Presenting phenotypes of acute heart failure patients in the ED: Identification and implications

Richard M. Nowak*, Brian P. Reed, Salvatore DiSomma, Prabath Nanayakkara, Michele Moyer, Scott Millis, Phillip Levy

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background There is little known about the baseline hemodynamic (HD) profiles (beyond pulse/blood pressure) of patients presenting to the Emergency department (ED) with acute heart failure (AHF). Assessing these baseline parameters could help differentiate underlying HD phenotypes which could be used to develop specific phenotypic specific approaches to patient care. Methods Patients with suspected AHF were enrolled in the PREMIUM (Prognostic Hemodynamic Profiling in the Acutely Ill Emergency Department Patient) multinational registry and continuous HD monitoring was initiated on ED presentation using noninvasive finger cuff technology (Nexfin, BMEYE, Edwards Lifesciences, Irvine, California). Individuals with clinically suspected and later confirmed AHF were included in this analysis and initial 15 minute averages for available HD parameters were calculated. K-means clustering was performed to identify out of 23 HD variables a set that provided the greatest level of inter-cluster discrimination and intra-cluster cohesions. Results A total of 127 patients had confirmed AHF. The final model, using mean normalized patient baseline HD values was able to differentiate these individuals into 3 distinct phenotypes. Cluster 1: normal cardiac index (CCI) and systemic vascular resistance index (SVRI); cluster 2: very low CI and markedly increased SVRI: and cluster 3: low CI and an elevated SVRI. These clusters were not differentiated using clinically available ED information. Conclusions Three distinct clusters were defined using novel noninvasive presenting HD monitoring technology in this cohort of ED AHF patients. Further studies are needed to determine whether phenotypic specific therapies based on these clusters can improve outcomes.

Original languageEnglish
Pages (from-to)536-542
Number of pages7
JournalAmerican Journal of Emergency Medicine
Volume35
Issue number4
DOIs
Publication statusPublished - 1 Apr 2017

Cite this