Rapid evaluation by lung-cardiac-inferior vena cava (LCI) integrated ultrasound for differentiating heart failure from pulmonary disease as the cause of acute dyspnea in the emergency setting

The study protocol was approved by our local ethics committee. From March 2011 to March 2012, 90 consecutive patients admitted to the ED of our hospital with acute dyspnea were enrolled. Patients with acute coronary syndrome or chest injury were excluded from this study. In addition, patients who had acute dyspnea due to neither cardiac nor pulmonary cause were excluded from this study. Within 30 minutes of admission, all enrolled patients received conventional physical examination, rapid lung, cardiac, and inferior vena cava [IVC] ultrasound with a hand-held device (Vscan^®), electrocardiography, blood tests (including brain natriuretic peptide assay), and chest X-ray in the emergency room. This study is being conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from the patient for publication of this report and any accompanying images.

The Vscan^® (GE Healthcare, Japan), a hand-held ultrasound device with a wide-bandwidth phased-array probe (1.7-3.5 MHz), was used in this study[26, 27]. The investigators were unaware of the chest X-ray findings and clinical data of each patient. First, lung ultrasound was performed. Bilateral scanning of the anterior and lateral chest walls was done with the patient in the supine or sitting position. The correct scan was intercostal with maximum extension of the visual pleural line. The chest wall was divided into 8 areas (2 anterior and 2 lateral areas per side), and 1 scan was obtained for each area[3, 14, 28‐30]. The anterior zone of the chest wall was designated from the sternum to the anterior axillary line and then was divided into upper and lower halves (from the clavicle to the third intercostal spaces and from the third space to diaphragm). The lateral zone was positioned from the anterior axillary line to the posterior axillary line and it was also divided into upper and lower halves. The investigators attempted to detect comet tail artifacts fanning out from the lung-wall interface and spreading to the edge of the screen, which was previously named B-lines[14, 28]. According to the definition used in previous reports, lung ultrasound examination is positive if B-lines are found in two or more zones bilaterally of the eight zones assessed[3, 28‐30]. Lung ultrasound examination was always completed within 1 minute. Subsequently, cardiac ultrasound was performed. Global LV systolic function and the severity of mitral or tricuspid regurgitation were estimated visually from images acquired in standard cardiac views, particularly the apical long-axis view combined with the four-chamber view[16, 17, 25, 31]. Preservation of the LV ejection fraction (EF) was defined as an estimated LVEF ≥40%, whereas a LVEF <40% indicated a reduced EF. Color scanning was also performed to assess flow across the mitral and tricuspid valves. Valvular regurgitation was semi-quantitatively assessed on a five-grade scale (none, trivial, mild, moderate, or severe), based on the width of the regurgitant jet at its origin estimated by visual inspection. A positive cardiac ultrasound examination meant that either a presence of moderate to severe mitral regurgitation (MR) in preserved EF subjects or a presence of moderate to severe MR or tricuspid regurgitation (TR) in reduced EF subjects was detected[19‐21, 32‐34]. Finally, ultrasound evaluation of the IVC was examined within 2.0 cm of the IVC-RA junction. The maximum diameter (IVC max) was measured at the end-expiration and minimum diameter (IVC min) was measured at the end-inspiration[35]. The IVC collapsibility index (IVC-CI) was calculated as (IVC max-IVC min)/IVC max[35, 36]. A positive IVC ultrasound examination, according to the definition in previous reports, required an IVC-CI <50% at baseline[23, 24, 35, 37]. The duration of LCI-integrated ultrasound examination was always less than 3 minutes (Figure 1)[14]. The images of lung-cardiac-IVC integrated ultrasound were shown in Figure 2.

Peripheral venous blood samples were obtained from each patient at admission, and then 5 ml of whole blood was placed into a prechilled vacuum tube containing EDTA for subsequent measurement of BNP. Immediately after blood sampling, each tube was placed on ice and centrifuged at 2,500 rpm and 4°C to obtain plasma. Then the plasma BNP level was measured by immunoradiometric assay using an antibody for human BNP (Shionogi Co. Ltd., Tokyo, Japan).

The initial diagnosis was determined for each patient by one or two cardiologists, who performed lung-cardiac-IVC (LCI) integrated ultrasound evaluation within 3 minutes on each patient in the ED. Confirmation that acute dyspnea was due to a cardiac etiology (AHFS) was based on a positive lung ultrasound examination combined with abnormal findings on either cardiac or IVC ultrasound in the ED (Figure 1). To determine the final diagnosis, two cardiologists and one pneumologist, who were blinded to the results of the LCI integrated ultrasound at admission, independently reviewed each patient’s medical records and classified them as having acute dyspnea due to AHFS, a history of HF but acute dyspnea due to a non-cardiac cause, or non-cardiac acute dyspnea. Confirmation of AHFS was based on the generally accepted Framingham criteria (two major criteria or one major and two minor criteria), with corroborative information including the medical history, hospital course (response to diuretics and vasoactive agents, or results of hemodynamic monitoring), and routine laboratory test data (including BNP)[5, 38]. The category of pulmonary acute dyspnea included pulmonary embolism and primary lung diseases (pneumonia, asthma, COPD, pulmonary fibrosis, or acute respiratory distress syndrome), with or without underlying LV systolic dysfunction but with no evidence of decompensated HF at admission.

Patients with acute dyspnea due to AHFS had a BNP level of 622.0 ± 505.3 pg/ml, which was significantly higher than the BNP level of 230.7 ± 208.2 pg/ml in patients with a final diagnosis of pulmonary disease (p < 0.001). In the group with acute dyspnea due to pulmonary disease, the 18 patients with a history of heart failure had a significantly higher BNP level compared to the 19 patients without a history of heart failure (396.7 ± 176.5 vs. 73.4 ± 59.6 pg/ml, p < 0.001). The BNP level of patients with a history of heart failure and dyspnea due to pulmonary disease showed no significant difference from that of the patients with acute dyspnea due to AHFS (396.7 ± 176.5 vs. 622.0 ± 505.3 pg/ml; p = 0.069). In addition, the BNP level of patients with ARDS (n = 5) showed no significant difference from that of patients with acute dyspnea due to AHFS (369.5 ± 246.3 vs. 622.0 ± 505.3 pg/ml; p = 0.277). The ability of BNP to differentiate AHFS from pulmonary disease was assessed by ROC analysis. The area under the ROC curve for differentiating AHFS from pulmonary disease with BNP was 0.750 (95% confidence interval: 0.698 to 0.804). A BNP value of 663.2 pg/ml had a sensitivity of 37.0%, specificity of 97.2%, negative predictive value of 50.7%, and positive predictive value of 95.2% for differentiating AHFS from pulmonary disease.

In Table 2, the sensitivity, specificity, negative predictive value (PV), positive PV, and total accuracy for differentiating AHFS from pulmonary disease in emergency patients with acute dyspnea are presented for Framigham criteria (two major or one major and two minor criteria), BNP (cut-off value: 100 pg/mL), lung ultrasound, both lung ultrasound and BNP (cut-off value: 100 pg/mL), IVC collapsibility (cut-off value: 50%), either MR or TR (≥ moderate), both preserved EF and MR (≥ moderate), both reduced EF and either MR or TR (≥ moderate), and LCI integrated ultrasound. Comparing these methods showed that LCI integrated ultrasound had the highest specificity (91.9%), negative PV (91.9%), positive PV (94.3%), and total accuracy (93.3%). While lung ultrasound alone had the highest sensitivity (96.2%), its specificity was much lower (54.0%). A reduced EF showed the lowest sensitivity and lowest total accuracy (26.4% and 51.1%, respectively), while BNP at a cut-off value of 100 pg/mL had the lowest specificity (35.1%).