Abstract: SA-OR017
Predicting Severe AKI, Fluid Overload, and Renal Replacement Therapy with the Renal Angina Index in Critically Ill Children
Session Information
- AKI: Risk Factors, Biomarkers, and Predictors
November 09, 2019 | Location: Salon C, Walter E. Washington Convention Center
Abstract Time: 05:42 PM - 05:54 PM
Category: Acute Kidney Injury
- 101 AKI: Epidemiology, Risk Factors, and Prevention
Authors
- Krallman, Kelli A., Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Roy, Jean-Philippe, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Chima, Ranjit S., Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Schmerge, Alexandra, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Gerhardt, Bradley S., Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Fei, Lin, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
- Goldstein, Stuart, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States
Background
The Renal Angina Index (RAI) and urinary Neutrophil Gelatinase Associated Lipocalin (uNGAL) can be used to risk stratify patients for development of severe AKI (sAKI). Our research has focused on combining the RAI and NGAL for early detection and acute kidney injury (AKI) prediction.
Methods
Patients admitted to the Pediatric Intensive Care Unit (PICU) from 7/1/18 to 11/30/18, underwent an automated RAI assessment 12 hours after admission. RAI+ patients were defined by RAI ≥ 8 and had uNGAL assessed. A cutoff value of 150ng/mL was used to stratify patients with the primary outcome measure the development of sAKI (KDIGO Stage 2 or 3) through PICU day 4. Secondary outcomes included fluid overload (FO) and renal replacement therapy (RRT) at PICU day 7, as well as PICU length of stay (LOS).
Results
Over our study period 1103 RAIs resulted from 1022 patients. After excluding patients with ESRD and with admissions less than 2 days, 627 RAIs from 569 patients were examined, of which 63 (10.0%) were RAI+. The incidence of Day 2-4 sAKI was higher in the RAI+ cohort compared to RAI- patients (38% vs. 1.8%, p<0.001). With the addition of uNGAL, the rate of sAKI in RAI+NGAL+ was higher than in RAI+NGAL- or RAI- patients (55.5% vs 17.6% vs 1.8%, p<0.0001). RAI+ patients had a higher need for RRT compared to RAI- patients (13% vs. 0.4%, p<0.001). PICU and hospital length of stay were also longer in RAI+ patients: PICU LOS 4.6 vs 3.1 days (p<0.02), Hospital LOS 17.8 vs 7.5 days (p<0.001). There was no significant difference between RAI+/- patients in the occurrence or duration of FO (p=0.14, p=0.11).
Conclusion
The RAI alone continues to be a good rule-out test for sAKI in the PICU, and addition of NGAL is promising for improved sAKI prediction. Additional research needs to be conducted on a larger sample size to better assess the clinical relevance of the RAI NGAL model in predicting sAKI and other relevant outcomes.
Performance of RAI and NGAL in Predicting sAKI
| RAI ≥ 8 on Day 2-4 sAKI | RAI ≥ 8 + NGAL ≥ 150ng/mL on Day 2-4 sAKI | |
| Negative Predictive Value (NPV) | 98.2% (97.1% to 99.0%) | 97.8% (96.7% to 98.5%) |
| Positive Predictive Value (PPV) | 38.1% (29.8% to 47.2%) | 55.2% (39.6% to 69.8%) |
| Sensitivity | 70.6% (52.5% to 84.9%) | 55.2% (35.7% to 73.6%) |
| Specificity | 93.4% (91.1% to 95.3%) | 97.8% (96.2% to 98.8%) |
Funding
- NIDDK Support