Abstract

BACKGROUND: Effective decongestion of heart failure patients predicts improved outcomes, but high dose loop diuretics (HDLD) used to achieve diuresis predict adverse outcomes. In the DOSE trial, randomization to a HDLD intensification strategy (HDLD-strategy) improved diuresis but not outcomes. Our objective was to determine if potential beneficial effects of more aggressive decongestion may have been offset by adverse effects of the HDLD used to achieve diuresis. METHODS AND RESULTS: A post hoc analysis of the DOSE trial (n=308) was conducted to determine the influence of post-randomization diuretic dose and fluid output on the rate of death, rehospitalization or emergency department visitation associated with the HDLD-strategy. Net fluid output was used as a surrogate for beneficial decongestive effects and cumulative loop diuretic dose for the dose-related adverse effects of the HDLD-strategy. Randomization to the HDLD-strategy resulted in increased fluid output, even after adjusting for cumulative diuretic dose (p=0.006). Unadjusted, the HDLD-strategy did not improve outcomes (p=0.28). However, following adjustment for cumulative diuretic dose, significant benefit emerged (HR=0.64, 95% CI 0.43-0.95, p=0.028). Adjusting for net fluid balance eliminated the benefit (HR=0.95, 95% CI 0.67-1.4, p=0.79). CONCLUSIONS: A clinically meaningful benefit from a randomized aggressive decongestion strategy became apparent after accounting for the quantity of loop diuretic administered. Adjusting for the diuresis resulting from this strategy eliminated the benefit. These hypothesis-generating observations may suggest a role for aggressive decongestion in improved outcomes.