Abstract
AIMS: Sodium excretion is a well-defined marker used to assess diuretic response in acute heart failure (AHF). Despite a strong pathophysiological background, the role of urine chloride excretion has not been described and established yet. We aimed to evaluate chloride trajectory during intensive diuretic treatment in AHF patients and examine its potential role in predicting poor diuretic response. METHODS: The study was conducted on 50 AHF patients. Participants were included within the first 36 h of hospitalization. They received furosemide dose adjusted for body weight (half in bolus, half in 2 h infusion). Post-diuretic hourly urine collection with biochemical analysis was performed. RESULTS: In general, the concentrations of urine chloride (uCl(-)) and sodium (uNa(+)) at the baseline samples exhibited a comparable level (71 ± 39 vs. 70 ± 44 mmol/L, respectively; P = 0.99), but across all post-furosemide study timepoints, uCl(-) remained significantly higher than uNa(+) since 1 to 6 h of the study. In this course, both ions (uCl(-) and uNa(+)) reached peak values in 2 h (114 ± 28 vs. 97 ± 34 mmol/L, respectively; P < 0.01). The pattern of uCl(-) dominance over uNa(+) concentration was also observed in separate analyses of patients naïve to furosemide and those chronically exposed to furosemide. Regardless of these patterns, naïve to furosemide individuals excreted more ions (both uCl(-) and uNa(+)) than chronically exposed patients at all timepoints. Additionally, a strong, linear correlation between uCl(-) and uNa(+) was observed in each post-furosemide timepoint (the strongest in 1 h r = 0.87; P < 0.001). Both interdependent ions concentration was almost parallel when analysed in chronic furosemide users and those naïve to furosemide separately [uCl(-) = 0.85 * uNa(+) + 28.82, P < 0.001, R(2) = 0.83 for chronic furosemide users, and uCl(-) = 0.72 * uNa(+) + 41.55, P < 0.001, R(2) = 0.65 for naïves to furosemide (linear regression model)]. Moreover, uCl(-) (with cutoff point: 72 mmol/L) was a satisfactory predictive factor for poor diuretic response (<100 mL/h in 6 h since the beginning of furosemide infusion) [odds ratio (OR) 95% confidence interval (CI): 39.0 (3.8-405.00)]. It presented those properties also after adjusting for urine creatinine [cutoff point: 0.296 mmol/mg-OR (95% CI): 81.0 (8.0-816.0)]. CONCLUSIONS: Urine chloride and sodium are highly interrelated during decongestion of AHF patients. The uCl(-) (cutoff 72 mmol/L) exhibits better prognostic abilities to identify poor diuretic response than uNa(+).