Performance of Creatinine and Cystatin Equations in People with Heart Failure

Background: Performance of eGFRs among patients with heart failure (HF) may be worse than in the average population, with implications for clinical care. Methods: We retrieved cases with mGFR (urinary iothalamate) performed at Mayo Clinic between 2011-23 that had an echocardiogram performed within 1 year and serum creatinine (Cr Cohort) or Cr and cystatin (CrCysC Cohort) available within 7 days of the mGFR. Patients were then classified based on ICD codes for HF present 1 year prior to the mGFR and ejection fraction (< or >=50%). CKD-EPI and EKFC eGFRcr, eGFRcys, and eGFRcrcys were computed. Bias was defined as mGFR minus eGFR, and P30 as the proportion of people with eGFR >=30% difference to mGFR. Results: Results are shown in Table 1. eGFRcr was the most biased equation, prone to overestimation of mGFR in HF groups compared to NoHF. eGFRcr also had the worst precision in the HF groups, with larger errors in HFrEF. eGFRcrcys was the most accurate in NoHF and HFpEF, while eGFRcysC was the most accurate in HFrEF. The CKD-EPI and EKFC performed similarly. Conclusion: In people with HF, eGFRcr tends to overestimate mGFR and has worse precision compared to people without HF. eGFRcrcys was the most accurate in NoHF and HFpEF, while eGFRcys was the most accurate in the HFrEF group. - Cr Cohort CrCys Cohort NoHF (n=13131) HFpEF (n=1517) HFrEF (n=1011) NoHF (n=1886) HFpEF (n=184) HFrEF (n=115) Median Bias (CI) CKDEPIcr 0.5 (0.2 , 0.8) -3.7 (-4.7 , -2.8) -4 (-5.1 , -3) 0 (-0.6 , 0.9) -6 (-8.1 , -4.5) -8.1 (-11.4 , -4.4) CKDEPIcys - - - 4.2 (3.3 , 5.3) 4.7 (3.7 , 6.4) 2.6 (-0.4 , 4.7) CKDEPIcrcys - - - 1.1 (0.5 , 1.8) 0.5 (-0.8 , 2.1) -1.5 (-3.4 , 1.1) EKFCcr 6 (5.6 , 6.3) 0.7 (0 , 1.7) -0.5 (-1.6 , 0.4) 5.1 (4.2 , 6) -2.2 (-3.9 , -0.4) -4 (-6.7 , -0.5) EKFCcys - - - 3.3 (2.2 , 4.4) 0.9 (-1.3 , 2.7) -0.9 (-3.4 , 0.7) EKFCcrcys - - - 4.1 (3.3 , 4.7) -1.2 (-3.1 , -0.3) -4.1 (-7.9 , -0.7) 1 - P30 (CI) CKDEPIcr 22.4 (21.7 , 23.1) 30.5 (28.1 , 32.8) 37.6 (34.4 , 40.4) 22.5 (20.6 , 24.4) 35.9 (28.8 , 42.4) 60 (51.3 , 68.7) CKDEPIcys - - - 24.7 (22.7 , 26.6) 32.1 (25.3 , 38.6) 27.8 (20.5 , 36.1) CKDEPIcrcys - - - 15.5 (14.0 , 17.2) 25.0 (19.0 , 31.0) 33.9 (26.1 , 42.6) EKFCcr 23.3 (22.5 , 24) 27.9 (25.4 , 30.2) 35 (32.2 , 37.8) 22.9 (21.0 , 24.6) 34.2 (27.2 , 42.1) 58.3 (49.6 , 67.8) EKFCcys - - - 20.8 (18.7 , 22.6) 31 (23.9 , 38) 33 (23.5 , 40.9) EKFCcrcys - - - 15.4 (13.7 , 17.1) 27.2 (19.6 , 33.4) 39.1 (28.7 , 47) RMSE CKDEPIcr 19.5 17.4 19.4 18.2 18.7 22.0 CKDEPIcys - - - 19.2 16.1 13.4 CKDEPIcrcys - - - 16.0 13.3 13.5 EKFCcr 20.7 16.3 18.0 19.3 16.6 19.6 EKFCcys - - - 18.2 14.9 12.8 EKFCcrcys - - - 17.0 13.2 13.8 Bias was computed as mGFR - eGFR. Bias and RMSE are expressed as mL/min/1.73m2.