Abstract: SA-PO157
A Decision Tree to Predict Renal Replacement Therapy Requirement in Rhabdomyolysis-Induced AKI
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
- Gómez Ruiz, Ismael Antonio, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico
- Correa-Rotter, Ricardo, Institutor Nacional de la Nutricion, Mexico City, Mexico
- Morales-Buenrostro, Luis E., Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, Ciudad De Mexico, Mexico
- Mejia-Vilet, Juan M., Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, Ciudad De Mexico, Mexico
Background
Rhabdomyolysis Induced Acute Kidney Injury (RI-AKI) develops in 10-40% of patients who present rhabdomyolysis. When damage is severe enough to develop complications such as life-threatening hyperkalemia or anuria, up to 85% of patients will require renal replacement therapy (RRT). The aim of the present study was to study admission variables that predict RRT requirement.
Methods
Retrospective cohort study. All patients hospitalized for RI-AKI between 2007-2017 were included. Patients were divided according to RRT requirement and their admission parameters compared with Mann-Whitney's U test. Using ROC curves, we determined the best cut-off for lactate dehydrogenase (LDH), creatine kinase (CK) and the MacMahon score to predict RRT requirement and a decision tree was generated.
Results
We identified 42 RI-AKI hospitalizations. All patients had CK>5000 U/L. The main etiologies were drug-induced (41%) and excessive physical activity (21%). Nine patients (21%) developed stage 1 AKI, 5 (12%) stage 2 AKI and 28 (67%) stage 3 AKI. Twenty-two patients (52%) required RST. The most frequent indications for RRT initiation were anuria (64%) and hyperkalemia (32%). Intermittent hemodialysis was used in 52% of cases. The median time on RRT was 17.5 days (range 4-59). Five patients (12%) died during hospitalization due to infectious causes. On follow-up, 6 patients (14%) developed CKD. Patients with RRT requirement presented with higher serum phosphorus (6.2mg/dl [5.5-7.6] Vs. 3.3mg/dl [3.0-3.9], p<0.001), potassium (5.5mEq/l [4.8-6.3] Vs. 4.3mEq/l [3.6-4.8], p<0.001) and LDH (2124 U/l [1067-3193] Vs. 553 U/l [322-744], p<0.001). The AUC of LDH, MacMahon score and CK to predict RRT requirement were 0.873, 0.900 and 0.620 respectively. A decision tree was generated and shown in Figure 1.
Conclusion
A simple decision tree based on LDH levels and the MacMahon score at presentation can predict RRT requirement in RI-AKI