The diagnostic value of lung ultrasound combine N -terminal pro -brain natriuretic peptide testing in patients with acute dyspnea

WU Wen; NIE Kun; MENG Jun; LI Min; CHEN Tao; XU Jian-qiang; CHANG Rui

doi:10.3969/j.issn.1002-1949.2016.06.007

Abstract：Objective The aim of this study was to study the diagnostic accuracy of lung ultrasound combined with N-terminal pro-brain natriuretic peptide ( NT-proBNP) in differentiating cardiogenic dyspneafrom non -cardiogenic dyspnea. Methods This was a prospective, observational study of 166 adult patients admitted with acute dyspnea. All patients underwent lung ultrasoundexaminations(the presence of a comet-tail sign), left ventricular ejectionfraction(LVEF)by echocardiography examination,NT - proBNP testing and clinical assessment ( according to the Boston criteria) . Two independent physicians reviewed all the medical records to establish the aetiologic diagnosis of dyspnea,served as the criterion standard. The patients were divided into two groups,namely cardiogenic dyspnea group and non - cardiogenic dyspnea group. The independent risk factors of the cardiogenic dyspnea were analyzed by univariate and multivariable Logistic regression. Receiver operating characteristic curve( ROC) was plotted to evaluate the value of ultrasound lung comets( ULCs) alone and combined with NT - proBNP in predicting acute decompensated heart failure ( ADHF ) . Results Compared with non - cardiogenic dyspnea group, the ULCs, NT - proBNP and Boston criteriain cardiogenic dyspnea group were significantly elevated, but LVEF was significantly worse(P <0. 01). Correlation analysis showed ULCs were significantly positively correlated with NT-proBNP(P<0. 01), and ULCs significantly negatively correlated with LVEF(P <0. 01). It was shown in the multivariable Logistic regression analysis that ULCs at admission[odds ratio(OR)1. 409, 95% confidence interval (95%CI) 1. 147 ~ 1. 732, P = 0. 001 ] was the most important independent prognostic factors for diagnosing the cardiogenic dyspnea. Receiver operating characteristic ( ROC) showed the optimal critical value of ULCs was 16 lines,the sensitivity was 82. 02%,and specificity was 89. 61%. ULCs was better in predicting the cardiogenic dyspnea than NT-proBNP,LVEF and Boston criteria in acute decompensated heart failure patients. The AUC of ULCs combined NT-proBNP was 0. 937,the sensitivity was 82. 02%, and specificity was 90. 91%. Conclusion Lung ultrasound alone or in combination with NT-proBNP is an effective tool for predicting the cardiogenic of dyspnea. A ULCs≥16 cutoff could be considered as a quick and reliable tool for diagnosing acute decompensated heart failure.