Adam Singer, MD – Professor and Vice Chairman for Research, Stony Brook University
Henry Thode, Jr., PhD – Assistant Professor, Stony Brook University
Kyle S. Fortner, BS – Senior Manager, Clinical Trials, Abbott
Brian Bales, MD – Professor, Vanderbilt University
Chad Cannon, MD – Professor, University of Kansas Medical Center
Anna Marie Chang, MD – Professor, Thomas Jefferson University
George Glass, MD – Assistant Professor, University of Virginia
Eric Gross, MD – Professor, University of California, Davis
Sergey Motov, MD – Professor, Maimonides Medical Center
Daniel Spoon, MD – Professor, International Heart Institute of Montana
Frank Peacock, MD – Professor, Baylor College of Medicine
Background and Objectives: Diagnosis and risk stratification of patients with suspected acute heart failure (AHF) is problematic; most patients directly admitted to the hospital. We determined whether inability to ambulate and lay flat predicted AHF and admission.
Methods: We conducted a prospective, multicenter study of adult ED patients with suspected AHF between 2014-2015. Data were collected and the ability to lay flat and ambulate while in the ED was determined. Outcomes were an adjudicated diagnosis of AHF and disposition. We determined the association of inability to lay flat and ambulate in the ED with adjudicated AHF, admission, and ICU admission using univariate and multivariate analyses using stepwise logistic regression.
Results: There were 330 patients. Mean age was 63 (14), 63% were male, 72% were white. New or worsening adjudicated AHF was diagnosed in 129 (39%) patients. Of all patients, 53% were able to lay flat and 83% could briefly ambulate in the ED. 240 patients (73%) were admitted (6% to an ICU). Median (IQR) ED LOS was 7.5 (5.4-11.5) hours; median (IQR) hospital LOS was 78 (43.6-141.7) hours. Other comorbidities requiring admission were present in 184 patients (56%). Inpatient mortality was 4%. Compared with patients without AHF, patients with AHF were older (65 vs. 61, P=0.03), more likely male (71% vs. 58%, P=0.03), and less likely to lay flat (41% vs. 61%, P=0.001). Multivariate (odds ratio; 95%CI) predictors of AHF included edema (2.24; 1.13-4.46), orthopnea (2.45; 1.23-4.88), weight gain (7.11; 2.47-20.46), sputum production (0.26; 0.11-0.64), CXR abnormalities (3.40, 1.50-7.73),and elevated BNP/NT-pro BNP (12.50; 6.26-24.99). Admitted patients were older than non-admitted (64 vs. 59, P=0.006), and less likely to lay flat (50% vs. 64%, P=0.03). Multivariate predictors of admission included heart rate (1.02/bpm; 1.00-1.04), edema (2.35; 1.11-4.97), confusion (0.06; 0.004-0.91), elevated BNP/NT-pro-BNP (4.72; 2.03-10.95), and co-morbidities (24.81, 9.80-62.84). Multivariate predictors of ICU admission included respiratory rate (1.09/RPM; 1.03-1.17); renal disease (4.03; 1.10-14.80), and inability to ambulate in the ED (2.67; 1.03-6.90).
Conclusion: In ED patients with suspected AHF, inability to lay flat while in the ED is predictive of ICU admission, but not associated with AHF diagnosis or admission after controlling for confounding variables.