Comparison of methods of estimating creatinine clearance in pediatric patients

• Published in: American journal of health-system pharmacy 2017-06, Vol.74 (11), p.826-830
• Main Authors: Padgett, Danielle, Ostrenga, Andrew, Lepard, Lindsey
• Format: Article
• Language: English
• Subjects: Analysis; Care and treatment; Chronic kidney failure; Correlation coefficients; Creatinine; Creatinine tests; Demographics; Dilution; Glomerular filtration rate; Mass spectrometry; Mass spectroscopy; Mathematical models; Measurement; Measurement methods; Methods; Pediatric diseases; Pediatrics; Renal function; Test procedures; Urine
• ISSN: 1079-2082; 1535-2900
• DOI: 10.2146/ajhp151004

A retrospective study was conducted to compare various methods of measuring serum creatinine (SCr) values for use in pediatric renal function assessments, including a method aligned with a recently implemented national SCr testing standard. Demographic, medication-use, and selected laboratory data were collected from the hospital records of a sample of pediatric patients ( = 91) who underwent 12- or 24-hour timed urine collection for determination of creatinine clearance (CL ) over a 2-year period. Documented CL values measured via the timed urine collection method were compared with investigator-calculated estimates of CL or glomerular filtration rate (GFR) derived using 3 SCr-based methods: the Counahan-Barratt equation; the original Schwartz equation; and the "bedside IDMS-traceable Schwartz equation," a modified version of the Schwartz equation reflecting the recent shift toward isotope dilution mass spectrometry (IDMS) methods of SCr measurement, which have been found to yield SCr values 10-20% lower than those derived by older methods, potentially resulting in GFR overestimation if traditional formulas for estimating GFR are used. Comparisons of timed urine collection-derived CL values with CL values derived from the 3 comparator equations indicated significant levels of bias in all cases, with calculated correlation coefficients of 0.71 for the original Schwartz equation, 0.72 for the bedside IDMS-traceable Schwartz equation, and 0.72 for the Counahan-Barratt equation. Pediatric CL values calculated using the original Schwartz, bedside IDMS-traceable Schwartz, and Counahan-Barratt equations were well correlated, but none of the 3 equations yielded values that correlated well with CL values derived via the gold-standard method of timed urine collection.