Abstract MP03: Epidemiology of Hospitalized Heart Failure Patients with Preserved Ejection Fraction: The ARIC Heart Failure Surveillance Project

Abstract

Background Despite the high burden of heart failure with preserved ejection fraction (HFpEF), information about its epidemiology is limited. We describe the population distribution of HFpEF according to criteria commonly used to ascertain heart failure (HF) and compare the characteristics and outcomes of patients with HFpEF to those with HF with reduced ejection fraction (HFrEF). Methods From the Atherosclerosis Risk in Communities (ARIC) HF surveillance of 21 hospitals in four US communities during 2005-2008, potential HF events were sampled based on ICD discharge codes from patients ages 55 years and older, . Trained study personnel abstracted charts (unweighted n =4784), and the presence of HF was determined both by computerized classification and by a panel of standardized physicians. Left ventricular ejection fraction (EF) was missing in 19%; HFpEF was defined as HF with EF ≥ 50%. All statistical analyses incorporated sampling weights. A multivariable logistic model of the probability of HFpEF was developed using variables associated with HFpEF in univariate analyses (p<0.2). Results The proportion of the hospitalized HF patients with HFpEF varied only slightly according to HF classification schema: NHANES (49%), Modified Boston (48%), Framingham (47%), ‘2008 Consensus criteria’ (44%), and ARIC review panel (43%). The following characteristics were independently associated with HFpEF (Odds Ratio and 95% Confidence Interval): female (2.28; 1.87, 2.82), history of sleep apnea (1.90; 1.30, 2.77), systolic blood pressure (1.08; 1.05, 1.12) per 10 mm of Hg increase), atrial fibrillation (1.29 (1.03, 1.61), heart rate (0.91; 0.87, 0.95 per 10 beats per minute increase), paroxysmal nocturnal dyspnea (0.70; 0.54, 0.92), and study community. Length of hospital stay was comparable for HFpEF and HFrEF patients (median = 5 days, p for geometric mean = 0.76). Absolute mortality at one-year was 4.4% lower among HFpEF compared to HFrEF (12.3% vs. 16.7%, p<0.05) with age, gender, and race adjusted case fatality OR= 0.75 (95% CI 0.56, 0.99). Conclusions The proportion of patients with HFpEF among HF patients hospitalized in 2005-2008 varied from 43 to 49% according to the main published HF classification schema. HFpEF patients had a similar length of hospital stay but lower one-year case fatality than those with HFrEF. These results support increasing efforts to better characterize HFpEF and its course to reduce the total burden of HF.