Usefulness of heart failure categories based on left ventricular ejection fraction

Abstract

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): This study was sponsored by AstraZeneca. Introduction Heart failure, a common disease with a mortality on par with many cancers, is one of the costliest diseases to society in the western world. Current heart failure guidelines are based on categories of left ventricular ejection fraction (LVEF) and have recently introduced a narrow category with mid-range/mildly reduced ejection fraction (HFmrEF; LVEF 41-49%) between the previous categories with reduced (HFrEF; LVEF ≤40%) and preserved (HFpEF; LVEF ≥50%) ejection fraction. The grouping of continuous variables into narrow categories can be questioned if the variability is high. Purpose To determine the shape of the LVEF distribution and quantify the stability of LVEF categories over time in a large real-world heart failure population, in order to establish the usefulness of LVEF categories for treatment decisions. Methods Population-based observational cohort study of all new cases of chronic heart failure in Region Stockholm, Sweden, from 2016-01-01 until 2020-12-31. All available values of LVEF were collected over time and patients were followed until death or end of study. Mixed models were used to quantify within-person variance in LVEF, and Multistate Markov models, with death as an absorbing state, to quantify the stability of LVEF categories. Results During the study period, 8,455 patients were diagnosed with heart failure. LVEF values followed a perfect normal distribution. The standard deviation of the within-person variance in LVEF was 7.4%. Regardless of initial LVEF category, age or sex, the mean time spent in any LVEF category before transition to another category (mean sojourn time) was on average less than 0.6 years (Figure 1). The one-year probabilities of transitioning between states were substantial (Figure 2). Patients with HFmrEF had a probability of 1/4 of remaining in that category one year later; the probability of patients with HFrEF of remaining in that category one year later was 1/4 for women and 1/3 for men; and patients with HFpEF had a probability of remaining in that category of 2/3 for women and 1/2 for men. Conclusion In patients with heart failure, LVEF has considerable variability over time. Hence, LVEF has limited use for categorisation of the heart failure population into narrow categories.