Abstract: SA-PO144
Impact of Clinical Variables at Dialysis Initiation for AKI in the ICU on In-Hospital Mortality
Session Information
- AKI: Epidemiology, Risk Factors, Prevention - III
November 09, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Acute Kidney Injury
- 102 AKI: Clinical, Outcomes, and Trials
Authors
- Ganguli, Anirban, Medstar Washington Hospital Center, Washington, District of Columbia, United States
- Adhikari, Shreedhar, Medstar Washington Hospital Center, Washington, District of Columbia, United States
- Shah, Vatsal, Medstar Washington Hospital Center, Washington, District of Columbia, United States
- Sherman, Michael J., Medstar Washington Hospital Center, Washington, District of Columbia, United States
- Veis, Judith H., Medstar Washington Hospital Center, Washington, District of Columbia, United States
- Moore, Jack, Medstar Washington Hospital Center, Washington, District of Columbia, United States
Background
Current research on timing of dialysis in critically ill AKI has focused on analyzing survival outcome with arbitrary definitions of “early” or “late” start. However, the competing effects of other variables such as clinical comorbidities, dialysis indication, or acuity of illness at dialysis start are unknown.
Methods
We analyzed new adult AKI patients initiated on renal replacement therapy (RRT) while in our 5 intensive care units (ICU) from 1/1/2010 through 12/31/2015, to identify clinical variables associated with survival to hospital discharge.
Results
Of the 235 patients initiated on RRT in medical and surgical ICUs, the mean age was 61.8+/- 14.3 yrs; 60 % were male, 47 % were Afro-American with a Charlson Comorbidity Score (CCS) of 5.5 +/- 3.1 and acuity scores of 29.6 +/-7.6 (APACHE-II) and 12.0 +/- 4.4 (SOFA) at dialysis start. The most common modality of RRT was continuous (67.2%). Logistic regression identified independent association of survival with low serum lactate, low SOFA scores, elevated serum creatinine at RRT initiation and hyperkalemia but not with CCS and time from KDIGO Stage 3 AKI to dialysis initiation(as a surrogate for timing). Serum lactate{odd ratio (O.R.) 0.74- 0.91} at initiation also correlated inversely with survival beyond 48 hours. Stratifying patients by SOFA scores at RRT initiation (<10=low-risk, >10= high-risk) identified severity of volume overload or hyperkalemia (low-risk group) and RRT modality type or serum lactate (high-risk group), as being associated with survival. Receiver-Operator Characteristics (ROC) of biochemical variables at dialysis initiation showed that only serum lactate had a moderate c-statistic of 0.759 in discriminating survivors from non-survivors.
Conclusion
Data from critically ill AKI patients initiated on RRT in the ICU primarily showed acuity of illness at the start of RRT affecting survival. Since time from KDIGO Stage 3 AKI to dialysis initiation was not associated with survival, the validity of definitions such as “early” or “late” RRT initiation remains uncertain. Triaging clinical decision based on acuity scores may optimize clinical outcomes. Finally, the absence of any association between hospital survival and co-morbid scores has great implications for prognostication and palliative care.