Feasibility and Accuracy of an Ultrasound Algorithm for Acute Dyspnea Diagnosis in the Emergency Department

Overview

The management of chest pain has revolutionized its prognosis, primarily by improving urgent diagnosis of myocardial infarction. Currently, acute dyspnea is twice as frequent as chest pain and its associated mortality is much higher (16% of acute dyspnea admitted to emergency departments (ED) ). Inappropriate treatment of acute dyspnea in the ED is frequent (30%) and is associated with a tripling of intra-hospital mortality after adjustment for confounding factors (2.83, IC 1.48 to 5.41, p=0.002). Other elements have also highlighted the importance of a quick and appropriate acute dyspnea diagnosis: * The 2015 European Guidelines on acute heart failure emphasize the need for appropriate treatment within 90 minutes after the first medical contact. * Inadequate treatment of chronic bronchitis decompensation is associated with a doubling of intra-hospital mortality. * An initiation of antibiotic treatment within 4 hours of admission for pneumonia is recommended. * 30% of pulmonary embolisms are not diagnosed during the initial emergency department visit, whereas their mortality in the absence of treatment is 25%. Lung, venous and (simplified) cardiac ultrasound is associated with improved diagnostic performance in ED. However, no ultrasound algorithm dedicated to emergency physicians has been formally validated. The Blue Protocol (Lichtenstein et al., Chest 2008) has been validated in intensive care patients with very different phenotypes than those admitted to the ED. Pivetta et al. (Chest 2015) proposed an algorithm focused solely for the diagnosis of heart failure, thus not providing a diagnosis for all the other causes of dyspnea in ED. Finally, Zanbonetti et al. (Chest 2017) proposed an "unguided" ultrasound use, notably integrating inferior vena cava evaluation. However, measuring the inferior vena cava is difficult at the start of ED management when patients are in acute respiratory distress.

The investigators aim to assess the feasibility and accuracy of a new pragmatic and original ultrasound algorithm adapted for acute dyspnea diagnosis in the emergency department. This primary objective of this prospective multicenter study is to assess the diagnostic accuracy of an ultrasound algorithm (EMERALD-US) dedicated to emergencies using lung, cardiac and vascular ultrasound for the 3 main dyspnea causes (heart failure, pneumonia and obstructive pulmonary disease exacerbation) in patients with acute non-traumatic dyspnea managed in the emergency department. Ultrasound exams will be blindly read by a centralized core laboratory after the standardized acquisition of all exams by a physician not involved in the care of patients in the ED. The main discharge diagnosis from initial hospitalization (heart failure, pneumonia and obstructive pulmonary disease exacerbation) will be adjudicated by a college of 3 senior physicians (emergency physician, cardiologist and internist) blinded to the use of ultrasound in the ED. The secondary objectives of the study are to: A/ Assess the feasibility of the ultrasound algorithm (EMERALD-US) in emergency departments. B/ Assess the association between the diagnosis obtained from the ultrasound algorithm (EMERALD-US) and the results of additional (laboratory and radiological exams. C/ Assess the diagnostic accuracy of the ultrasound algorithm (EMERALD-US) for less frequent dyspnea causes (pulmonary embolism, pleural effusion). D/ Assess, the diagnostic accuracy of clinical (including BREST and PREDICA score), laboratory and radiological variables. E/ Assess, the improvement in diagnosis accuracy with the ultrasound algorithm (EMERALD-US) on top of the diagnostic accuracy of clinical, laboratory and radiological exams. F/ Assess the association between misdiagnosis (without using ultrasound) and survival at D30.