The Sask formula to estimate glomerular filtration rate in renal transplant patients

The aim of this study was to develop a glomerular filtration rate (GFR) equation for renal transplant and compare its performance with Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and isotope dilution mass spectrometry (IDMS) equations. Using genetic symbolic regression analysis, the Sask equation was developed from a training sample of 772 isotope GFR (iGFR) scans performed in 99 transplanted patients. It was then validated in two other samples of 269 scans with the same number of patients. Standard methods including accuracy at 30% range from reference values were compared. In two validation samples, the Sask equation maintained the lowest bias of -0.7 ± 19.0 and 0.4 ± 18.4 ml/min/1.73 m(2) (p < 0.05) versus -3.1 ± 19.6 and -7.2 ± 18.8 ml/min/1.73 m(2) for CKD-EPI and -2.2 ± 19.2 and -6.5 ± 18.3 ml/min/1.73 m(2) for IDMS, respectively. In those with iGFR between 90 and 30 ml/min/1.73 m(2), the Sask equation demonstrated: (1) the lowest bias of -1.0 ± 15.7 and -0.4 ± 15.7 ml/min/1.73 m(2) (p < 0.05 vs. other tests); (2) an accuracy of 75.5 and 76.1% (p < 0.05 vs. other tests), and (3) a mean percentage error of 1.9 ± 30.5 and -4.1 ± 31.4 ml/min/1.73 m(2) (p < 0.05 vs. other tests). Analysis based on gender demonstrated improved performance in the total and subtotal female populations with GFR between 90 and 30 ml/min/1.73 m(2). The CKD-EPI and Sask equations performed better than IDMS. The Sask equation demonstrated improved bias over CKD-EPI, with iGFR between 90 and 30 ml/min/1.73 m(2), particularly in females.