Kidney Week

Abstract: FR-PO448

Examining Mortality in the Multinational Observational Study of Continuous Renal Replacement Therapy (CRRT) Practices

Session Information

• Hemodialysis and Frequent Dialysis - IV
  November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Dialysis

• 701 Dialysis: Hemodialysis and Frequent Dialysis

Authors

• Neyra, Javier A., University of Kentucky Medical Center, Lexington, Kentucky, United States

Javier A. Neyra,

• Secic, Michelle, Secic Statistical Consulting, Inc., Cleveland, Ohio, United States

Michelle Secic,

• Rewa, Oleksa G., University of Alberta, Edmonton, Alberta, Canada

Oleksa G. Rewa,

• House, Andrew A., LHSC University Hospital, London, Ontario, Canada

Andrew A. House,

• Monga, Divya, University of Mississippi Medical Center, Madison, Mississippi, United States

Divya Monga,

• Juncos, Luis A., Central Arkansas Veterans Healthcare System, Little Rock, Arkansas, United States

Luis A. Juncos,

• Piazza, Robin, Watermark Research Partners, Inc, Indianapolis, Indiana, United States

Robin Piazza,

• Bagshaw, Sean M., University of Alberta, Edmonton, Alberta, Canada

Sean M. Bagshaw,

• Goldstein, Stuart, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States

Stuart Goldstein,

• Heung, Michael, University of Michigan, Ann Arbor, Michigan, United States

Michael Heung,

Background

Risk-stratification of patients receiving CRRT to guide practice is direly needed. Utilizing data from a registry of critically ill adults receiving CRRT, we report mortality and kidney recovery rates and examine clinical parameters associated with mortality.

Methods

Multi-national observational registry study including critically ill adults (age ≥18 and ≤89 years) receiving CRRT. Patients who received CRRT for <48 h or other type of acute RRT before CRRT were excluded. Moderate and severe kidney disease at baseline were defined as baseline serum creatinine >265 µmol/L and ESKD/kidney transplant, respectively.

Results

566 critically ill patients receiving CRRT were included. Mean (SD) age was 58.7 (14.3) yrs., 58.5% were male. Mean BMI was 31.1 (8.0) kg/m². Median [range] Charlson and APACHE-II scores were 2.0 [0.0-18.0] and 29.0 [9.0-50.0], respectively and 106 (18.7%) patients had moderate/severe kidney disease at baseline. Median days from ICU admission to CRRT initiation was 1.0 [0.0-52.0] and 213 (37.6%) patients had fluid overload (FO%) ≥10% at CRRT initiation. Median SOFA at CRRT initiation was 15.0 [6.0-23.0]. Mean delivered CRRT dose was 33.1 (14.1) ml/kg/h and the median FO% per CRRT-day was 1.4% [0.3%-7.0%]. ICU mortality rate was 48.4%. In a multivariable Cox model, age (>60 vs ≤60), BMI (>29.7 vs ≤29.7), moderate/severe kidney disease at baseline, FO% per CRRT-day (>1.4% vs ≤1.4%), the absence of circuit anticoagulation, diagnosis of sepsis or acute-on-chronic liver failure, and serum bicarbonate at CRRT initiation (≤20 vs >20 mmol/L) independently associated with higher ICU mortality. Among 217 survivors without moderate/severe kidney disease at baseline, 142 (65.4%) recovered kidney function no longer needing RRT at hospital discharge.

Conclusion

Approximately one of two critically ill adults receiving CRRT died in the ICU. However, almost two-thirds of survivors recovered kidney function sufficiently enough to be independent of RRT. We identified clinical parameters independently associated with ICU mortality. This registry can help develop and validate risk-stratification tools in this susceptible population, and identify modifiable risk factors for evaluation in clinical trials.

Funding

• Private Foundation Support