Short, and long-term mortality among cardiac intensive care unit patients started on continuous renal replacement therapy

Purpose: Patients requiring continuous renal replacement therapy (CRRT) are at high risk of death. Predictors of hospital mortality and post-discharge survival in cardiac intensive care unit (CICU) patients requiring CRRT have not been reported. Materials and methods: Retrospective review of 198 CICU patients undergoing CRRT from 2006 to 2015. Multivariable regression identified predictors of hospital mortality and Cox proportional-hazards identified predictors of post-discharge mortality among hospital survivors. Results: The indication for CRRT was volume overload in 129 (65%) and metabolic abnormalities in 76 (38%). 105 (53%) subjects died in hospital, with 22% dialysis-free hospital survival. Cardiogenic shock was present in 159 (80%) subjects; 150 (76%) subjects received vasopressors and 101 (51%) subjects required mechanical ventilation. Hospital mortality was similar in cardiogenic and non-cardiogenic causes of CICU admission. Predictors of hospital death included semi-quantitative RV function, Braden score, VIS, and PaO2/FIO2 ratio. Median post-discharge Kaplan-Meier survival was 1.9 years. Predictors of post-hospital death included age, VIS, diabetes, Braden score, semi-quantitative RV function, prior heart failure, and dialysis dependence. The indication for CRRT was not predictive of survival. Conclusion: Mortality is high among CICU patients requiring CRRT, and is predicted by the Braden score, RV dysfunction, respiratory failure and vasopressor load. (C) 2019 Elsevier Inc. All rights reserved.