In-Hospital Observation on Oral Diuretics After Treatment for Acute Decompensated Heart Failure: Evaluating the Utility

Background: Following treatment for acute decompensated heart failure, in-hospital observation on oral diuretics (OOD) is recommended, assuming it provides actionable information on discharge diuretic dosing and thus reduces readmissions.

Methods: In the Mechanisms of Diuretic Resistance (MDR) cohort, we analyzed in-hospital measures of diuretic response, provider's decisions, and diuretic response ≈30 days postdischarge. In a Yale multicenter cohort, we assessed if in-hospital OOD was associated with 30-day readmission risk. The main objective of this study was to evaluate the utility of in-hospital OOD.

Results: Of the 468 patients in the MDR cohort, 57% (N=265) underwent in-hospital OOD. During the OOD, weight change and net fluid balance correlated poorly with each other (r=0.36). Discharge diuretic dosing was similar between patients who had increased, stable, or decreased weight (decreased discharge dose from OOD dose in 77% versus 72% versus 70%, respectively), net fluid status (decreased discharge dose from OOD dose in 100% versus 69% versus 74%, respectively), and urine output (decreased discharge dose from OOD dose in 69% versus 79% versus 72%, respectively) during the 24-hour OOD period (P>0.27 for all). In participants returning at 30 days for formal quantification of outpatient diuretic response (n=98), outpatient and inpatient OOD natriuresis was poorly correlated (r=0.26). In the Yale multicenter cohort (n=18 454 hospitalizations), OOD occurred in 55% and was not associated with 30-day hospital readmission (hazard ratio, 0.98 [95% CI, 0.93-1.05]; P=0.51).

Conclusions: In-hospital OOD did not provide actionable information on diuretic response, was not associated with outpatient dose selection, did not predict subsequent outpatient diuretic response, and was not associated with lower readmission rate. Additional research is needed to replicate these findings and understand if these resources could be better allocated elsewhere.

Registration: URL: https://www.

Clinicaltrials: gov; Unique identifier: NCT02546583.