Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

Predictive criteria for acute heart failure in emergency department patients with acute dyspnoea

the PREDICA study

Roncalli, J.a; Picard, F.b; Delarche, N.c; Faure, I.d; Pradeau, C.e; Thicoipe, M.e; Galinier, M.a; Charpentier, S.f

European Journal of Emergency Medicine: December 2019 - Volume 26 - Issue 6 - p 400–404
doi: 10.1097/MEJ.0000000000000622
Original Articles

Objectives The early identification of patients with Acute Heart Failure Syndrome (AHFS) among patients admitted to the Emergency Department (ED) with dyspnoea can facilitate the introduction of appropriate treatments. The objectives are to identify the predictive factors for AHFS diagnosis in patients with acute dyspnoea (primary objective) and the clinical ‘gestalt’ (secondary objective) in ED.

Methods PREDICA is an observational, prospective, multicentre study. The enrolment of patients admitted to the ED for nontraumatic acute dyspnoea and data collection on admission were recorded by the patient’s emergency physician. The AHFS endpoints were assessed following a duplicate expert evaluation by pairs of cardiologists and emergency physicians. Step-by-step logistic regression was used to retain predictive criteria, and the area under the receiver operating characteristic (ROC) curve of the model was constructed to assess the ability of the selected factors to identify real cases. The probability of AHFS was estimated on a scale from 1 to 10 based on the emergency physician’s perception and understanding (gestalt).

Results Among 341 patients consecutively enrolled in three centres, 149 (44%) presented AHFS. Eight predictive factors of AHFS were detected with a performance test showing an area under the model ROC curve of 0.86. Gestalt greater than or equal to five showed sensitivity of 78% and specificity of 90% (AUC 0.91) and diagnosed 88% of AHF in our population.

Conclusions We identified several independant predictors of final AHFS diagnosis. They should contribute to the development of diagnostic strategies in ED. However, unstructured gestalts seem to perform very well alone.

aDepartment of Cardiology, CARDIOMET Institute, Rangueil Hospital, University Hospital of Toulouse, Toulouse

bDepartment of Cardiology, University Hospital of Bordeaux, Bordeaux

cDepartment of Cardiology, General Hospital of Pau, Pau

dEmergency Department, St-André Hospital of Bordeaux

eEmergency Department, University Hospital of Bordeaux, Bordeaux

fEmergency Department, University Hospital of Toulouse, Toulouse, France

Received 15 February 2019 Accepted 12 June 2019

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website (

Correspondence to J. Roncalli, MD, PhD, Department of Cardiology, Institute CARDIOMET, University Hospital of Toulouse, Rangueil Hospital, 1, Avenue Jean Poulhès TSA50032, 31059 Toulouse Cedex9, France, Tel: +33 5 61 32 33 34; fax: +33 5 61 32 22 46; e-mail:

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.