Incorporating Time from Donation (TFD) to the MDRD and CKD-EPI Formulas Enhances Their Performance in Kidney Donors
Department of Medicine, University of Minnesota, Minneapolis, MN
Surgery, University of Minnesota, Minneapolis, MN
Meeting: 2013 American Transplant Congress
Abstract number: C1222
Introduction:
The reexpressed MDRD study equation is least biased in estimating GFR in former kidney donors. Hyperfiltration is not limited to the first year after donation. We hypothesized that adding a term for time from donation (TFD) may enhance performance of the MDRD and CKD-EPI formulas.
Methods:
We randomly selected two groups of donors with iohexol GFR (n=257): a development (n=129) and a validation cohort (n=128). For the development cohort, a linear regression model was fit with the covariates of log creatinine, age and gender. The outcome was log iohexol GFR. The new MDRD equation from the development cohort was as follows:
MDRD*=289*(Age-0.339)*(SCr-0.485)*(0.850 (if Female)). With TFD:
MDRD**=310*(Age-0.371)*(SCr-0.432)*((TFD) 0.025)*(0.858 (if Female))
The CKD-EPI Creatinine formula was first refit and TFD was added:
CKD-EPI Creatinine**=327*(min(SCr/7,1)-0.409)*(max(Scr/7,1)-0.538)*(Age-0.371)*((TFD)0.013) [if Female]
CKD-EPI Creatinine**=302*(min(SCr/9,1)0.186)*(max(Scr/9,1)-0.362)*(Age-0.365) *((TFD)0.047) [if Male]
These four models were then fit to the validation cohort.
Results:
Mean age at donation was 41.0 ± 11.0 years and 11.9 ± 9.0 years have elapsed since donation. Iohexol GFR was 71.9 ± 11.9 ml/min/1.73m2. Incorporating TFD improved precision (from 0.51 to 0.58), reduced the IQR of bias and increased estimates within 10% of iohexol GFR from 40.7% to 65.6%. Performance improved similarly with the refit CKD-EPI Creatinine model.
| Original MDRD (n=257) | MDRD with time after donation (n=128) | |
|---|---|---|
| Mean (SD) | 71.6 (14.9) | 71.1 (8.3) |
| Bias (SD) | -0.3 (11.7) | -1.0 (7.7) |
| IQR of Difference | 15.7 | 9.8 |
| Relative Bias (%) (SD) | 12.9 (10.6) | 8.4 (7.1) |
| R² | .41 | .58 |
| RMSE | 9.1 | 7.8 |
| Within 10% of mGFR (%) | 116 (45.1) | 84 (65.6) |
| Within 30% of mGFR (%) | 242 (94.2) | 126 (98.4) |
| Within 50% of mGFR (%) | 255 (99.2) | 128 (100.0) |
| Original CKD-EPI Creatinine (n=257) | CKD-EPI Creatinine With time after donation (n=128) | |
|---|---|---|
| Mean (SD) | 78.4 (16.3) | 71.3 (8.6) |
| Bias (SD) | 6.5 (12.1) | -0.9 (7.8) |
| IQR of Difference | 17.7 | 10.0 |
| Relative Bias (%) (SD) | 15.6 (9.8) | 8.5 (6.9) |
| R² | .45 | .57 |
| RMSE | 8.8 | 7.8 |
| Within 10% of mGFR (%) | 101 (39.3) | 84 (65.6) |
| Within 30% of mGFR (%) | 224 (87.2) | 126 (98.4) |
| Within 50% of mGFR (%) | 252 (98.1) | 128 (100.0) |
Conclusion:
Incorporating a term for time from donation in the reexpressed MDRD and CKD-EPI Creatinine formulas provides major improvement in this exclusively Caucasian group of donors.
To cite this abstract in AMA style:
Ibrahim H, Jackson S, Matas A, Issa N, Spong R, Kukla A. Incorporating Time from Donation (TFD) to the MDRD and CKD-EPI Formulas Enhances Their Performance in Kidney Donors [abstract]. Am J Transplant. 2013; 13 (suppl 5). https://atcmeetingabstracts.com/abstract/incorporating-time-from-donation-tfd-to-the-mdrd-and-ckd-epi-formulas-enhances-their-performance-in-kidney-donors/. Accessed November 22, 2024.« Back to 2013 American Transplant Congress