The association of the heart failure score with mortality and heart failure hospitalizations in elderly patients: insights from the Mode Selection Trial (MOST).

BACKGROUND: Patient and physician preferences as well as cost favor an increasingly higher threshold for hospital admission for heart failure (HF) treatment. This trend risks masking the severity and prevalence of HF as hospitalization for HF may decrease. METHODS: Heart Failure Score (HFS) has 4 ordinal subscales assessing (1) HF symptoms, physical signs of left (2) and (3) right HF, and (4) therapy changes for HF. Heart Failure Score was calculated for 1257 of 2010 (63%) patients enrolled in the MOST trial in sinus node dysfunction, who survived and had complete first-year HFS data at 4 postpacemaker implant visits (1, 3, 6, and 12 months). Heart Failure Score was summed and ranged from 0 to 14, with lower scores representing less HF. RESULTS: There were 1257 patients (median age 74 years [interquartile range 68-79], 47% were women, 61% had hypertension, 20%, diabetes mellitus, and 23%, prior myocardial infarction). The median HFS accumulated during 1 year was 4 (interquartile range 1-8). Of patients with a benign first year, those with a higher HFS were more likely to die during subsequent follow-up compared with patients with lower HFS (hazard ratio 1.07, 95% CI 1.04-1.10 for each 1-point increase, P < .001). CONCLUSIONS: Increasing HFS is associated with an increased risk of mortality in mostly elderly patients without pre-existing HF. Heart Failure Score may be a useful surrogate HF end point for clinical trials.

Duke Scholars

Author Manesh Raman Patel Medicine, Cardiology

Author Kerry L. Lee Biostatistics & Bioinformatics, Division of Biostatistics