がん研究と免疫腫瘍学ジャーナル
オープンアクセス

概要

In-Hospital Mortality in Acute Heart Failure: Predicted ADHERE Risk vs. Observed Mortality

Irene Marques*, Manuela V. Bertão, Denisa Mendonça, Laetitia Teixeira

Background: Acute heart failure in-hospital mortality varies between 3.8-28%. The Acute Decompensated Heart Failure National Registry (ADHERE) risk tree is a landmark risk prediction model for in-hospital mortality. This study compares in-hospital mortality predicted by the ADHERE risk tree with the observed mortality of patients from the Pre Heart Failure Clinic (PRECIC-Pré Clínica de Insuficiência Cardíaca) study.

Methods and findings: This is a retrospective, single-centre, observational sub-analysis of the previously reported PRECIC study. The ADHERE risk tree was applied. In-hospital predicted and observed mortality was compared in 419 patients with a mean age of 79.5 years. The majority were female and had left ventricle ejection fraction ≥ 40%. In-hospital mortality was 8.1% (n=34). After being classified in the ADHERE risk groups, the observed mortality rates between 3 groups were significantly different (p<0.001): Low, Intermediate 2, and Intermediate 1 (4.5% vs. 2.3%, p=0.0017; 12.5% vs. 5.6%, p=0.024; and 40% vs. 13.2%, p=0.003, respectively). The observed mortality was twice the predicted mortality in the Low and Intermediate two groups and threefold in the Intermediate 1 group. The small number of patients in the High group impairs solid conclusions to be drawn for this group.

Conclusion: The ADHERE risk tree cannot accurately predict the in-hospital mortality rates of patients with the characteristics of those included in the PRECIC study. This finding supports the need to locally validate risk trees to take into account different patient characteristics and healthcare settings.