Association Between Loop Diuretics and Mortality in Patients With Cardiac Surgery–Associated Acute Kidney Injury: A Retrospective Propensity Score–Weighted Analysis

Abstract

BACKGROUND: Although loop diuretics (LDs) have been widely used in clinical practice, their effect on mortality when administered to patients experiencing cardiac surgery–associated acute kidney injury (CS-AKI) remains unknown. The study aimed to investigate the effectiveness of LD use in patients with CS-AKI. METHODS: Patients who underwent cardiac surgery with AKI were identified from the Medical Information Mart for Intensive Care III. Postoperative LD use in intensive care units (ICUs) was exposure. There were 2 primary outcome measures, the in-hospital mortality and ICU mortality; both were treated as time-to-event data and were analyzed via multivariable Cox proportional hazard models. Inverse probability weighting (IPW) was used to minimize bias. RESULTS: The study enrolled a total of 5478 patients, with a median age of 67 years, among which 2205 (40.3%) were women. The crude in-hospital and ICU mortality rates were significantly lower in the LD use group (525 of 4150 [12.7%] vs 434 of 1328 [32.7%], P < .001; 402 of 4150 [9.69%] vs 333 of 1328 [25.1%], P < .001). Adjusted hazard ratios suggested significant reductions in both in-hospital (hazard ratio [HR], 0.428; 95% confidence interval [CI], 0.374–0.489) and ICU mortality (HR, 0.278; 95% CI, 0.238–0.327). The IPW data showed a similar reduction, in-hospital mortality (HR, 0.434; 95% CI, 0.376–0.502) and ICU mortality (HR, 0.296; 95% CI, 0.251–0.349). Such association may act differently for patients with different fluid balance (P value for interaction < .001). CONCLUSIONS: LD use is associated with lower hospital and ICU mortality in CS-AKI patients in general. Patients under different conditions showed diverse responses toward such treatment indicating that personalized management is needed.