Abstract
1301
Background: While serum labs allow estimation of GFR (eGFR), measurement of GFR by scintigraphy (nuGFR) remains the clinical reference standard. The aim of our study was to characterize agreement between eGFR and nuGFR in a large pediatric cohort.
Methods: This was a retrospective, cross-sectional study of patients who had both nuGFR and serum labs (creatitinine, cystatin-c, BUN) obtained within a 7 day interval between January 2017 and August 2019. eGFR was calculated using multiple equations (Bedside Schwartz, Schwartz-Lyon, Hoek, Larsson, CKiD 2012 and Zappitelli). Agreement was calculated using IntraClass correlation and Bland-Altman analysis with linear regression used to define predictors of agreement. Results: 504 patients (216 male) had paired nuGFR and eGFR measurements. Mean ±SD age, body weight, and height were 7.6±5.5 years, 30.4±22.7 kg, and 1.2±0.4 m. Mean nuGFR was 140±50 mL/min/1.73m2. Of the equations for eGFR, Zappitelli had the highest absolute agreement with nuGFR (ICC=0.60; 95% CI: 0.55-0.64). Mean nuGFR-eGFR difference was 1.51 mL/min/1.73m2 (95% limits: -89.44 to 92.46). For categorical prediction of a nuGFR <60 mL/min/1.73m2, eGFR <60 mL/min/1.73m2 by Zappitelli was 65% sensitive, 99% specific, and 96% accurate. Univariate regression identified age (β=-2.18; p<0.001), height (β=-0.36; p<0.0001), weight (β=-0.48; p<0.0001), serum creatinine (β=-25.2; p=0.0006), male gender (β=-16.5; p=0.0008), soft tissue tumor diagnosis (β=11.8; p=0.02), liquid tumor diagnosis (β=-33.03; p<0.001), and blood dyscrasia (β=15.7 ; p=0.03) as predictors of agreement between eGFR and nuGFR. Conclusion: In a diverse pediatric sample, agreement between eGFR and nuGFR is fair to good but sensitivity to a GFR <60 mL/min/1.73m2 is only 65%. Agreement between eGFR and nuGFR is impacted by measures of patient size and by tumor type. In patients with concern for impaired renal function nuGFR should remain the reference standard for diagnosis.