Kidney Week

Abstract: PO2296

Developing a Strategy for Routine Reporting of Estimated Glomerular Filtration Rate on Pediatric Laboratory Results

Session Information

• Pediatric Nephrology: Benign Urology, AKI, Neonatal Nephrology, and Case Reports
  October 22, 2020 | Location: On-Demand
  Abstract Time: 10:00 AM - 12:00 PM

Category: Pediatric Nephrology

• 1700 Pediatric Nephrology

Authors

• Modi, Zubin J., University of Michigan Medical School, Ann Arbor, Michigan, United States

Zubin J. Modi,

• Blazius, Brooke Ann, University of Michigan Medical School, Ann Arbor, Michigan, United States

Brooke Ann Blazius,

• Troost, Jonathan P., University of Michigan Medical School, Ann Arbor, Michigan, United States

Jonathan P. Troost,

• Luckritz, Kera E., University of Michigan Medical School, Ann Arbor, Michigan, United States

Kera E. Luckritz,

• Balis, Ulysses G. J., University of Michigan Medical School, Ann Arbor, Michigan, United States

Ulysses G. J. Balis,

• Gipson, Debbie S., University of Michigan Medical School, Ann Arbor, Michigan, United States

Debbie S. Gipson,

Background

Routine calculation of estimated glomerular filtration rate (eGFR) is not present in laboratory reporting for children, potentially leading to low recognition of decreased eGFR. The study aim was to define options for pediatric eGFR reporting that optimize accuracy and minimize the impact of missing concurrent height in routine lab reporting.

Methods

Data was extracted from the Michigan Medicine Data Warehouse on all patients aged 1-19 years with serum creatinine between 2017-2018. Creatinine-cystatin C-based CKiD (Cr-CysC, Schwartz, 2012) and creatinine-based ‘Bedside’ (Schwartz 2009) equations were used to calculate eGFR. Correlations were tested between eGFR calculated with same day height values against eGFR calculated with height imputation of 50^th percentile for age and sex or patient-specific historical height percentiles. Scatterplots, Pearson’s correlation coefficients, and predictive characteristics were calculated.

Results

There were 109,090 serum creatinine measurements, and 35% had concurrent heights. There were 1770 Cystatin C measurements. eGFR by Bedside equation was abnormal (<90 mL/min/1.73m2) in 25% of measurements. Cr-CysC (r=0.99) and Bedside (r=0.98) equations had excellent correlations for measured vs imputed 50^th percentile height eGFRs (Figure). Patient-specific height percentile imputed for the Cr-CysC (r=0.99) and Bedside (r=0.99) eGFR also had tight correlations. Discrimination of abnormal eGFR was excellent for the Cr-CysC (94% sensitivity and 95% specificity) and Bedside (90% sensitivity and 97% specificity) equations using 50^th percentile height.

Conclusion

Imputation of 50^th percentile or patient-specific historic height percentile enables eGFR calculation that correlates with eGFR using same day height. For ease of implementation and with high sensitivity and specificity, use of 50^th percentile height may be the preferred approach for pediatric eGFR reporting when patient-specific heights are not available.