Kidney Week

Abstract: TH-OR047

Adaptation of the Kidney Failure Risk Equation to Predict Risk of ESRD in Children

Session Information

• Featured Research in Pediatric Nephrology
  October 25, 2018 | Location: 26A, San Diego Convention Center
  Abstract Time: 05:42 PM - 05:54 PM

Category: Pediatric Nephrology

• 1600 Pediatric Nephrology

Authors

• Ku, Elaine, UC San Francisco, San Francisco, California, United States

Elaine Ku, MD, FASN

• Wuehl, Elke, University of Heidelberg, Heidelberg, BW, Germany

Elke Wuehl,

• Winnicki, Erica, UC San Francisco, San Francisco, California, United States

Erica Winnicki,

• Mitsnefes, Mark, Cincinnati Children's Hospital, Cincinnati, Ohio, United States

Mark Mitsnefes,

• McCulloch, Charles E., University of California San Francisco, San Francisco, California, United States

Charles E. McCulloch,

• Warady, Bradley A., Children's Mercy Kansas City , Kansas City, Missouri, United States

Bradley A. Warady,

• Furth, Susan L., The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States

Susan L. Furth,

• Schaefer, Franz S., University of Heidelberg, Heidelberg, BW, Germany

Franz S. Schaefer,

Background

The Kidney Failure Risk Equations (KFREs) have been shown to predict ESRD risk in children with mild-to-moderate CKD in the US. Whether the KFRE accurately predicts progression to ESRD in children with more advanced CKD and children outside of the US is unknown. KFREs to predict ESRD risk in North American (NA) and non-North American (NNA) adults are currently available.

Methods

We studied 653 children from 4C study, 341 children from ESCAPE trial, and 633 children from CKiD study with an eGFR <60 mL/min/1.73 m². The primary predictors were the 4-variable (age, sex, bedside eGFR, albumin-to-creatinine ratio) NA and NNA KFRE score which provides estimates of the risk of ESRD over a 2- and 5-year horizon. C-statistics were used to compare discrimination of the risk for ESRD by the KFRE, and this predicted risk of ESRD was compared to observed failure rates.

Results

Median follow-up time was 2.8 years in 4C, 4.7 years in ESCAPE, and 3.8 years in CKiD. The NA and NNA 4-variable KFRE strongly discriminated risk of ESRD in children outside the US, with a C-statistic of 0.82 and 0.79 in 4C and 0.90 and 0.87 in ESCAPE over a 2- and 5-year horizon, respectively. The predicted risk closely matched the observed ESRD rates in CKD stage 3, but over-predicted ESRD risk in CKD stage 4-5 across all cohorts (Figure 1). After we re-calibrated the equation, performance of the KFRE in advanced CKD improved (Figure 2).

Conclusion

The KFRE is a simple tool that strongly discriminates the risk of ESRD in children with CKD in the US and internationally, although its performance was less robust in children with advanced CKD. With re-calibration, its performance in more advanced CKD was substantially improved.

Funding

• NIDDK Support