Evaluation of the estimated variables for scaling glomerular filtration rate of renal patients: a repeated...

Using a best variable to scale glomerular filtration rate (GFR) is important for clinical practice. The variables, estimated by equations regressed from a healthy population, are usually used in scaling GFR of renal patients. However, because the predicted variables may deviate in renal patients, it is necessary to verify whether these variables can be used to reduce the variability of GFR of renal patients. This study was designed to use repeated measures analyses to identify the best variable for scaling GFR of renal patients.

Patients with non-obstructive renal diseases were enrolled in this study. The absolute GFRs of (99m)Tc-DTPA renography (gGFR) and plasma clearance (pGFR) were measured. The indices relating to between-subjects variability, such as Passing and Bablok regression, intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC) were used to identify the best variable from body surface area (BSA), extracellular fluid volume (ECV), lean body mass (LBM), total body water (TBW), body mass index (BMI), and metabolic rate (MR).

For the scaled indices related to between-subjects variability, ICC and CCC identified the same ranking sequence (BMI < LBMB(B; Boer) < LBMJ(J; James) < TBW < ECVB(B; Bird) < ECVS(S; Silva) < BSA < MR). In the Passing and Bablok regression, the ratio of residual standard deviation to pooled standard deviation (RSD/PSD) produced the same ranking sequence as that identified by ICC and CCC.

The estimated metabolic rate can explain most between-subjects variability of GFR, and seems to be the best variable for scaling GFR of renal patients.