Accuracy of Several Lung Ultrasound Methods for the Diagnosis of Acute Heart Failure in the ED

[

1

Original Research

]

56

2

57

3

58

4

59

5

60

Accuracy of Several Lung Ultrasound Methods for the Diagnosis of Acute Heart Failure in the ED

6 7 8 9 10 11

61 62 63 64 65 66

A Multicenter Prospective Study

12

67

13 14

68 Q24

15 16 17 18

Q2

Q1

Q3

Aurélien Buessler, MD; Tahar Chouihed, MD, PhD; Kévin Duarte, PhD; Adrien Bassand, MD;

69

Matthieu Huot-Marchand, MD; Yannick Gottwalles, MD; Alice Pénine, MD; Elies André, MD; Lionel Nace, MD;

70 71

Déborah Jaeger, MD; Masatake Kobayashi, MD; Stéfano Coiro, MD, PhD; Patrick Rossignol, MD, PhD;

72

and Nicolas Girerd, MD, PhD

73

19

74

20

75

21

Early appropriate diagnosis of acute heart failure (AHF) is recommended by international guidelines. This study assessed the value of several lung ultrasound (LUS) strategies for identifying AHF in the ED. BACKGROUND:

22 23 24

This prospective study, conducted in four EDs, included patients with diagnostic uncertainty based on initial clinical judgment. A clinical diagnosis score for AHF (Brest score) was quantified, followed by an extensive LUS examination performed according to the 4-point (BLUE protocol) and 6-, 8-, and 28-point methods. The primary outcome was AHF discharge diagnosis adjudicated by two senior physicians blinded to LUS measurements. The C-index was used to quantify discrimination. METHODS:

25 26 27 28 29 30 31

39

Among the 117 included patients, AHF (n ¼ 69) was identified in 27.4%, 56.2%, 54.8%, and 76.7% of patients with the 4-point (two bilateral positive points), 6-point, 8-point ($ 1 bilateral positive point), and 28-point (B-line count $ 30) methods, respectively. The C-index (95% CI) of the Brest score was 72.8 (65.3-80.3), whereas the C-index of the 4-, 6-, 8, and 28-point methods were 63.7 (58.5-68.8), 72.4 (65.0-79.8), 74.0 (67.1-80.9), and 72.4 (63.9-80.9). The highest increase in the C-index on top of the BREST score was observed with the 8-point method in the whole population (6.9; 95% CI, 1.6-12.2; P ¼ .010) and in the population with an intermediate Brest score, followed by the 6-point method.

40

CONCLUSIONS:

RESULTS:

32 33 34 35 36 37 38

In patients with diagnostic uncertainty, the 6-point/8-point LUS method (using the 1 bilateral positive point threshold) improves AHF diagnosis accuracy on top of the BREST score.

41 42 43

TRIAL REGISTRY:

44

ClinicalTrials.gov; No.: NCT03194243; URL: www.clinicaltrials.gov. CHEST 2019;

45 46 47

Q7

KEY WORDS:

76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

-(-):---

100 101

acute heart failure; dyspnea; emergency; lung ultrasound

102

48

103

49

104

50 51 52 53 54 55

ABBREVIATIONS:

AHF = acute heart failure; LUS = lung ultrasound From the Emergency Department (Drs Buessler, Chouihed, Bassand, Huot-Marchand, and Jaeger), University Hospital of Nancy, France; Université de Lorraine (Dr Duarte), Institut Elie Cartan de Lorraine, Unité Mixte de Recherche 7502, Vandœuvre-lèsNancy, France; Centre National de la Recherche Scientifique, Institut Elie Cartan de Lorraine, Unité Mixte de Recherche 7502, Vandœuvrelès-Nancy, France; INRIA, Project-Team BIGS, Villers-lès-Nancy, AFFILIATIONS:

France; Université de Lorraine (Dr Chouihed, Duarte, Rossignol, and Girerd), Centre d’Investigations Cliniques Plurithématique 1433, Institut Lorrain du Cœur et des Vaisseaux, Vandoeuvre les Nancy France Groupe choc, INSERM U1116, Faculté de Médecine, 54500 Vandoeuvre les Nancy, France; Emergency Department (Dr Gottwalles), Colmar Hospital, Colmar, France; Emergency Department (Dr Pénine), Charleville-Maizières Hospital, Charleville-Maizières France; Emergency Department (Dr André), Mercy Hospital, Metz, France;

1

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

105 106 107 108 109 110

111 112 113 114 115 116 117 118 119 120 121

Dyspnea is one of the most frequent causes of admission to the ED1 and represents a significant diagnostic challenge for emergency physicians. Acute heart failure (AHF) is one of the most common etiologies of acute dyspnea.2 Guidelines recommend that diagnosis should be made as soon as possible to promptly begin appropriate early treatment.3,4 Prognosis is related to initiation time of specific therapies.5 In-hospital mortality is typically reported to be > 10%6 and has remained stable in the last 30 years.

122 123 124 125 126 127 128 129 130 131 132 133 134 135

Diagnostic approaches include clinical evaluation, chest radiograph, biological tests, and specific biomarkers. Nevertheless, diagnosis remains difficult, especially in ED patients, many of whom feature atypical clinical presentation due to several previous comorbidities and mixed/concomitant etiologies of acute dyspnea.2 Basset et al7 developed the Brest score for the diagnosis of AHF in ED patients. However, this score classified 50% of cases in the intermediate probability group, hence supporting the importance of developing and promoting “new tools”8 that are complementary to clinical scores to achieve a quick

166

diagnosis of AHF in patients admitted for acute dyspnea in the ED.

167 168

Ultrasound has gained widespread use in recent years and is now a highly valuable tool in the ED. Lung ultrasound (LUS) is a quick, reliable, and easy-to-use examination that can improve the diagnostic accuracy for dyspneic patients.9,10 Lichtenstein and Mezière11 further highlighted the advantages of LUS in ICUs for the evaluation of patients with respiratory distress (ie, the BLUE protocol). Several methods have been secondarily proposed to assess pulmonary congestion using different analysis points, interpretation thresholds, and various assessment conditions.12-14 However, all of these studies focused on patients outside of the ED. Given these factors, the current study aimed to evaluate and compare the diagnostic performance of currently available ultrasound protocols for pulmonary congestion assessment (ie, the 4-point [BLUE protocol] and 6-, 8-, and 28-point methods) in patients admitted for acute dyspnea in the ED. The study further aimed to evaluate the diagnostic performance of these methods in patients with intermediate Brest scores (ie, 4-8).

169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190

136

191

137

192

138 139 140 Q8 141 142 143 144 145 146 147 148 149 150

Methods

Ultrasound Methods

193

Study Protocol and Design

Ultrasounds were performed by ultrasound-certified emergency physicians. Twenty-eight-point LUS were performed in all patients: for each point, a B-line grading from 0 to 10 was used. Using the data of this 28-point method, patients were able to be classified Q9 according to four published methods (Fig 1).11–13,16

194

This study is a part of the prospective PURPLE study (CNIL DR2017-098).15 Patients admitted to the ED in four different hospitals, including a university hospital, over a 3-month period were included. All patients aged > 50 years admitted for acute dyspnea for whom the treating physician had diagnostic uncertainty based on his or her initial clinical evaluation were included. Exclusion criteria consisted of traumatic dyspnea and systolic BP < 70 mm Hg. For each patient, the Brest score was calculated,7 and a standardized LUS was performed. All clinical and ultrasound analysis data were collected by the emergency physicians and entered into the Clinical Research Form of the study.

151 153 154 155 156 157 158 159 160 Q5 161 162 Q6 163 164 165

199

 Two scanning sites on each hemithorax:  Second intercostal space, mid-clavicular line  Fourth intercostal space, anterior axillary line  A positive point was defined as the presence of at least three B-lines.  A positive examination was defined, according to the seminal publication,17 by the presence of at least three B-lines on each scanning site.

200

2 Original Research

201 202 203 204 205 206

Six-point method : Intensive Care Unit (Dr Nace), University Hospital of Nancy, France; Department of Cardiology (Dr Kobayashi), Tokyo Medical University, Tokyo, Japan; F-CRIN INI-CRCT (Cardiovascular and Renal Clinical Trialists) (Drs Chouihed, Rossignol, and Girerd), Nancy, France; University of Perugia (Dr Coiro), School of Medicine, Perugia, Division of Cardiology; and Département de Cardiologie (Drs Girerd), Institut Lorrain du Cœur et des Vaisseaux, CHRU Nancy, France. FUNDING/SUPPORT: The authors have reported to CHEST that no funding was received for this study. CORRESPONDENCE TO: Tahar Chouihed, MD, PhD, University Hospital of Nancy, Emergency Department, Nancy, Lorraine France; e-mail: [email protected] Copyright Ó 2019 American College of Chest Physicians. Published by Elsevier Inc. All rights reserved. DOI: https://doi.org/10.1016/j.chest.2019.07.017

197 198

Four-point method (BLUE protocol)11:

12

152

195 196

 Three scanning sites on each hemithorax:  Second intercostal space, mid-clavicular line  Fourth intercostal space, anterior axillary line  Fifth intercostal space, mid-axillary line  A positive point was defined as the presence of at least three B-lines in a given scanning site.  A positive examination was defined, according to the seminal publication,12 by the presence of at least three B-lines on two scanning sites on each hemithorax. Eight-point method13:

207 208 209 210 211 212 213 214 215 216

 Four scanning sites on each hemithorax:  Two anterior points, between the sternum and the anterior axillary line, comprising two scanning sites.  Two lateral points between the anterior and the posterior axillary line, comprising two scanning sites.

[

-#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

]

217 218 219 220

221

276

222

277

223

278

224

279

225

280

226

281

227

282

228

283

229

284

230

285

231

286

232

287

233

288

234

289

235

290 291

237

292

238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269

print & web 4C=FPO

236

293 294 295 296

Figure 1 – A-C, LUS methods: 4- to 28-point method described on a frontal (A) and lateral (B) view. Examples of LUS recordings showing 0 to 3 B-lines (C). LUS ¼ lung ultrasound.  A positive point was defined as the presence of at least three B-lines in a given scanning site. Twenty-eight-point method16,18:  Sixteen points on the right side and 12 points on the left as described in Figure 1.  This examination was used both as a continuous count of overall B-lines as well as in the form of dichotomous variables ($ 15 or $ 30). A positive point was defined as the presence of at least three B-lines in a given scanning site. The examinations were then categorized according to the presence and number of bilateral positive points. We considered two definitions of positive examinations: a positive examination was either defined as at least one positive zone bilaterally (ie, at least one on the right lung and at least one on the left lung) or as at least two positive zones bilaterally. The presence of two positive points on each hemithorax, irrespective of their locations (ie, positive points on the superior part of the right thorax and on the inferior part of the left thorax), qualified for being considered as having “$ 2 bilateral positive points.” Outcome Diagnostic outcome was the final diagnosis at discharge collected from the patients’ medical records. The final diagnosis of the hospital stay was adjudicated by two senior physicians (emergency physician and cardiologist) blinded to the LUS measurements. Sample Size A random sample of 120 patients (60 with AHF and 60 without AHF) was necessary, when the sample C-index was equal to 80%, to achieve a two-sided 95% CI width of 16% (ie, with a lower limit equal to 72% and

a upper limit equal to 88%) according to the Hanley and McNeil method. This setting also allows use of a CI width < 0.18 for a C-index equal to 75% and a CI width of 0.14 for a C-index of 0.85.

297 Q21

298 299 300 301 302

Statistical Analysis

303

All analyses were performed by using R software (the R foundation for Statistical Computation). The two-tailed significance level was set at P < .05.

304

Baseline characteristics are described as mean  SD or median (interquartile range) for continuous variables and frequency (percentage) for categorical variables. Comparison of baseline characteristics according to the AHF and non-AHF groups were conducted by using the nonparametric Wilcoxon test for continuous variables and the c2 or Fisher exact tests for categorical variables. Associations between LUS measurements and AHF were assessed by using logistic regression. ORs with 95% CIs are reported. For certain variables, quasi-complete separation was detected. ORs with CIs were therefore estimated by using a logistic regression model with Firth’s penalized likelihood method. This method provides a solution to the phenomenon of monotone likelihood, which causes parameter estimates of the usual logistic regression model to diverge, with infinite SEs. Individual performance of LUS measurements for diagnosing AHF was assessed by the calculation of the C-index, which is very similar to the area under the curve of the receiver-operating characteristic used on univariable data. In addition, the increase in the C-index was calculated to assess the additional value of LUS measurements in addition to the Brest score for the diagnosis of AHF.

305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324

270

325

271

326

272

Results

273

A total of 117 patients were included, 62% of whom had a hospital discharge diagnosis of AHF (n ¼ 73); only 54% (n ¼ 63) had an AHF diagnosis in the ED (three

274 275

patients with AHF at the ED had a non-AHF discharge diagnosis and 13 had an AHF diagnosis at discharge but not in the ED), however (Table 1). The population was elderly (mean age, 79.6  11.8 years), mainly female

3

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

327 328 329 330

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385

(56%), and frequently had comorbidities. The majority of patients were hospitalized subsequent to ED admission (96%; n ¼ 112), primarily in medical wards (n ¼ 68; 58%); 25% (n ¼ 29) were admitted to ICUs, and only 13% (n ¼ 15) were admitted to a cardiology ward. A majority of patients had an intermediate Brest score (64%; n ¼ 75), both in the AHF group (67%; n ¼ 49) and in the non-AHF group (59%; n ¼ 26). Diagnostic Performances in the Overall Study Population

In a first instance, the Brest score had a good diagnostic value when considered as a continuous variable (C-index ¼ 81.8; 95% CI, 74.2-89.4), which subsequently decreased when using BREST score categories (C-index ¼ 72.8; 95% CI, 65.3-80.3). Among the LUS methods, the 4-point method (two bilateral positive points) had the lowest C-index (63.7; 95% CI, 58.5-68.8), whereas the other methods had very similar C-indices (6-point method for $ 1 bilateral positive point, 72.4 [95% CI, 65.0-79.8]; 8-point method for $ 1 bilateral positive point, 74.0 [95% CI, 67.1-80.9]; and 28-point method for B-lines $30, 72.4 [95% CI, 63.9-80.9]). The 6-point method ($ 1 bilateral positive point) had a specificity near 90% with a relatively low sensitivity (56.2%; 95% CI, 41.1-67.8). The 8-point method ($ 1 bilateral positive point) had a higher specificity (93.2%; 95% CI, 81.3-98.6) and similar sensitivity (54.8%; 95% CI, 42.7-66.5). In contrast, the 28-point method had high sensitivity (B-lines $ 15, 89.0 [95% CI, 79.595.1]; B-lines $ 30, 76.7 [95% CI, 65.4-85.8]) but low specificity (B-lines $ 15, 43.2 [95% CI, 28.3-59.0]; B-lines $ 30, 68.2 [95% CI, 52.4-81.4]). For the 6- and 8-point methods, the use of the $ 1 bilateral positive point threshold yielded a higher C-index as well as a better sensitivity (13% and 6%, respectively) and moderately lower specificity (–4% and –11%). Each method provided significant added value to the Brest score as assessed by changes in the C-index. However, the highest increase in the C-index was observed for the 6-point method (6.7; 95% CI, 0.9-12.5; P ¼ .024) and the 8-point method (6.9; 95% CI, 1.6-12.2; P ¼ .010) (Fig 2, Table 3). Diagnostic Performances With Intermediate Brest Scores

In patients (n ¼ 75) with intermediate Brest scores (4-8), the 4-point method (two positive points bilaterally) had

4 Original Research

386

a C-index of 61.2 (95% CI, 55.3-67.1) and an added value to the Brest score of < 5 as measured by an increase in the C-index. In contrast, the 6- and 8-point methods had a C-index > 70 when considering $ 1 positive point bilaterally (71.8 [95% CI, 62.4-81.2] and 72.7 [95% CI, 63.9-81.5], respectively).

387 388 389 390 391 392 393

Similarly to the results in the overall population, the 6and 8-point methods ($ 1 bilateral positive point) had a specificity near 90% and a sensitivity near 50%. For the 8-point method, the use of the $ 1 bilateral positive point threshold yielded a higher C-index as well as better sensitivity (14% increase) and moderately lower specificity (4% decrease).

394 395 396 397 398 399 400 401

A significant increase in C-index over the BREST score was only identified for the 8-point method (increase in C-index ¼ 10.7; 95% CI, 1.7 to 19.7; P ¼ .020). However, the increase in the C-index with the 6-point method had a very similar point estimate (increase in C-index ¼ 8.9; 95% CI, –0.2 to 17.9; P ¼ .054). Importantly, the 28-point method had a lower increase in the C-index of 6.8 (95% CI, –2.6 to 16.1), which was not statistically significant (P ¼ .16) (Fig 2, Table 3).

402 403 404 405 406 407 408 409 410 411 412

Discussion

413

In the present study, the 6- and 8-point methods were found to be the most relevant LUS methods for establishing an AHF diagnosis in the ED. This result was further confirmed among patients with intermediate Brest scores. In addition, all ultrasound methods (particularly the 6- and 8-point methods) provided a diagnostic added value in addition to the Brest score, both in the whole population (increase in C-index 8point method ¼ 6.9; 95% CI, 1.6-12.2; P ¼ .010) and in patients with intermediate Brest scores (increase in C-index 8-point method ¼ 10.7; 95% CI, 1.7-19.7; P ¼ .020). The main results and techniques used are summarized in Figure 2. Importantly, we identified a somewhat lower C-index for the diagnosis of AHF than that previously reported in a meta-analysis19 in which AHF identified on LUS proved to be a diagnostic variable with discriminatory value (positive LR, 7.4 [95% CI, 4.2-12.8]; negative LR, 0.16 [95% CI, 0.05-0.51]) and for which the authors acknowledged the high statistical heterogeneity for these pooled estimates (I2 ¼ 78% and I2 ¼ 99%, respectively). However, contrary to the aforementioned studies, the current analysis was conducted in the specific setting of “real-life” patients admitted to the ED for whom the treating physician had diagnostic uncertainty based on

[

-#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

]

414 415 416 417 418 419 420 421 422 423 424 425 426 427 Q10

428 429 430 431 432 433 434 435 436 437 438 439 440

441

TABLE 1

] Characteristics of the Study Population

442

Population (N ¼ 117)

443 444 445 446

Q18

Characteristic

No AHF (n ¼ 44)

496 497

AHF (n ¼ 73)

498

No.

Mean  SD or No. (%)

No.

Mean  SD or No. (%)

No.

Mean  SD or No. (%)

P Value

79.6  11.8

44

77.0  13.6

73

81.2  10.3

.088

499 500

Clinical characteristics

501

447

Age, y

117

448

Male sex

117

52 (44)

44

16 (36)

73

36 (49)

.19

503

449

Chronic heart failure

117

19 (16)

44

3 (7)

73

16 (22)

.039

504

450

Chronic pulmonary disease

117

38 (33)

44

18 (41)

73

20 (27)

.16

505 506

451

SBP, mm Hg

117

137.1  25.0

44

131.8  21.2

73

140.3  26.7

.11

DBP, mm Hg

117

73.1  15.5

44

70.6  12.7

73

74.6  16.9

.18

454

Heart rate, beats/min

117

93.6  24.6

44

95.8  21.2

73

92.2  26.5

.25

455

Respiratory rate, beats/min

106

27.3  8.6

39

28.6  10.6

67

26.6  7.3

.54

456

SpO2, %

117

93.5  5.8

44

91.7  8.2

73

94.6  3.4

.045

457

NYHA functional score

117

458

NYHA class III

452 453

507 508 509 510 511 512

51 (44)

20 (45)

459

NYHA class IV

460

Jugular venous distension

117

19 (16)

44

461

502

59 (50)

22 (50)

31 (42)

513

37 (50)

514

1 (2)

73

18 (24)

.0001

515 516

Hepato-jugular reflux

117

19 (16)

44

3 (7)

73

16 (22)

.039

463

Peripheral edema

117

64 (54)

44

18 (41)

73

46 (63)

.026

464

Lungs auscultation

117

465

Crackles

52 (44)

4 (9)

48 (66)

466

Focal auscultatory findings

18 (15)

12 (27)

6 (8)

521 522

462

467

Rhonchi

468

Wheezing

469 470

44

517 518

< .0001

73

23 (20)

13 (29.5)

10 (13)

8 (7)

7 (16)

1 (1)

519 520

523 524

Biology

525

eGFR MDRD, mL/min/1.73 m2

116

60.1  27.5

44

71.6  26.7

72

472

Natremia, mmol/L

115

135.9  5.6

44

135.7  5.9

71

136.0  5.4

.81

527

473

BNP, pg/mL

86

946  1017

27

274  281

59

1,254  1,083

< .0001

528

474

NT-proBNP, pg/mL

15

2,815  3,741

5

575  448

10

3,936  4,183

.13

529

475

Hemoglobin, g/dL

117

12.4  2.0

44

13.2  1.9

73

11.9  2.0

.0004

530

476

Hematocrit, %

113

38.6  5.8

43

40.8  5.4

70

37.2  5.6

.0008

13 (29)

73

471

477 478 479

53.1  25.7

.0004

Radiology Cardiomegaly

117

68 (58)

44

55 (75)

< .0001 < .0001

Pulmonary congestion

117

59 (50)

44

7 (16)

73

52 (71)

481

Pleural effusion

117

35 (30)

44

12 (27)

73

23 (3)

.68

117

39 (33)

44

22 (50)

73

17 (23)

.004

Pulmonary infection

483

Brest score

484

Continuous

485 486

117

44 5.9  2.7

4.1  2.0

7.0  2.4

5 (7)

488

4-8

75 (64)

26 (59)

49 (67)

489

9-15

20 (17)

1 (2)

19 (26)

492 493

537 539

< .0001

540

Categorized 17 (38)

491

535 536 538

22 (19)

490

533 534

73

0-3

487

531 532

480 482

526

541

< .0001

542 543 544

Diagnosis: AHF

545

ED

117

63 (54)

44

At hospitalization discharge

117

73 (62)

44

3 (7) 0

73

60 (82)

< .0001

73

73 (100)

< .0001

546 547 548

LUS

494

549

(Continued)

495

550

5

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

551

TABLE 1

] (Continued)

552

Population (N ¼ 117)

553 554 555

606

Characteristic

No AHF (n ¼ 44)

607

AHF (n ¼ 73)

No.

Mean  SD or No. (%)

No.

Mean  SD or No. (%)

No.

Mean  SD or No. (%)

7.1  1.6

43

6.9  1.7

72

7.3  1.5

608 609

P Value

610

556

Ultrasound quality

115

557

4-point method

117

558

B-line count

< .0001

613

559

$1 bilateral positive point

40 (34)

5 (11)

35 (47)

< .0001

614

560

$2 bilateral positive points

20 (17)

0

20 (27)

< .0001

561 562 563

44

117

612

3.6  3.9

11.3  9.8

44 5.2  5.4

17.0  13.8

46 (39)

5 (11)

41 (56)

< .0001

565

$2 bilateral positive points

29 (25)

0

29 (39)

< .0001

566

8-point method (superomedial point)

567

117

44 15.5  16.6

5.3  5.8

21.6  18.0

$1 bilateral positive point

43 (37)

3 (7)

40 (54)

< .0001

570

$2 bilateral positive points

31 (26)

1 (2)

30 (41)

< .0001

B-line count

573 574 575

117

44

625 626

22.0  21.3

78.5  63.6

90 (77)

25 (57)

65 (89)

< .0001

B-lines $ 30

70 (60)

14 (32)

56 (78)

< .0001

72 (98)

.009

577

Medical ward

68 (58)

32 (73)

36 (49)

578

ICU

29 (24)

7 (16)

22 (30)

633

579

Cardiology ward

15 (12)

1 (2)

14 (19)

634

582 583

73

629 630

Hospitalization

581

40 (91)

628

576

580

44

627

< .0001

B-lines $ 15

112 (95)

623 624

73

57.3  58.6

117

622

< .0001

569

28-point method

619 620 621

B-line count

572

618

73

568

571

617

< .0001

$1 bilateral positive point

564

615 616

73

12.6  12.8

B-line count

611

73

8.4  8.9

6-point method

.21

631 632

635

AHF ¼ acute heart failure; BNP ¼ brain natriuretic peptide; eGFR ¼ estimated glomerular filtration rate; MDRD ¼ Modification of Diet in Renal Disease; LUS ¼ lung ultrasound; NT-proBNP ¼ N-terminal pro-B-type natriuretic peptide; NYHA ¼ New York Heart Association; SBP/DPB ¼ systolic/diastolic blood pressure; SpO2 ¼ blood oxygen saturation.

636 637 638

584

639

585

640

586 587

TABLE 2

Analysis and Following Adjustment on the Brest Score)

588

Univariable Association

589 590 591 592 593 594 595

Variable

OR (CI 95%)

B-line count

1.20 (1.09-1.31)

$ 1 bilateral positive point

7.18 (2.54-20.29)

$ 2 bilateral positive points

597

B-line count

598

$ 1 bilateral positive point

599

$ 2 bilateral positive points

602 603 604 605

Adjusted on Brest Score (Continuous)

P Value

OR (CI 95%)

645

P Value

34.10 (4.46-4381.20)

< .0001 .0002 < .0001

1.22 (1.10-1.36)

.0003

7.49 (2.29-24.53)

.0009

23.96 (2.57-3248.84)

647 648 649

.002

650 651

1.14 (1.07-1.22)

1.17 (1.08-1.26)

9.99 (3.53-28.26) 59.00 (7.84-7559.37)

< .0001

652

.0002

12.08 (3.51-41.53)

< .0001

653

51.15 (6.08-6740.40)

< .0001

654 655

8-point method 1.15 (1.08-1.23)

< .0001

1.15 (1.07-1.24)

$ 1 bilateral positive point

16.57 (4.70-58.38)

< .0001

15.68 (3.87-63.48)

.0001

$ 2 bilateral positive points

30.00 (3.91-229.96)

38.75 (4.19-358.43)

.001

B-line count

.001

656

< .0001

657 658 659 660

See Table 1 legend for expansion of abbreviation.

6 Original Research

643 644

4-point method

6-point method

601

Q19

642

646

596

600

641

] Association Between the Different Lung Ultrasound Techniques and AHF Diagnosis (in Univariable

[

-#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

]

661

716

1.0

1.0

662

717

663

718

664

719

0.8

0.8

665

720 721

667

722

668 669 670

0.6

Sensitivity

Sensitivity

666

0.4

0.6

723 724 725

0.4

671

726

672

727

0.2

0.2

673

728

674

729

675

730

0.0

0.0

676 677

0.0

678

0.4 0.6 1-Specificity

679

Curve 1.a

680

0.2

Brest score 6-points method 28-points method

681 682 683

0.8

1.0

731

0.0

0.2

0.4 0.6 1-Specificity

0.8

732

1.0

733 734

Curve 1.b Brest score 6-points method 28-points method

4-points method 8-points method

735

4-points method 8-points method

736 737 738

684

739

685

740

1.0

1.0

686

741

687

742

0.8

0.8

688

743

689

744

691 692 693 694

0.6

Sensitivity

Sensitivity

690

0.4

745

0.6

746 747 748

0.4

749

695

750

696

751

0.2

0.2

697

752

698

753

699 700

0.0

701

0.2

702 704 705 706 707 708

0.4 0.6 1-Specificity

0.8

1.0

Curve 2.a print & web 4C=FPO

703

754

0.0

0.0

Brest score 6-points method 28-points method

4-points method 8-points method

0.0

0.2

0.4 0.6 1-Specificity

0.8

755

1.0

756 757

Curve 2.b Brest score 6-points method 28-points method

758 759

4-points method 8-points method

760 761

Figure 2 – Receiver-operating characteristic curves for acute heart failure diagnosis (B-line count). See Figure 1 legend for expansion of abbreviation.

709 710 711

Q25 Q22

763 764

his or her initial clinical evaluation. Our results can be summarized as shown in Figure 3.

712 713

Brest Score and AHF

714

Brest score is a clinical score recently developed for AHF diagnosis, with three probability categories: low,

715

762

intermediate, and high. Our study confirmed its good diagnostic capacity when considered as a continuous value analysis, although it was decreased (C-index, 72.8; 95% CI, 65.3-80.3) when dichotomized as risk categories. Indeed, the Brest score efficiently rules out AHF diagnosis for scores < 4 and affirms the diagnosis

7

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

765 766 767 768 769 770

771

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

796

797

798

799

800

801

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

817

818

819

820

821

822

823

824

825

8 Original Research

TABLE 3

] Diagnostic Performance of the Various LUS Techniques in Conjunction With the Brest Score for Pulmonary Congestion Assessment Performance

Variable

Specificity (95% CI)

Sensitivity (95% CI)

C-Index Value of Brest Score and Considered Parameter (95% CI)

C-Index Increase in Addition to the Brest Score (95% CI)

P Value

P Value

Overall population Brest score 81.8 (74.2 to 89.4)

.

.

.

.

72.8 (65.3 to 80.3)

.

.

.

.

B-line count

76.7 (68.2 to 85.1)

.

.

88.1 (82.0 to 94.1)

< .0001

6.3 (1.0 to 11.6)

.020

$ 1 bilateral positive point

68.3 (60.8 to 75.8)

88.6 (75.4 to 96.2)

47.9 (36.1 to 60.0)

86.6 (80.1 to 93.1)

< .0001

4.8 (–0.1 to 9.6)

.053

$ 2 bilateral positive points

63.7 (58.5 to 68.8)

100.0 (92.0 to 100.0)

27.4 (17.6 to 39.1)

85.3 (78.6 to 91.9)

< .0001

3.5 (0.4 to 6.5)

.026

Continuous Categories (0-3, 4-8, 9-15) 4-point method

6-point method B-line count

78.2 (70.1 to 86.4)

.

.

89.1 (83.3 to 94.8)

< .0001

7.3 (1.7 to 12.8)

.010

$ 1 bilateral positive point

72.4 (65.0 to 79.8)

88.6 (75.4 to 96.2)

56.2 (44.1 to 67.8)

88.5 (82.5 to 94.5)

< .0001

6.7 (0.9 to 12.5)

.024

$ 2 bilateral positive points

69.9 (64.2 to 75.5)

100.0 (92.0 to 100.0)

39.7 (28.5 to 51.9)

88.4 (82.6 to 94.2)

< .0001

6.6 (2.3 to 10.8)

.002

B-line count

81.8 (74.3 to 89.3)

.

..

90.6 (85.2 to 96.0)

< .0001

8.8 (2.8 to 14.7)

.004

$1 bilateral positive point

74.0 (67.1 to 80.9)

93.2 (81.3 to 98.6)

54.8 (42.7 to 66.5)

88.7 (82.9 to 94.6)

< .0001

6.9 (1.6 to 12.2)

.010

$ 2 bilateral positive points

69.4 (63.3 to 75.5)

97.7 (88.0 to 99.9)

41.1 (29.7 to 53.2)

88.7 (82.8 to 94.7)

< .0001

6.9 (1.7 to 12.1)

.009

.

.

NA

.

.

8-point method

Patients with intermediate Brest score Brest score Continuous

71.7 (59.9 to 83.6)

.

Q20

4-point method

[ -#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

C-Index Value of the Considered Parameter (95% CI)

Diagnostic Value of LUS Techniques in Addition to the Brest Score

B-line count

75.9 (65.0 to 86.8)

.

.

81.6 (71.7 to 91.5)

< .0001

9.9 (0.1 to 19.6)

.047

$ 1 bilateral positive point

68.7 (59.3 to 78.2)

88.5 (69.8 to 97.6)

49.0 (34.4 to 63.7)

78.5 (67.8 to 89.2)

< .0001

6.8 (–2.1 to 15.7)

.13

$ 2 bilateral positive points

61.2 (55.3 to 67.1)

100.0 (86.8 to 100.0)

22.4 (11.8 to 36.6)

76.5 (65.9 to 87.1)

< .0001

4.8 (0.3 to 9.3)

.037 (Continued)

] 827

826

829

828

831

830

832

833

835

834

837

836

839

838

841

840

842

843

845

844

847

846

849

848

851

850

852

853

855

854

857

856

859

858

861

860

862

863

865

864

867

866

869

868

871

870

872

873

875

874

877

876

879

878

880

913 914 915

C-Index Value of the Considered Parameter (95% CI)

917

78.4 (68.0 to 88.7)

916 918 919 920 921 922 923

935

0.029 8.6 (0.9 to 16.4) $ 2 bilateral positive points

$ 1 bilateral positive point

934

B-line count

933

8-point method

932

B-line count

931

6-point method

930

Variable

929

TABLE 3

928

] (Continued)

927

$1 bilateral positive point

926

$2 bilateral positive points

924 925

LUS is recommended by international guidelines.4 It is reliable, reproducible, quick, and easy to use, which prompted its increasing use in patients with acute dyspnea. Its diagnostic performance was reported to be excellent in a large meta-analysis (sensitivity, 94.1% [95% CI, 81.3-98.3]; specificity, 92.4% [95% CI, 84.2-96.4]) for an AHF diagnosis.21 In addition, Zanobetti et al22 reported that the diagnostic accuracy of LUS is better for AHF than for other etiologies of acute dyspnea and that 30 min of training is sufficient to provide good expertise.23,24 However, in these previous studies, a number of LUS methods were used, such that the indicated method in the aforementioned metaanalysis is unclear. Moreover, a head-to-head comparison of each available method for AHF diagnosis was not conducted. In addition, previous studies typically did not specify if the clinical setting of the patients required the use of LUS. Indeed, it is likely that in patients with very unequivocal clinical pictures, the added value of LUS is moderate. Importantly, to the best of our knowledge, its added value on top of the Brest score, a recent and powerful clinical diagnostic tool, has not been previously assessed.

< .0001

0.020

.009 13.6 (3.4 to 23.8)

10.7 (1.7 to 19.7) < .0001

< .0001

.054

.004 9.6 (3.1 to 16.1)

8.9 (–0.2 to 17.9) 88.5 (69.8 to 97.6)

912

71.8 (62.4 to 81.2)

Performance

911

.

910

Specificity (95% CI)

909

In the current study, the 6- and 8-point methods were the most discriminative LUS tools for identifying AHF in elderly patients (mean age, 79.6 years) in whom the ED physicians perceived diagnostic uncertainty. Importantly, in our study, uncertainty was purely physician driven. This explains why only two-thirds of the population would qualify for uncertainty (ie, intermediate risk of HF) using the Brest score. In this “real-life” clinical setting, the 6- or 8-point method significantly increased the discrimination for AHF diagnosis in addition to the Brest score (Table 3) along with an isolated C-index (ie, not taking into account clinical features) > 70. In addition, the diagnostic performance of LUS was maintained in patients with intermediate BREST scores, which further strengthens the ability of LUS to correctly identify AHF in patients with the most clinical uncertainty. Although the current study reports less evocative C-index values than in previous reports,19,25 it should be emphasized that only patients with true diagnostic uncertainty were considered in this analysis, which

9

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

936 937 938 939 940 941

LUS Methods Using Six or More Scanning Sites

See Table 1 legend for expansion of abbreviations.

908

80.4 (70.3 to 90.4)

907

38.8 (25.2 to 53.8)

906

for scores > 9. However, for patients with an intermediate score (4-8), other complementary tools (biomarkers and/or LUS)9 seemingly appear necessary to improve diagnostic accuracy.20

96.2 (80.4 to 99.9)

905

67.5 (59.6 to 75.3)

904

82.4 (72.6 to 92.2)

Sensitivity (95% CI)

903

< .0001

.020

P Value

902

85.4 (76.4 to 94.3)

901

.

900

53.1 (38.3 to 67.5)

899

.

898

92.3 (74.9 to 99.1)

897

72.7 (63.9 to 81.5)

896

81.0 (71.2 to 90.8)

895

< .0001

894

81.4 (71.9 to 90.8)

893

80.6 (70.4 to 90.8)

892

22.4 (11.8 to 36.6)

891

49.0 (34.4 to 63.7)

890

100.0 (86.8 to 100.0)

889

69.4 (62.5 to 76.3)

888

11.6 (1.9 to 21.4)

887

< .0001

886

83.4 (74.0 to 92.7)

885

.

884

P Value

883

C-Index Increase in Addition to the Brest Score (95% CI)

Diagnostic Value of LUS Techniques in Addition to the Brest Score

882

C-Index Value of Brest Score and Considered Parameter (95% CI)

881

942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990

991

Brest Score

993 994 996 997

Age >65 years Sudden dyspnea Night outbreak Orthopnea Prior CHF episode COPD Myocardial infarction Pulmonary crackles Pitting leg edema ST abnormalities Atrial fibrillation/flutter Maximal score

998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1011

84

1012

82

1013

80

1014

78

1015

76

1016

74

1018

72

1019

70

1020

68

1023 1024 1025 1026

print & web 4C=FPO

1017

6-point LUS method

1047

8-point LUS method

1048 1049 1050 1051

+2 1 2 1 1 2 –2 1 2 1 1 1 15

1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063

81.8 (74.2 to 89.4)

1066 1067

8.9 (–0.2 to 17.9) 6.8 (–2.1 to 15.7)

8 6

1065

10.7 (1.7 to 19.7)

12 10

71.7 (59.9 to 83.6)

1064

IAUC by LUS methods in addition to the Brest score

AUC for Brest score

1010

1022

4-point LUS method Points score

Variables

995

1021

1046

Diagnostic performance of the Brest score and LUS methods

992

1068 1069

6.9 (1.6 to 12.2)

6.7 (0.9 to 12.5)

1070

4.8 (–0.1 to 9.6)

1071 1072

4

1073 1074

2

1075 1076

0

66 Overall Population Intermediate Brest score (4-8)

1077 1078

Overall population: Brest score + LUS ≥ 1 bilat. positive zone Intermediate Brest score (4-8): Brest score + LUS ≥ 1 bilat. positive zone

1079 1080

Figure 3 – Diagnostic performance of the Brest score and LUS methods. See Figure 1 legend for expansion of abbreviation.

1081

1027 1028 1029 1030 1031 1032 1033 1034 1035 1036

1082

could have decreased the diagnostic performance of LUS. In this particular setting, an isolated C-index $ 70% together with a significant increase of 6% to 10% in the C-index suggest a strong and clinically relevant improvement in diagnostic accuracy for AHF in actual clinical settings focusing on the most difficult cases. These results further confirm the strong diagnostic ability of LUS.

1037 1038 1039 1040 1041 1042 1043 1044 1045

LUS Methods Using Four Scanning Sites

The BLUE protocol technique, developed in an ICU by Lichtenstein and Mezière,11 is the most widely used and taught LUS technique. However, in the current study, the diagnostic ability of a 4-point LUS technique for diagnosing AHF was somewhat less than that of other methods which rely on a greater number of scanning

10 Original Research

points (6-28). The BLUE protocol, relying on four anterior scanning sites to identify AHF, may be less effective in ED patients due to the lower severity of patients with dyspnea (and subsequent pulmonary features/lesions) admitted in the ED compared with patients admitted in ICUs. Patients admitted in the ED are likely to exhibit less extensive pulmonary abnormalities than patients admitted in the ICU and may therefore benefit from LUS techniques involving six or more scanning sites.

1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095

Perspectives

LUS is a new helpful tool in the ED as well as in the prehospital setting. Although echocardiography can assess cardiac dysfunction and filling pressures, the latter requires trained practitioners and can be difficult to

[

-#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

]

1096 1097 1098 1099 1100

1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139

Q11

perform in the setting of acute dyspnea. Our results show that LUS using a 6- or 8-point method, as in other reports,26 improves the diagnostic accuracy of AHF in the ED. Notwithstanding, although the specificity of LUS using either a 6- or 8-point method herein was similar to other reports, the sensitivity documented in the current study was only about 50%, which is much lower than the 90.5% (87.4-93) reported by Pivetta et al.12 However, this previous study was performed by an ED group with extensive experience in LUS, which may have resulted in its higher diagnostic performance. In addition, the differences in diagnostic performance could also be partly related to the absence of identification of lung sliding and condensation in the current study. In addition, LUS alone may not be sufficient to fully identify AHF in patients with high diagnostic uncertainty. Nazerian et al27 reported a good diagnostic performance for simplified echocardiography performed by emergency physicians for AHF diagnosis. Other studies also suggest that using the size and collapsibility of the inferior vena cava, or other markers, can improve diagnostic accuracy in dyspneic patients.28-30 Furthermore, Laursen et al31 showed that an algorithm using cardiac, vascular, and LUS resulted in an improved early diagnostic accuracy. Thus, the use of an ultrasound-based algorithm rather than an LUScentered algorithm may be needed to further improve the accuracy of AHF diagnosis. Importantly, studies advocating a multimodal ultrasound approach for improving early diagnostic accuracy do not provide a precise algorithm. We believe that such an algorithm should be validated. It is hoped that the Evaluation of the Feasibility and Accuracy of an Ultrasound Algorithm for Acute Dyspnea Diagnosis in the Emergency Department (EMERALD-US) study will be able to provide reliable evidence regarding an integrated

ultrasound algorithm in the field of acute dyspnea admitted in the ED.32

1156 1157 1158

Limitations

1159

The present prospective multicenter study has certain limitations. First, various ultrasound devices were used as well as various patient positions33 (it is, however, likely that most patients were in a semi-seated position), which could have resulted in some heterogeneity. However, given that LUS is likely to occupy an increasing place in emergency settings, including with various ultrasound devices, in various positions, pragmatic studies such as the current one more aptly reflect this intrinsic heterogeneity. Uncertainty was an inclusion criteria but was purely physician-driven. This could have introduced some heterogeneity in the data as the perception of uncertain situations might vary across physicians. The adjudicated diagnosis used for the current analysis was based on the hospitalization report extracted from the medical record. This diagnosis could have been influenced by the LUS results. However, the final diagnosis was adjudicated by two senior physicians blinded to the LUS measurements.

1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184

Conclusions The current study suggests that LUS using the 8-point/6point method improves AHF diagnosis in addition to the BREST score, especially in patients with intermediate BREST scores. Validated algorithms centered not only on the positive diagnosis of AHF but also on the competing diagnosis of dyspnea (eg, pneumonia) using LUS, vascular ultrasound, and simplified echocardiography could further improve LUS diagnostic accuracy in the ED.

1185 1186 1187 1188 1189 1190 1191 1192 1193 1194

1140

1195

1141

1196

1142

1197

1143

1198

1144

1199

1145

1200

1146

1201

1147

1202

1148

1203

1149

1204

1150

1205

1151

1206

1152

1207

1153

1208

1154

1209

1155

1210

11

chestjournal.org

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

1211 1212

Acknowledgments Q12 Q23

1213

Q13

Financial/nonfinancial disclosures: The authors have reported to CHEST the following: P. R. and N. G. are funded by a public grant overseen by the French National Research Agency (ANR) as part of the second “Investissements d’Avenir” program FIGHTHF [reference: ANR-15-RHU-0004] and by the French PIA project “Lorraine Université d’Excellence” [reference ANR-15-IDEX-04LUE]. P. R. has received board membership fees from Novartis, Relypsa, and Steathpeptides. T. C. reports honoraria from Novartis. N. G. reports honoraria from Novartis, Boehringer, and Servier. None declared (A. Buessler, K.D., A. Bassand, M. H.-M., Y. G., A. P., E. A., L. N., D. J., M. K., S. C.).

Q14

Additional information: The e-Tables can be found in the Supplemental Materials section of the online article.

1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228

Author contributions:

1229 1230

Q15

1231 1232 1233 1234

Q16

1235 1236 1237 1238 1239 1240 Q17 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264

References 1. Hunold KM, Caterino JM. High diagnostic uncertainty and inaccuracy in adult emergency department patients with dyspnea: a national database analysis. Acad Emerg Med. 2018. 2. Ray P, Birolleau S, Lefort Y, et al. Acute respiratory failure in the elderly: etiology, emergency diagnosis and prognosis. Crit Care Lond Engl. 2006;10(3):R82. 3. Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail. 2016;18(8):891-975. 4. Mebazaa A, Yilmaz MB, Levy P, et al. Recommendations on pre-hospital & early hospital management of acute heart failure: a consensus paper from the Heart Failure Association of the European Society of Cardiology, the European Society of Emergency Medicine and the Society of Academic Emergency Medicine. Eur J Heart Fail. 2015;17(6):544-558. 5. Matsue Y, Damman K, Voors AA, et al. Time-to-furosemide treatment and mortality in patients hospitalized with acute heart failure. J Am Coll Cardiol. 2017;69(25):3042-3051. 6. Chouihed T, Buessler A, Bassand A, et al. Hyponatraemia, hyperglycaemia and worsening renal function at first blood sample on emergency department admission as predictors of in-hospital death in patients with dyspnoea with suspected acute heart failure: retrospective observational analysis of the PARADISE cohort. BMJ Open. 2018;8(3):e019557. 7. Basset A, Nowak E, Castellant P, GutGobert C, Le Gal G, L’Her E. Development of a clinical prediction score

for congestive heart failure diagnosis in the emergency care setting: the Brest score. Am J Emerg Med. 2016. 8. Girerd N, Seronde MF, Coiro S, et al. Integrative assessment of congestion in heart failure throughout the patient journey. JACC Heart Fail. 2018;6(4):273285. 9. Pirozzi C, Numis FG, Pagano A, Melillo P, Copetti R, Schiraldi F. Immediate versus delayed integrated point-of-careultrasonography to manage acute dyspnea in the emergency department. Crit Ultrasound J. 2014;6(1):5. 10. Silva S, Biendel C, Ruiz J, et al. Usefulness of cardiothoracic chest ultrasound in the management of acute respiratory failure in critical care practice. Chest. 2013;144(3): 859-865. 11. Lichtenstein DA, Mezière GA. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest. 2008;134(1):117-125. 12. Pivetta E, Goffi A, Lupia E, et al. Lung ultrasound-implemented diagnosis of acute decompensated heart failure in the ED: a SIMEU multicenter study. Chest. 2015;148(1):202-210. 13. Volpicelli G, Mussa A, Garofalo G, et al. Bedside lung ultrasound in the assessment of alveolar-interstitial syndrome. Am J Emerg Med. 2006;24(6):689-696. 14. 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. 15. ClinicalTrials.gov. NCT 03194243. 16. Frassi F, Gargani L, Tesorio P, Raciti M, Mottola G, Picano E. Prognostic value of extravascular lung water assessed with ultrasound lung comets by chest sonography in patients with dyspnea and/ or chest pain. J Card Fail. 2007;13(10): 830-835. 17. 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. 18. Jambrik Z, Monti S, Coppola V, et al. Usefulness of ultrasound lung comets as a nonradiologic sign of extravascular lung water. Am J Cardiol. 2004;93(10):12651270. 19. Martindale JL, Wakai A, Collins SP, et al. Diagnosing acute heart failure in the emergency department: a systematic review and meta-analysis. Acad Emerg Med. 2016;23(3):223-242. 20. Chouihed T, Coiro S, Zannad F, Girerd N. Lung ultrasound: a diagnostic and prognostic tool at every step in the pathway of care for acute heart failure. Am J Emerg Med. 2016;34(3):656-657. 21. Al Deeb M, Barbic S, Featherstone R, Dankoff J, Barbic D. Point-of-care

ultrasonography for the diagnosis of acute cardiogenic pulmonary edema in patients presenting with acute dyspnea: a systematic review and meta-analysis. Acad Emerg Med. 2014;21(8):843-852. 22. Zanobetti M, Scorpiniti M, Gigli C, et al. Point-of-care ultrasonography for evaluation of acute dyspnea in the ED. Chest. 2017;151(6):1295-1301. 23. Chiem AT, Chan CH, Ander DS, Kobylivker AN, Manson WC. Comparison of expert and novice sonographers’ performance in focused lung ultrasonography in dyspnea (FLUID) to diagnose patients with acute heart failure syndrome. Acad Emerg Med. 2015;22(5):564-573. 24. Noble VE, Lamhaut L, Capp R, et al. Evaluation of a thoracic ultrasound training module for the detection of pneumothorax and pulmonary edema by prehospital physician care providers. BMC Med Educ. 2009;9:3.

1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283

25. Wang Y, Shen Z, Lu X, Zhen Y, Li H. Sensitivity and specificity of ultrasound for the diagnosis of acute pulmonary edema: a systematic review and metaanalysis. Med Ultrason. 2018;1(1):32.

1284

26. Pivetta E, Goffi A, Nazerian P, et al. Lung ultrasound integrated with clinical assessment for the diagnosis of acute decompensated heart failure in the emergency department: a randomized controlled trial. Eur J Heart Fail. 2019.

1288

27. 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.

1292

1285 1286 1287 1289 1290 1291 1293 1294 1295 1296 1297 1298 1299

28. Bataille B, Riu B, Ferre F, et al. Integrated use of bedside lung ultrasound and echocardiography in acute respiratory failure: a prospective observational study in ICU. Chest. 2014;146(6):1586-1593.

1300

29. Laffin LJ, Patel AV, Saha N, et al. Focused cardiac ultrasound as a predictor of readmission in acute decompensated heart failure. Int J Cardiovasc Imaging. 2018.

1304

1301 1302 1303 1305 1306

30. Öhman J, Harjola VP, Karjalainen P, Lassus J. Rapid cardiothoracic ultrasound protocol for diagnosis of acute heart failure in the emergency department. Eur J Emerg Med Off J Eur Soc Emerg Med. 2017.

1308

31. Laursen CB, Sloth E, Lambrechtsen J, et al. Focused sonography of the heart, lungs, and deep veins identifies missed lifethreatening conditions in admitted patients with acute respiratory symptoms. Chest. 2013;144(6):1868-1875.

1312

1307 1309 1310 1311 1313 1314

32. ClinicalTrials.gov. NCT03691857.

1315

33. Frasure SE, Matilsky DK, Siadecki SD, Platz E, Saul T, Lewiss RE. Impact of patient positioning on lung ultrasound findings in acute heart failure. Eur Heart J Acute Cardiovasc Care. 2015;4(4):326-332.

1316

1265

1317 1318 1319 1320

12 Original Research

[

-#- CHEST - 2019

FLA 5.6.0 DTD  CHEST2466_proof  5 September 2019  12:13 am  EO: CHEST-19-0454

]