Estimated GFR for Living Kidney Donor Evaluation

Abstract

All living kidney donor candidates undergo evaluation of GFR. Guidelines recommend measured GFR (mGFR), using either an endogenous filtration marker or creatinine clearance, rather than estimated GFR (eGFR), but measurement methods are difficult, time consuming and costly. We investigated whether GFR estimated from serum creatinine (eGFRcr) with or without sequential cystatin C is sufficiently accurate to identify donor candidates with high probability that mGFR is above or below thresholds for clinical decision making. We combined the pretest probability for mGFR thresholds <60, <70, ≥80, and ≥90 mL/min per 1.73 m(2) based on demographic characteristics (from the National Health and Nutrition Examination Survey) with test performance of eGFR (categorical likelihood ratios from the Chronic Kidney Disease Epidemiology Collaboration) to compute posttest probabilities. Using data from the Scientific Registry of Transplant Recipients, 53% of recent living donors had predonation eGFRcr high enough to ensure ≥95% probability that predonation mGFR was ≥90 mL/min per 1.73 m(2) , suggesting that mGFR may not be necessary in a large proportion of donor candidates. We developed a Web-based application to compute the probability, based on eGFR, that mGFR for a donor candidate is above or below a range of thresholds useful in living donor evaluation and selection.