Which method is best for an early accurate diagnosis of acute heart failure? Comparison between lung ultrasound, chest X-ray and NT pro-BNP performance: a prospective study.
Intern Emerg Med. 2016 Jul 11;
Authors: Sartini S, Frizzi J, Borselli M, Sarcoli E, Granai C, Gialli V, Cevenini G, Guazzi G, Bruni F, Gonnelli S, Pastorelli M
Acute heart failure is a common condition among adults presenting with dyspnea in the Emergency Department (ED), still the diagnosis is challenging as objective standardized criteria are lacking. First line work-up, other then clinical findings, is nowadays made with lung ultrasound imaging study, chest X-ray study and brain natriuretic peptide (BNP) level determination; however, it is not clear which is the best diagnostic test to be used and whether there is any real benefit for clinical judgement. We set up this study to compare the performances of these three diagnostic tools; furthermore, we combined them to find the best possible approach to dyspneic patients. This is a prospective observational study based in the ED. We enrolled adults presenting with dyspnea not trauma-related, they underwent lung ultrasound, and chest X-ray studies, and NT pro-BNP level determination. Then we compared the results with the diagnosis of acute heart failure established by an independent panel of experts. 236 patients were enrolled in the study. We find sensitivity and specificity for lung ultrasound of 57.73 and 87.97 %, for chest X-ray 74.49 and 86.26 %, for NT pro-BNP 97.59 and 27.56 %, respectively. Combining together the chest X-ray and lung ultrasound, we find the best overall performance with 84.69 % sensitivity, 77.69 % specificity and 87.07 % negative predictive value. From our results, we could not identify the "best test" to diagnose acute heart failure in an emergency setting, although we could suggest that a stepwise workup combining chest X-ray and lung ultrasound at first, then for those negative, a determination of NT pro-BNP assay would be a reasonable approach to the dyspneic patient.
PMID: 27401330 [PubMed - as supplied by publisher]