Abstract: SA-PO358
Performance of CKD-EPI GFR Estimating Equations in a Cohort of Young Adults with Pediatric CKD
Session Information
- Pediatric Nephrology - III
November 04, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Pediatric Nephrology
- 1900 Pediatric Nephrology
Authors
- Ng, Derek K., Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland, United States
- Inker, Lesley Ann, Tufts Medical Center, Boston, Massachusetts, United States
- Furth, Susan L., The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
- Warady, Bradley A., Children's Mercy Kansas City, Kansas City, Missouri, United States
- Schwartz, George J., University of Rochester Medical Center, Rochester, New York, United States
Group or Team Name
- Chronic Kidney Disease in Children Study.
Background
The Chronic Kidney Disease in Children (CKiD) study previously documented an overestimation bias by the 2009 age-sex-race CKD-EPI eGFR among young adults aged 18 to 24 with a history of pediatric CKD. However, the recently updated 2021 race-free CKD-EPI equations have not been evaluated among young adults with a history of pediatric CKD.
Methods
Using CKiD data from young adults 18 to 25 years, we evaluated three CKD-EPI equations: 2021 SCr (with age and sex; AS), 2012 CysC, and the 2021 SCr-CysC (AS) Bias (eGFR - mGFR), accuracy (proportion within 10% and within 30%) and root mean square error (RMSE) were estimated.
Results
Among 313 person-visits, the median age was 19.4 years [IQR: 18.6, 20.6] and the median mGFR was 44.8 ml/min|1.73m2 [IQR: 31.3, 65.1]. Table 1 summarizes agreement metrics. The average biases were +8.81 (95%CI: +7.09, +10.53), +3.73 (+2.04, +5.41), and +4.31 (+2.84, +5.78) for the SCr, CysC, and combined CKD-EPI equations, respectively. The proportion of eGFR within 30% of mGFR was 69.0%, 78.3% and 81.5%, respectively. The lowest RMSE was observed for the combined equation (11.75) and the creatinine-based and cystatin c-based equations RMSEs were 14.83 and 14.03, respectively.
Conclusion
Application of the new CKD-EPI SCr equation to CKiD data demonstrated overestimation of measured GFR and high root mean square errors. The cystatin-based and combination equations performed better than creatinine alone. As previously published, the race-free CKiD U25 equations had no significant bias (range: -0.89 to +0.70) and p30 between 85-90% among young adult participants, but performance is expected to be better in internal validation because it randomly differed from the data used to develop the U25 eGFR. For pediatric CKD patients transitioning to adulthood, care needs to be exercised when using disparate equations that yield abrupt differences in estimates of GFR. Use of U25 eGFR through 25 years is recommended for this population.
Agreement between CKD-EPI eGFR and mGFR (bias, p10, p30, and root mean square error (RMSE) among 208 young adults participants in CKiD (≥ 18 years) with 313 assessments of mGFR
| CKD-EPI eGFR | eGFR – mGFR (95% CI) | p10: eGFR within 10% of mGFR | p30: eGFR within 30% of mGFR | RMSE ml/min|1.73m2 |
| 2021 SCr-only | 8.81 (7.09, 10.53) | 26.2% | 69.0% | 14.83 |
| 2012 CysC only | 3.73 (2.04, 5.41) | 30.0% | 78.3% | 14.05 |
| 2021 SCr- and CysC-combined | 4.31 (2.84, 5.78) | 37.4% | 81.5% | 11.75 |
Funding
- NIDDK Support