Effects of analytic variations in creatinine measurements on the classification of renal disease using estimated glomerular filtration rate (eGFR)

BACKGROUND: This study uses the Isotope Dilution Mass Spectrometer-Traceable Modification of Diet in Renal Disease (MDRD) equation to evaluate the effects of analytic variation of serum creatinine on estimated glomerular filtration rate (eGFR). The equation was inverted to provide critical creatinine levels corresponding to eGFRs equal to 15, 30, and 60 mL/min/1.73 m(2). METHODS: eGFRs were calculated for two populations of patients representing clinic outpatients and hospitalized inpatients. Simulation studies evaluated the effects of analytic bias for these populations. Differences between creatinine tests ordered on consecutive blood draws were analyzed for within-subject eGFR variation. Finally, propagation of error techniques established the relationship between creatinine imprecision and eGFR imprecision. RESULTS: eGFR of 60 mL/min/1.73 m(2) corresponds to creatinine levels of 80-156 micromol/L. A 20-micromol/L negative shift of creatinine approximately doubles the percentage of patients classified in various stages of decreased renal function, whereas a positive shift approximately halves the percentage. The central 95% limits for within-subject variation of eGFR are 33% for outpatients, 38% for inpatients and 40% across the groups. eGFR imprecision is approximately 15%-20% higher than creatinine imprecision. CONCLUSIONS: Small analytic changes in serum creatinine create major shifts in the distributions of eGFR, which can cause large differences in the classification of patients. The within-subject variations in eGFR over time, especially between hospital and clinic measurements, can be large. Therefore, tight control of laboratory analysis is important.