The FAST-FURO study: effect of very early administration of intravenous furosemide in the prehospital setting to patients with acute heart failure attending the emergency department

Abstract

Aims: The effect of early administration of intravenous (IV) furosemide in the emergency department (ED) on short-term outcomes of acute heart failure (AHF) patients remains controversial, with one recent Japanese study reporting a decrease of in-hospital mortality and one Korean study reporting a lack of clinical benefit. Both studies excluded patients receiving prehospital IV furosemide and only included patients requiring hospitalization. To assess the impact on short-term outcomes of early IV furosemide administration by emergency medical services (EMS) before patient arrival to the ED.

Methods and results: In a secondary analysis of the Epidemiology of Acute Heart Failure in Emergency Departments (EAHFE) registry of consecutive AHF patients admitted to Spanish EDs, patients treated with IV furosemide at the ED were classified according to whether they received IV furosemide from the EMS (FAST-FURO group) or not (CONTROL group). In-hospital all-cause mortality, 30-day all-cause mortality, and prolonged hospitalization (>10 days) were assessed. We included 12 595 patients (FAST-FURO = 683; CONTROL = 11 912): 968 died during index hospitalization [7.7%; FAST-FURO = 10.3% vs. CONTROL = 7.5%; odds ratio (OR) = 1.403, 95% confidence interval (95% CI) = 1.085-1.813; P = 0.009], 1269 died during the first 30 days (10.2%; FAST-FURO = 13.4% vs. CONTROL = 9.9%; OR = 1.403, 95% CI = 1.146-1.764; P = 0.004), and 2844 had prolonged hospitalization (22.8%; FAST-FURO = 25.8% vs. CONTROL = 22.6%; OR = 1.189, 95% CI = 0.995-1.419; P = 0.056). FAST-FURO group patients had more diabetes mellitus, ischaemic cardiomyopathy, peripheral artery disease, left ventricular systolic dysfunction, and severe decompensations, and had a better New York Heart Association class and had less atrial fibrillation. After adjusting for these significant differences, early IV furosemide resulted in no impact on short-term outcomes: OR = 1.080 (95% CI = 0.817-1.427) for in-hospital mortality, OR = 1.086 (95% CI = 0.845-1.396) for 30-day mortality, and OR = 1.095 (95% CI = 0.915-1.312) for prolonged hospitalization. Several sensitivity analyses, including analysis of 599 pairs of patients matched by propensity score, showed consistent findings.

Conclusion: Early IV furosemide during the prehospital phase was administered to the sickest patients, was not associated with changes in short-term mortality or length of hospitalization after adjustment for several confounders.

Keywords: Acute heart failure; Diuretics; Emergency department; Furosemide; Mortality; Outcome.

Figure 1

Flow chart for patient inclusion in the FAST-FURO study. ED, emergency department; EMS, emergency medical service; LVEF: left ventricular ejection fraction.

Figure 2

Percentage of in-hospital mortality, 30-day mortality, and prolonged hospitalization in the FAST-FURO and CONTROL groups. The analysis is presented for the whole cohort (left), by each risk category of the MEESSI scale (middle), and in 599 pairs of patients matched by the propensity score to be treated with intravenous furosemide before arrival to the emergency department (right). ^aOnly 8213 out of 12 594 (65.2%) were classified by the MEESSI risk score and in 599 pairs of patients matched by propensity score. ^bOnly 8148 out of 12 498 (65.2%) were classified by the MEESSI risk score and in 599 pairs of patients matched by propensity score. ^cOnly 8127 out of 12 447 (65.3%) were classified by the MEESSI risk score and in 599 pairs of patients matched by propensity score. ED, emergency department; EMS, emergency medical service; PS, propensity score.

Figure 3

Unadjusted and adjusted odds ratio for the assessed endpoints. Variables corresponding to baseline characteristics used for adjustment were those resulting in significant differences between groups in the univariable analysis: diabetes mellitus, ischaemic heart disease, atrial fibrillation, peripheral artery disease, NYHA class, and left ventricular ejection fraction. The variable used for adjustment for the severity of the AHF decompensation was the MEESSI score, taken as a continuous variable. Bold numbers denote statistical significance (P < 0.05). AHF, acute heart failure; CI, confidence interval; LL, lower limit; OR, odds ratio; UL, upper limit.

Figure 4

Stratified analysis by age, sex, left ventricular ejection fraction, on treatment with loop diuretic at home, creatinine concentration at emergency department arrival and severity of the acute heart failure episode assessed by the MEESSI-AHF scale. Bold numbers denote statistical significance (P < 0.05). CI, confidence interval; LL, lower limit; LVEF, left ventricular ejection fraction; OR, odds ratio; UL, upper limit.