0.46 (0.38-0.56) See Table 4 legend for expansion of abbreviations.

(7)

Acknowledgments

Author contributions: M. Z. is the guarantor of the manuscript. M. Z. contributed to study conception and design and data acquisition, analysis, and interpretation; drafted the manuscript; edited the manuscript for important intellectual and scientific content; served as the principal author; and edited the revision. M. S. and C. G. contributed to study conception and design, conducted statistical analysis, drafted the manuscript, and edited the revision. P. N. contributed to study design and data acquisition, drafted and edited the manuscript for important intellectual and scientific content, and edited the revision. S. V., F. I., V. T. S., and C. S. contributed to data acquisition, conducted statistical analysis, and edited the revision. A. C., S. B., F. C., I. T., A. C., and S. G. contributed to data acquisition. R. P. contributed to study conception and design, as well as to data analysis and interpretation. All authors approved thefinal draft.

Financial/nonﬁnancial disclosure: None declared.

References

1. Mulrow CD, Lucey CR, Farnett LE. Discriminating causes of dyspnea through clinical examination. J Gen Intern Med. 1993;8(7):383-392.

2. Schmitt BP, Kushner MS, Wiener SL. The diagnostic usefulness of the history of the patient with dyspnea. J Gen Intern Med. 1986;1(6):386-393.

3. Nielsen LS, Svanegaard J, Wiggers P, Egeblad H. The yield of a diagnostic hospital dyspnoea clinic for the primary health care section. J Intern Med. 2001;250(5):422-428.

4. Brent RL. The effect of embryonic and fetal exposure to x-ray, microwaves, and ultrasound: counseling the pregnant and nonpregnant patient about these risks. Semin Oncol. 1989;16(5):347-368. 5. Greenbaum DM, Marschall KE. The

value of routine daily chest x-rays in intubated patients in the medical intensive care unit. Crit Care Med. 1982;10(1):29-30.

6. Lichtenstein D, Goldstein I, Mourgeon E, Cluzel P, Grenier P, Rouby JJ.

Comparative diagnostic performances of auscultation, chest radiography, and lung ultrasonography in acute respiratory distress syndrome. Anesthesiology. 2004;100(1):9-15.

7. Ivatury RR, Sugerman HJ. Chest radiograph or computed tomography in the intensive care unit? Crit Care Med. 2000;28(4):1234-1235.

8. Heggie JC. Patient doses in multi-slice CT and the importance of optimisation. Australas Phys Eng Sci Med. 2005;28(2): 86-96.

9. Jehle D, Davis E, Evans T, et al. Emergency department sonography by emergency physicians. Am J Emerg Med. 1989;7(6):605-611.

10. Plummer D. Whose turf is it, anyway? Diagnostic ultrasonography in the emergency department. Acad Emerg Med. 2000;7(2):186-187.

11. Lichtenstein D, Axler O. Intensive use of general ultrasound in the intensive care unit. Prospective study of 150 consecutive patients. Intensive Care Med. 1993;19(6): 353-355.

12. Lichtenstein D, Mezière G. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest. 2008;134(1):117-125.

13. Reissig A, Copetti R, Mathis G, et al. Lung ultrasound in the diagnosis and follow-up of community-acquired pneumonia: a prospective, multicenter, diagnostic accuracy study. Chest. 2012;142(4): 965-972.

14. Zanobetti M, Poggioni C, Pini R. Can chest ultrasonography replace standard chest radiography for evaluation of acute dyspnea in the ED? Chest. 2011;139(5): 1140-1147.

15. Nazerian P, Volpicelli G, Vanni S, et al. Accuracy of lung ultrasound for the diagnosis of consolidations when compared to chest computed tomography. Am J Emerg Med. 2015;33(5):620-625. 16. Mueller X, Stauffer JC, Jaussi A, Goy JJ,

Kappenberger L. Subjective visual echocardiographic estimate of left ventricular ejection fraction as an alternative to conventional

echocardiographic methods: comparison with contrast angiography. Clin Cardiol. 1991;14(11):898-902.

17. Moore CL, Rose GA, Tayal VS, Sullivan M, Arrowood JA, Kline JA. Determination of left ventricular function by emergency physician echocardiography of hypotensive patients. Acad Emerg Med. 2002;9(3):186-193.

18. Melamed R, Sprenkle MD, Ulstad VK, Herzog CA, Leatherman JW. Assessment of left ventricular function by intensivists using hand-held echocardiography. Chest. 2009;135(6):1416-1420.

19. Nazerian P, Vanni S, Zanobetti M, et al. Diagnostic accuracy of emergency Doppler echocardiography for identiﬁcation of acute left ventricular heart failure in patients with acute dyspnea: comparison with Boston criteria

and N-terminal prohormone brain natriuretic peptide. Acad Emerg Med. 2010;17(1):18-26.

20. Frassi F, Gargani L, Gligorova S, Ciampi Q, Mottola G, Picano E. Clinical and echocardiographic determinants of ultrasound lung comets. Eur J Echocardiogr. 2007;8(6):474-479. 21. Gargani L. Lung ultrasound: a new tool

for the cardiologist. Cardiovasc Ultrasound. 2011;27(9):6.

22. Stawicki SP, Braslow BM, Panebianco NL, et al. Intensivist use of hand-carried ultrasonography to measure IVC collapsibility in estimating intravascular volume status: correlations with CVP. J Am Coll Surg. 2009;209(1):55-61. 23. Feissel M, Michard F, Faller JP, Teboul JL.

The respiratory variation in inferior vena cava diameter as a guide toﬂuid therapy. Intensive Care Med. 2004;30(9): 1834-1837.

24. ACEP. Emergency ultrasound fellowship guidelines. ACEP. 2011.

25. Volpicelli G, Elbarbary M, Blaivas M, et al. International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med. 2012;38(4):577-591.

26. Mathis G, Blank W, Reissig A, et al. Thoracic ultrasound for diagnosing pulmonary embolism: a prospective multicenter study of 352 patients. Chest. 2005;128(3):1531-1538.

27. Barbier C, Loubières Y, Schmit C, et al. Respiratory changes in inferior vena cava diameter are helpful in predictingﬂuid responsiveness in ventilated septic patients. Intensive Care Med. 2004;30(9): 1740-1746.

28. Goldﬂam K, Saul T, Lewiss R. Focus on:

inferior vena cava ultrasound. ACEP. 2011.

29. Porter TR, Shillcutt SK, Adams MS, et al. Guidelines for the use of echocardiography as a monitor for therapeutic intervention in adults: a report from the American Society of Echocardiography. J Am Soc Echocardiogr. 2015;28(1):40-56. 30. McGinn T, Wyer PC, Newman TB,

Keitz S, Leipzig R, Guyatt G. Tips for learners of evidence-based medicine: 3. Measures of observer variability (kappa statistic). CMAJ. 2004;171(11): 1369-1373.

31. Hawass N. Comparing the sensitivities and speciﬁcities of two diagnostic procedures performed on the same group of patients. Br J Radiol. 1997;70(832): 360-366.