How does the MDRD Study equation compare with serum creatinine in routine healthcare? Anatomy of MDRD‐eGFR

Abstract

The Modification of Diet in Renal Disease Study equation–estimated glomerular filtration rate (MDRD–eGFR) as a marker for chronic kidney disease, with a single decision level for both genders and all adult ages, requires that the calculated quantity is independent of gender and age. In a retrospective study of S‐Creatinine concentrations from laboratory information systems of hospitals in the UK and Sweden, comprising about 140,000 results in total, it was found that the MDRD–eGFR indeed differs between genders and that it varies with age more than the S‐Creatinine concentration does. If the age compensation is deleted from the algorithm, the relative changes in the MDRD–eGFR decrease and become almost the same as those for S‐Creatinine concentrations. The difference between the genders could probably be overcome by increasing the “if female factor”. We used Pt‐Iohexol and S‐Creatinine concentrations measured simultaneously to estimate the performance of the MDRD–eGFR in relation to Pt‐Iohexol clearance. The Pt‐Iohexol varies considerably between patients with the same S‐Creatinine concentrations, a difference that is not reflected in the MDRD–eGFR. It is concluded that the mathematical transformation of S‐Creatinine concentrations does not add any diagnostic value. On the contrary, an increased measurement uncertainty is unavoidable with the use of factors and exponents. The uncertainty is greater than any difference between age and gender in adjacent age groups. There is no compensation for the individual relation between S‐Creatinine and Iohexol clearance, and the equation does not consider the individual body surface area; it is therefore inadvisable to use the MDRD–eGFR values as the basis for administration of drugs excreted by the kidneys.

• Diagnostic efficiency
• likelihood ratio
• modelling