Accuracy to Estimate Rates of Decline in Glomerular Filtration Rate in Renal Transplant Patients

BACKGROUND: We examined the use of the Cockroft Gault (C-G) test, Modified Diet in Renal Disease 2 (MDRD2) test, and inverse serum creatinine (Delta1/Scr) to estimate rates of decline in renal transplant function using isotope glomerular filtration rate (GFR) as a reference test. METHODS: Percent changes in estimated GFR (DeltaeGFR) were compared to simultaneous changes in isotope GFR (DeltaiGFR) in 72 patients. RESULTS: The number of iGFR was 508 with a mean of 7.15+/-3.15 scans per patient. There was a decline in iGFR of 16.14+/-21.37 ml/min over the study duration of 88.9+/-57.6 months. DeltaeGFR and Delta1/Scr correlated significantly with DeltaiGFR. Accuracy to predict DeltaiGFR from the eGFRs was limited to <65% concordance within 30% range from changes in iGFR. Slope analyses showed a significantly lower percent annual loss in mean iGFR of 6.03% than that of the C-G of 8.62% and MDRD2 of 8.96% (P<0.001). The within patient variability measured from the standard deviation (ml/min) of root mean square of 4.69 for iGFR was significantly higher than that for C-G and MDRD2 of 2.46 and 2.94, respectively. iGFR and eGFR at first observation correlated significantly (P<0.001) with last observation. CONCLUSIONS: iGFR is significantly more variable within patient than the other predictors, and the two estimators predict the iGFR with a high sensitivity but low specificity. This is a clinically reasonable combination. Predicted percent of annual loss in iGFR appears to be smaller than that using the two estimators.