Abstract 187: Loop Diuretic Therapy in Acute Decompensated Heart Failure among Premier Hospitals

Abstract

Background: Although loop diuretics are frequently used for patients with heart failure (HF), little is known about the variation in patterns of diuretic therapy in US hospitals. We sought to describe such treatment patterns among a diverse group of hospitals. Methods: We studied HF hospitalizations occurring during 2009-10 in Premier Inc. hospitals participating in a collaborative project to pool administrative and charge data, which includes information about drug types, average daily dose, and duration of therapy. We excluded hospitals with less than 25 HF hospitalizations. For ease of comparison, all diuretic doses were converted to bioequivalent doses of intravenous (IV) furosemide: 40mg IV furosemide ∼ 80mg oral furosemide ∼ 20mg (oral or IV) torsemide ∼ 1mg (oral or IV) bumetanide. Summary statistics were calculated. Results: Among 366 studied hospitals (264,675 HF hospitalizations), use of any loop diuretic had an interquartile range (IQR) from 92% to 96% (median: 94%). At the hospital level, the average daily dose IQR varied from 45mg to 64 mg (median: 55 mg) and the median duration of therapy was 4 days (IQR: 4 to 4; median: 4), as was the median length of stay. The IQR for use of furosemide varied from 89% to 94% (median: 92%), and its median average daily dose had an IQR from 40mg to 60 mg (median: 53 mg). Hospital use of bumetanide had an IQR from 2% to 11%, and hospital use of torsemide had an IQR from 0% to 4% (medians of 5% and 1%, respectively). The variation in median average daily dose for bumetanide and torsemide was greater than for furosemide (bumetanide IQR: 79mg to 127 mg, with median of 89 mg; torsemide IQR: 53mg to 120 mg, with median of 80 mg). Use of IV diuretics on the last day before home discharge had an IQR from 16% to 33% (median: 24%) across hospitals. Conclusion: US hospitals administer loop diuretics, particularly furosemide, to the vast majority of HF inpatients. The duration and daily dosage of therapy was similar across most hospitals. In contrast, a minority of hospitals used bumetanide and torsemide for several patients. The daily dosage of these agents showed more marked variation. We observed a high rate of intravenous diuretic use on the last day of hospitalization, with considerable variation across hospitals.