SP212DIURETIC RESISTANCE: A STRONG MORTALITY PREDICTOR IN ACUTE DECOMPENSATED HEART FAILURE ADMISSIONS

Abstract

Introduction and Aims: Despite diuretic treatment is not associated with a better prognosis in acute heart failure (AHF), loop diuretics are the mainstay of initial pharmacological treatment. Failure to loop diuretic response seems to be related with a worse outcome. We sought to describe and compare patient (P) features and outcome in AHF admissions who developed diuretic resistance (RDIUR) and to identify its predictors. Methods: A total of 728 consecutive P were hospitalized for AHF between July 2011 and November 2015 and included in the analysis. RDIUR was defined as failure to achieve a urine output ≥1.5 ml/kg/hr due to a pre-established protocol that included furosemide in an initial 40 mg-IV bolus, followed by a two hours 5 mg/hr-continuous infusion. Under failure to achieve the goal, furosemide dose was doubled for two additional hours. Failure to respond to this strategy was called “diuretic resistance”. Results: RDIUR was observed in 71 P (9.8%). No differences were detected after analyzing sex, ventricular function or structural heart disease. These P were younger (69±15 vs 72±14 years old; p= 0.03), a greater history of CKD (53 vs 29%; p< 0.001), greater Furosemide doses prior to admission (74 vs 49 mg/d; p< 0.001). Severe hyponatremia on admission (<125 mEq/L) and new severe hyponatremia were more frequent in RDIUR (12.7 vs 4.3%; p< 0.005 and 26 vs 9.6%; p= 0.02 respectively). Clinical hypoperfusion and right sided AHF evidences were more frequent in RDIUR (27 vs 7.7 %; p< 0.001 and 38.6 vs 24%; p= 0.01), while pulmonary congestion was less frequent (34 vs 68%; p< 0.001). RDIUR was associated with in-hospital events, as worsening heart failure (WHF) (49 vs 13%; p< 0.001), inotropic support (60 vs 13%; p< 0.001) and longer hospitalizations (> 7 days)(74 vs 35%; p< 0,001). In-hospital mortality was 10.4%, higher in RDIUR (35.2 vs 7.9%; OR 6.3; CI95% 3.6-11; p< 0.001). Six months all-cause mortality was 36.2%, higher in RDIUR too (63.4 vs 33.3%; OR 3.5; CI95% 2-5.8; p< 0.001). No differences were detected after analysing readmission rates. RDIUR (HR 3.1; IC95% 1.5-6; p= 0.003), WHF (HR 3; IC95% 1.5-6; p= 0.002) and low T3 concentrations on admission (HR 4.2; IC95% 1.4-12; p= 0.01) were found as independent predictors for in-hospital mortality. Multivariate analysis by multiple logistic regression revealed need for inotropic support (HR 3.9; IC95% 1.9-8; p< 0,001), right sided heart failure evidences (HR 3.4; IC95% 1.7-6.6; p< 0.001), previous CKD (HR 3.2; IC95% 1.6-6; p< 0,001) and WHF (HR 3; IC95% 1.4-6; p= 0.003) as independent predictors for RDIUR development. Conclusions: RDIUR is a serious problem in P hospitalized for AHF. It is associated with a higher in-hospital event rate and longer admissions. Early detection might allow early therapy to reduce adverse outcome.