Article, Pulmonology

Clinical application of rapid B-line score with lung ultrasonography in differentiating between pulmonary infection and pulmonary infection with acute left ventricular heart failure

a b s t r a c t

Background: We have found that there are usually 2 causes of acute dyspnea in our emergency department:

(1) Pulmonary infection only and (2) pulmonary infection in the setting of acute left ventricular heart failure (LVHF). These conditions are sometimes difficult to differentiate. Lung ultrasonography (LUS) is easily performed at the bedside and provides accurate information for diagnosis. In this study, we propose a simple B-line score to allow rapid differential diagnosis between these 2 lung conditions.

Methods: A prospective, single-blind trial was conducted on 98 patients with acute dyspnea in the emergency de- partment. Lung ultrasonography and transthoracic echocardiography were performed within 30 minutes after enrollment. The final clinical diagnosis was recorded for all patients. Using the Bedside lung ultrasound in Emer- gency protocol, we recorded the number of B lines at 4 standardized points. Based on the theory of Lichtenstein, scores of 1, 2, 3, and 4 were categorized by the number of B lines on a static screen (0 to b 3, 3 to b 6, 6 to b 8, and

>=8, respectively). The B-line score of 4 Bedside Lung Ultrasound in Emergency protocol points was recorded, and the total B-line score was calculated. receiver operating characteristic curves were used to evaluate the accuracy of the rapid Ultrasound measurements for the final clinical diagnosis.

Results: In our study, 27 patients were diagnosed with pulmonary infection and acute LVHF. The total number of B

lines and the B-line score in patients with pulmonary infection in the setting of acute LVHF were 24.2 +- 2.5 and

11.5 +- 1.5, respectively, which were significantly higher than those in patients with pulmonary infection (12.5 +- 6.4 and 7.2 +- 1.9) (P = .000). In patients with pulmonary infection and acute LVHF, the effective diag- nostic value of left ventricular ejection fraction and the total B-line score were similar (area under the ROC curve: 0.986 vs 0.962, P = .2607). The cutoff value of the total B-line score was 8, with a sensitivity of 80.7% and a spec- ificity of 100%. A combination of LUS and echocardiography might improve the diagnostic accuracy (area under the ROC curve: 0.994; 95% confidence interval, 0.981-1.000; P = .000).

Conclusions: This simple B-line score with LUS can help make a rapid differential diagnosis between pulmonary infection and pulmonary infection with acute LVHF. The diagnostic accuracy may be enhanced when used in con- junction with echocardiography.

(C) 2015

Introduction

Acute dyspnea is a distressing condition, and its cause is sometimes difficult to identify [1]. Care of patients with dyspnea in the emergency department (ED) can be challenging, as traditional physical examina- tion techniques and laboratory tests may be misleading and result in an incorrect diagnosis. Initial mistakes may lead to disastrous outcomes.

* Corresponding author at: 100730, No. 2, Chongwenmennei St, Dongcheng District, Beijing.

E-mail addresses: [email protected], [email protected] (N. Ding).

Therefore, Initial diagnosis and care must be accurate to optimize pa- tient outcome. Lung ultrasonography (LUS) is easily performed at the bedside by the physician and provides accurate information of diagnos- tic and therapeutic relevance; therefore, it is considered a “visual stethoscope.”

We have found that there are usually 2 causes of acute dyspnea in our department, especially in elderly patients in winter: (1) pulmonary infec- tion and (2) pulmonary infection in the setting of acute left ventricular heart failure (LVHF). A rapid and correct diagnosis is difficult to make on the basis of symptoms and physical examination only. In 2008, Lichtenstein and his team [2] proposed the Bedside Lung Ultrasound in Emergency

http://dx.doi.org/10.1016/j.ajem.2015.10.050

0735-6757/(C) 2015

Z. Liu et al. / American Journal of Emergency Medicine 34 (2016) 278281 279

(BLUE) protocol, which provides a step-by-step diagnosis of the main causes of acute respiratory failure–that is, the 6 diseases seen in 97% of pa- tients in the ED–providing an overall accuracy of 90.5%. However, although the BLUE protocol can distinguish between pulmonary edema and pneu- monia, it cannot distinguish between pulmonary edema and interstitial pneumonia. The B line is the core measurement of LUS, which moves in concert with Lung sliding. In practice, counting the number of B lines is dif- ficult and time consuming. Therefore, in the present study, we attempted to devise a simple and rapid B-line score based on the BLUE protocol to differ- entiate pulmonary infection from pulmonary infection with acute LVHF.

Materials and methods

Study subjects and design

This was a prospective study. The study protocol was approved by the local institutional review boards in accordance with the Declaration of Helsinki.

Patients received routine care and treatment according to our standard clinical practice, and this was not affected by participation in the study. In total, 98 patients with acute dyspnea admitted to our ED from November 2014 to April 2015 were enrolled. Enrollment took place in the ED, and im- mediately after enrollment, the general clinical data of patients were col- lected. Bedside LUS and transthoracic echocardiography examination were performed within 30 minutes of enrollment by the attending physi- cians, who each had at least 5 years of experience of point-of-care emergen- cy Ultrasonography . The physicians were aware of the presenting symptoms but were blinded to the other clinical information. The number and score of B lines for 4 BLUE protocol standardized points (Fig. 1) and the left ventricular ejection fraction (LVEF) were the main results of the point-of-care US that were recorded. At the same time, routine clinical indi- cators such as the blood levels of troponin T (TNT), B-type natriuretic pep- tide (BNP), and C-reactive protein (CRP) were recorded for the final diagnosis. We followed up all patients and recorded the final clinical diag- nosis, which was determined by 3 attending doctors in the ED who each had at least 6 years of experience. Diagnosis was based on a combination of clinical assessments, laboratory tests, and imaging scans. Patients with rapid arrhythmia or Severe bradycardia, conduction disturbance, acute cor- onary syndrome, acute pulmonary embolism, Hypertensive crisis, cardiac tamponade, peripartum cardiomyopathy, anemia, kidney dysfunction, hy- pothyroidism or hyperthyroidism, and iatrogenic causes were excluded from the study.

Methods

A multiprobe machine (GE Co, Wuxi City, Jiangsu, China) was used. Lung ultrasonography was performed using a 4- to 8-MHz linear

Fig. 1. The 4 standardized BLUE points. Two hands are placed as shown (size equivalent to the patient’s hands), with the upper hand touching the clavicle and thumbs excluded; this placement corresponds to the location of the lung. The upper BLUE point is at the middle of the upper hand, and the lower BLUE point is at the middle of the lower hand.

probe to recognize B lines and a 3.5- to 5-MHz curved array probe to find the consolidation or pleural effusion. The lung was examined by longitudinal and oblique scans at 4 points, following the BLUE protocol. The examination was performed with the patient in a supine or semirecumbent position, depending on the patient’s clinical condition. In accordance with the conclusion of the study of Volpicelli et al [3], we counted the number of B lines at 4 points only, following the BLUE protocol: the upper BLUE points and the lower BLUE points (Fig. 1). To allow rapid examination, we devised a simple B-line scoring system based on the theory of Lichtenstein, using 3-4 B lines called septal rockets, which correlate with Keler B lines [4], and 6-8 B lines called ground-glass rockets, which correlate with ground-glass areas [5]. The scoring system is a 4-point scale, with the points 1, 2, 3, and 4 corre- sponding to the number of B lines: 0 to b 3, 3 to b 6, 6 to b 8, and >= 8, re- spectively, on a static screen. We recorded the B-line score of every BLUE point and calculated the total B-line score.

We performed echocardiography using a 2- to 5-MHz phased array probe. The value of LVEF was measured by the Simpson method. The mean of 5 measurements taken was used for statistical analysis.

Analysis

Normally distributed data are expressed as mean +- SD and nonnormally distributed data as median (minimum-maximum). The unpaired Student t test and the nonparametric test were used to com- pare normally distributed and nonnormally distributed data, respec- tively. Fisher exact test was used for the comparison of noncontinuous variables. P b .05 indicates statistical significance. The diagnostic perfor- mance of US measurements was assessed by the receiver operating characteristic curve. Calculations were performed using SPSS software (version 22.0; IBM Inc, Armonk, NY).

Results

Baseline characteristics

A total of 58 patients were included in the study. There were 31 pa- tients with pulmonary infection: 16 cases of pneumonia, 6 cases of Acute bronchitis, 3 cases of lung interstitial fibrosis complicated by infection, 5 cases of infection with chronic obstructive pulmonary disease (COPD), and 1 case of lung metastasis carcinoma complicated by infection. There were also 27 patients with pulmonary infection and acute LVHF: 6 cases with pneumonia, 18 cases with acute bronchitis, and 3 cases with acute exacerbation COPD. Mean age of the patients was

76.6 +- 10.1 years, and 60.3% of the patients were male. Mean (min- max) number of and score for B lines were 20.0 (0.0-29.0) and 9.5 (4.0-14.0), respectively. Based on the final clinical diagnosis, 31 patients were assigned to the pulmonary infection group and 27 patients to the pulmonary infection with acute LVHF group. Patient data are presented in the Table. There were significant differences between the 2 groups with respect to blood pressure; levels of TNT, BNP, and CRP; value of LVEF; number of B lines; and the total B-line score (all P b .001). There was no significant difference in age, sex, or patient prognosis between the 2 groups.

Main results

Pulmonary infection is a common cause of acute LVHF. The diagnosis of pulmonary infection in the setting of acute LVHF needs an appropri- ate combination of clinical assessment, laboratory tests, and imaging scans. Left ventricular ejection fraction, a measure reflecting the real- time left ventricular systolic dysfunction, is widely used as a diagnostic basis of acute LVHF. Therefore, LVEF was used as an important reference indicator for the accuracy of final diagnosis in this study. As mentioned above, traditional physical examination techniques may be misleading and lead to incorrect diagnosis, whereas laboratory tests may delay

280 Z. Liu et al. / American Journal of Emergency Medicine 34 (2016) 278281

Table

Baseline characteristics of patients stratified by final clinical diagnosisa

Characteristic

Group

P

Pulmonary infection (n = 31)

Pulmonary infection and acute LVHF (n = 27)

Age, y

76.9 +- 11.0

76.3 +- 9.2

.821

Female

13 (41.9)

10 (37.0)

.704

At admission

Duration of symptoms, d

3.2 +- 2.1

1.9 +- 0.8

.003

Heart rate, beats/min

102.4 +- 17.9

100.0 +- 18.6

.631

Systolic blood pressure, mm Hg

114.5 +- 21.0

148.6 +- 24.1

.000

Diastolic blood pressure, mm Hg

70.3 +- 10.2

82.6 +- 10.4

.000

Ultrasonography measurementsb No. of B lines at R1 point

3.0 (0.0-6.0)

6.0 (4.0-8.0)

.000

No. of B lines at R2 point

4.0 (0.0-7.0)

6.0 (3.0-8.0)

.000

No. of B lines at L1 point

3.0 (0.0-6.0)

6.0 (3.0-8.0)

.000

No. of B lines at L2 point

4.0 (0.0-7.0)

6.0 (3.0-8.0)

.000

Total no. of B lines

12.5 +- 6.4

24.2 +- 2.5

.000

Total B-line score

7.2 +- 1.9

11.5 +- 1.5

.000

Positive pleural effusion

20 (64.5)

13 (48.1)

.209

Positive consolidation

20 (64.5)

13 (48.1)

.209

Value of LVEF, %

56.8 +- 6.1

34.8 +- 6.9

.000

Laboratory values

Positive TNT, ng/mL

3 (9.7)

16 (59.3)

.000

CRP, mg/L

97.0 (16.0-266.0)

17.0 (9.0-61.0)

.000

BNP, pg/mL

180.0 (11.0-1669.7)

798.5 (34.6-5000.0)

.000

Outcomes

Death

8 (25.8)

9 (33.3)

.530

Hospitalization time, d

12.0 (1.0-60.0)

11.0 (2.0-53.0)

.166

IQR, interquartile range.

a Data are expressed as mean +- SD, median (min-max), or number (percentage), as appropriate.

b R1 point, right upper BLUE point; R2 point, right lower BLUE point; L1 point, left upper BLUE point; L2 point, left lower BLUE point.

treatment. In this study, we compared only the rapid results achieved by ultrasonography.

The effective diagnostic values of LVEF and total B-line score in pa- tients with pulmonary infection and acute LVHF were comparable (area under the receiver operating characteristic curve [AUC]: 0.986 vs 0.962, P = .2607) (Fig. 2). The cutoff value of LVEF was 42%, with a sensitivity of 100% and a specificity of 88.9%. The cutoff value of the total B-line score was 8, with a sensitivity of 80.7% and a specificity of 100%.

The AUC for the combination of total B-line score plus LVEF was 0.994,

which was greater than that for either the total B-line score or the LVEF individually. Therefore, there were no significant differences between them (AUCcombination vs AUCtotal B-line score, P = .0902; AUCcombination vs AUCLVEF, P = .5092). The cutoff values of the combination were 8% and 62%, with a sensitivity of 96.67% and a specificity of 96.3%.

Sensitivity

1-Specificity

Fig. 2. Receiver operating characteristic curves for the diagnosis of pulmonary infection with acute LVHF. Area under the curve is 0.986 for LVEF (95% confidence interval [CI], 0.913-1.000; P b .0001), 0.962 for total B-line score (95% CI, 0.875-0.993; P b .0001), and

0.993 (95% CI, 0.923-1.000; P b .0001) for the combination of both.

Discussion

This study demonstrated that LUS is a useful tool for the differential diagnosis of pulmonary infection and pulmonary infection with acute LVHF. Our data showed that the total B-line score of the pulmonary in- fection with acute LVHF group was significantly higher than that of the pulmonary infection only group (11.5 +- 1.5 vs 7.2 +- 1.9, P b .001). The AUC of the total B-line score was 0.962 for pulmonary infection with acute LVHF. These findings indicate that the B-line score can be used to make a rapid differential diagnosis for patients with pulmonary infec- tion or patients with pulmonary infection and acute LVHF.

Ultrasonographic technology has been rapidly integrated into emer- gency medicine over the past decade [6]. The findings of the present study are comparable with those of previous studies; however, our study was conducted based on a rapid B-line score in a number of pa- tients in the ED, which is the first such study to our knowledge. Previous studies on LUS performed on patients in the intensive care unit used B lines as a marker of lung rocket sign or alveolar-interstitial syndrome [2,5,7,8]. It has been shown that multiple B lines 7 mm apart (B7 lines) are caused by interlobular septa, characterizing alveolar-interstitial syn- drome, whereas B lines 3 mm apart (B3 lines) are caused by ground- glass opacity, characterizing alveolar edema [9,10]. In their study of the correlation between B lines and extravascular lung water quantified by the indicator method, Agricola et al [11] found that the number of B lines provided reliable information on interstitial pulmonary edema. In practice, we found that recognizing B7 and B3 lines was difficult because of the nonuniformity of the B-line spacing, and counting the number of B line accurately was time consuming because of their movement with lung sliding. Therefore, we devised an easy-to-use B-line score for the rapid differential diagnosis of cardiogenic and pulmonary causes of dys- pnea in patients in ED.

The B line is the artifact produced by a high-acoustic impedance gra-

dient between air and fluid corresponding to physiopathologic patterns that have been in clinical use since 1994 [8]. However, Volpicelli et al [3] confirmed that B lines can be present in the normal lung and that this finding was mainly confined to the laterobasal areas. In acute LVHF, the transudate under pressure is pushed along the interlobular septa

Z. Liu et al. / American Journal of Emergency Medicine 34 (2016) 278281 281

against gravity, up to the anterior chest wall, explaining the symmetric anterior interstitial edema seen [12]. B lines are often detected in a focal chest area surrounding the consolidation and are found in interstitial pulmonary infection or in decompensated COPD [2,9,13,14]. As a result, inflammatory transudates in different pulmonary regions can be detect- ed by B lines. In our group of patients with pulmonary infection and acute LVHF, we found that the average number of B lines was the same at each of the 4 points of the BLUE protocol. By contrast, in our group of patients with pulmonary infection only, the average number of B lines was not the same at the 4 points. Therefore, according to the above theory, the number of B lines is increased if both pressure and in- flammation causes are collaborating at the same time. Our study has dem- onstrated that not only was the total B-line score significantly higher in the pulmonary infection and acute LVHF group compared with the pul- monary infection only group (11.5 +- 1.5 vs 7.2 +- 1.9 P b .001) but also the number of B lines at the 4 anterior BLUE-protocol points was signifi- cantly different between the 2 groups (all P b .001).

The differential diagnosis for our patients was of pulmonary infec- tion with acute LVHF or pulmonary infection only; however, the BLUE protocol would have classified these as pulmonary edema or pneumo- nia, respectively. The BLUE protocol excluded interstitial pneumonia also. For pulmonary edema, the B profile had 97% sensitivity and 95% specificity in the BLUE protocol [2]. In our study, the total B-line score for pulmonary edema had a sensitivity of only 80.7% because we quan- tified the B line in a way that is more detailed than the BLUE protocol and we did not exclude interstitial lung disease.

The LVEF value is a routine clinical indicator for acute LVHF. In prac- tice, if LVEF is less than 50%, there is a possibility of left heart failure. In our study, the cutoff value of LVEF was 42% because we had 3 patients with sepsis or septic shock who had damage to the myocardium. Al- though the effective diagnostic values of LVEF and total B-line score in patients with pulmonary infection and acute LVHF were comparable, the sensitivity of the total B-line score was 80.7%, which was less than the sensitivity of LVEF. As mentioned before, the number of B lines was affected by both the inflammatory reaction of pneumonia and the cardiac function. The value of LVEF as the clinical “criterion standard” for heart failure is mainly decided by cardiac systolic function. As a re- sult, the sensitivity of LVEF for diagnosis of acute LVHF was increased.

Point-of-care US examination can be used for assessing the real-time situation of the lungs and heart. Theoretically, the combination of LUS and echocardiography should improve the diagnostic accuracy for pul- monary infection in the setting of acute LVHF. The AUC for the combina- tion of the total B-line score and LVEF was greater than that for the total B-line score and LVEF individually, but this was not statistically signifi- cant, which may be due to the small sample used in the study. Therefore, further studies are needed to confirm the diagnostic accuracy of this combination for pulmonary infection with acute LVHF.

The study has some limitations. Firstly, 3 cases of lung interstitial fi- brosis complicated by infection were included in the pulmonary infec- tion only group. The total B-line score of these patients was no more than 12; thus, a total B-line score of 8-12 may be regarded as an appro- priate reference range for the diagnosis of pulmonary infection with

LVHF. In addition, the B-line score must not be less than 1 at any point in our protocol. If the total B-line score is more than 12, pulmonary in- terstitial disease must be excluded. Furthermore, transthoracic echocar- diography allows exclusion of acute LVHF but fails to identify diastolic dysfunction, which may be the main cause of heart failure in approxi- mately 50% of cases [15]. The normal value of LVEF has an influence on the final clinical diagnosis. As a result, patients with acute cardiac di- astolic dysfunction but with normal LVEF might have been placed into the pulmonary infection group, so the sensitivity and specificity of the total B-line score would then be affected. Finally, this investigation was a single-center observational study with a small sample size, and confounding factors were difficult to avoid.

Conclusions

The results of this study show that our simple B-line score can help make a rapid differential diagnosis between pulmonary infection and pulmonary infection with acute LVHF in the ED. A total B-line score of 8-12 is an appropriate reference range for diagnosis of pulmonary infec- tion with acute LVHF. When the score is greater than 12, pulmonary in- terstitial disease must be excluded. The diagnostic accuracy of LUS may be enhanced when used in conjunction with echocardiography.

References

  1. Ray P, Birolleu S, Lefort Y, Becquemin M, Beigelman C, Isnard R, et al. Acute respira- tory failure in the elderly: etiology, emergency diagnosis and prognosis. Crit Care 2006;10:R82 1-2.
  2. Lichtenstein D, Gilbert A, Meziere G. Relevance of lung ultrasound in the diagnosis of acute respiratory failure–the BLUE protocol. Chest 2008;134:117-25.
  3. Volpicelli G, Caramello V, Cardinale L. Dection of sonographic B-lines in patients with normal lung or radiographic alveolar consolidation. Med Sci Monit 2008;14: 122-8.
  4. Kerley P. Radiology in heart disease. Br Med 1933;2:594-6.
  5. Lichtenstein D. Lung ultrasound in the critically ill. Ann Intensive Care 2014;4:1-12.
  6. Perera D, Maihot D, Riley D, Riley D, Mandavia D. The RUSH Exam 2012: rapid ultra- sound in shock in the evaluation of the critically ill patient. Ultrasound Clin 2012;7: 255-78.
  7. Copetti R, Soldati G, Copetti P. Chest sonography: a useful tool to different acute car- diogenic pulmonary edema from acute respiratory distress syndrome. Cardiovasc Ultrasound 2008;6:1-10.
  8. Lichtenstein D. FALLS-protocol: lung ultrasound in hemodynamic assessment of shock. Heart Lung Vessels 2013;5:142-6.
  9. Lichtenstein D, Meziere G, Biderman P, Gepner A, Barre O. The comet-tail artifact: an ultrasound sign of alveolar-interstitial syndrome. Am J Respir Crit Care Med 1997; 156:1640-6.
  10. Hasan A, Makhlouf H. B-line: transthoracic chest ultrasound signs useful in assess- ment of interstitial lung diseases. Ann Thorac Med 2014;9:99-103.
  11. Agricola E, Bove T, Oppizzi M. Ultrasound comet-tail images: a maker of pulmonary edema. Chest 2005;127:1690-5.
  12. Safran NC. Pulmonary edema. Physiol Rev 1974;54:678-811.
  13. Chavez MA, Shams N, Ellington LE, Naithani N, Gilman RH, Steinhoff MC, et al. Lung ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta- analysis. Respir Res 2014;15:50-8.
  14. Nazerian P, Volpicelli G, Vanni S, Gigli C, Betti L, Bartolucci M, et al. Accuracy of lung ultrasound for the diagnosis of consolidation when compared to chest computed to- mography. Am J Emerg Med 2015;33:620-5.
  15. Cadinale L, Volpicelli G, Binello F, Garofalo G, Priola SM, Veltri A, et al. Clinical appli- cation of lung ultrasound in patients with acute dyspnea: differential diagnosis be- tween cardiogenic and pulmonary causes. Radiol Med 2009;114:1053-64.