The Effect of Loop Diuretics on 28-Day Mortality in Patients With Acute Respiratory Distress Syndrome

Abstract

Background: Diuretics have been widely used in critically ill patients while it remains uncertain whether they can reduce mortality in patients with acute respiratory distress syndrome (ARDS). This study aimed to investigate the associations between diuretics and 28-day mortality in patients with ARDS. Methods: This is a secondary analysis of the ARDS Network Fluid and Catheter Treatment Trial (FACTT) of National Heart, Lung, and Blood Institute. Those patients who did not receive renal replacement therapy within the first 48 h after enrollment in the FACTT were included in the analysis. A marginal structural Cox model (MSCM) was used to investigate the associations between diuretics and 28-day mortality after correction of both the baseline and time-varying variables. The latent class analysis (LCA) and subgroup analysis were performed to identify the kind of patients that could be benefited from diuretics. Results: A total of 932 patients were enrolled, i.e., 558 patients in the diuretics group and 374 patients in the no diuretics group within the first 48 h. The 28-day mortality was lower in the diuretics group (15.1 vs. 28.1%, p < 0.001). In MSCM, diuretics use was related to the improved 28-day mortality (HR 0.78; 95% CI 0.62–0.99; p = 0.04). LCA identified three subtypes, and diuretics were associated with reduced mortality in subtype 3, which was characterized by worse renal function and higher central venous pressure (CVP). A subgroup analysis indicated survival advantage among the female patients, sepsis induced ARDS, and those with the ratio of partial pressure of oxygen to the fractional concentration of inspired oxygen (PaO2/FiO2) ≤ 150 mmHg, and mean arterial pressure (MAP) ≥ 65 mmHg. Conclusion: Loop diuretics were associated with the reduced 28-day mortality in the patients with ARDS, after controlling for time-varying confounders. Randomized trials are required to verify the association.