Serum creatinine is an unreliable marker of renal function in patients undergoing heart transplantation

BackgroundRenal dysfunction is a common complication after heart transplantation (Htx). Glomerular filtration rate (GFR) can be assessed by various estimating equations (eGFR). We evaluated the correlation, agreement, and accuracy between eGFR and mGFR and the ability of eGFR to track changes in mGFR early after Htx.MethodsA single-center prospective observational study on 55 patients undergoing Htx. Serum creatinine and mGFR (plasma clearance of Cr51-EDTA or iohexol) were measured preoperatively and on the fourth postoperative day. The accuracy of eGFR to predict true mGFR was calculated as the percentage of patients with an eGFR within 30% of mGFR (P30). The agreement between eGFR and mGFR was assessed according to Bland and Altman. A four-quadrant plot was made to evaluate the ability of eGFR to track changes in mGFR.ResultsThe accuracy of eGFR to assess mGFR was 52%. The bias was 11.2 +/- 17.4 mL/min/1.72 m2. The limits of agreement were -23.0 to 45.4 mL/min/1.72 m2 and the error 58%. The concordance rate between eGFR and mGFR was 72%.ConclusionseGFR underestimated mGFR and the agreement between eGFR and mGFR was low with an unacceptably large between-group error and low accuracy. Furthermore, the ability of eGFR to assess changes in mGFR, postoperatively, was poor. Thus, the use of estimating equations from serum creatinine will not adequately assess renal function early after major heart surgery. To gain adequate information on renal function early after Htx, GFR needs to be measured, not estimated.