Performance of creatinine-based estimates of GFR in kidney transplant recipients: A systematic review

Background: Glomerular filtration rate (GFR) commonly is estimated in kidney transplantation by using creatinine-based estimation equations. The performance of these equations in kidney transplant recipients is unclear, with conflicting results between studies. Study Design: Systematic review. Setting & Population: Stable adult kidney transplant recipients more than 6 months posttransplantation. Selection Criteria: Reporting of or ability to calculate from available data the GFR estimation equation bias (mean difference between measured GFR and estimated GFR) and percent accuracy (percentage of GFR estimates within 10%, 20%, or 30% of measured GFR). Index Tests: Creatinine-based GFR estimation equations (Cockcroft-Gault, 6-variable Modification of Diet in Renal Disease [MDRD] Study, 4-variable MDRD Study, and Nankivell). Reference Tests: GFR determination using plasma or renal clearance of inulin, radioisotopes, or nonradiographic contrast. Results: The search yielded 23 studies. For the 4-variable MDRD Study equation, bias ranged from -11.4 to +9.2 mL/min/1.73 m(2) (0.15 mL/s/1.73 m(2)). Only 76% of estimates were within 30% of measured GFR. For the Cockcroft-Gault equation, bias ranged from -4.0 to +16 mL/min/1.73 m(2) and 73% of estimates were within 30% of measured GFR. For the Nankivell equation, bias ranged from -1.4 mL/min to 36.3 mL/min with a 30% accuracy of only 68%. Limitations: This review is limited by the inability to pool bias data, lack of calibration of serum creatinine in the majority of studies, and inclusion of nonindependent observations in many studies. Conclusions: Differences in patient populations, baseline GFRs of the study group, reference standard GFR used, and creatinine assay calibration likely account for the heterogeneity in results. These factors need to be considered by investigators and clinicians when interpreting estimates of GFR in kidney transplant recipients.