To the Editor: I read with interest the position statement1 recommending that Australian laboratories use the recently published MDRD (Modification of Diet in Renal Disease) equation2 to estimate glomerular filtration rate (GFR) and automatically report the estimated GFR (eGFR) when the value is less than 60 mL/min/1.73m2.

Serum creatinine concentration provides an imperfect guide to renal function because it is affected by factors other than glomerular filtration. To improve the accuracy of the estimate of GFR, more than 40 equations and nomograms have been developed over the years. One of these, the MDRD equation, was found to give a better estimate than the most commonly used equation (the Cockcroft–Gault equation3) in a study in a study that (it might be argued) favoured the MDRD method.2 The abbreviated MDRD equation recommended in the position statement has only been published as an abstract.4

The position statement notes that “an uncorrected eGFR may be preferred for clinical use in some situations, such as drug dosing”, which requires the reported eGFR to be modified for body surface area (BSA). However, the formula quoted in the position statement is incorrect. The correct formula, published in 19165 is:

This complicated equation requires data that are infrequently available, and it is arguable that performing this extra calculation may negate any benefit from the “improved” MDRD estimate. While the authors of the position statement do not recommend that reported eGFRs be used to calculate drug doses for patients with renal impairment, it is likely that they will be used for that purpose, without modification, by prescribers unaware of the limitations.

For patients requiring drug dosage reductions, it is doubtful whether the MDRD formulas give a more accurate eGFR than the Cockcroft–Gault equation, which has not only provided useful bedside estimates of GFR for decades, but has also been a valuable teaching tool. By incorporating age, sex and weight, the Cockcroft–Gault equation reminds us that these factors affect creatinine production and hence serum creatinine concentration. The same cannot be said for the MDRD formulas, which are complex, difficult to remember, and require a personal digital assistant or similar device for doing the calculations.

In estimating the dose of renally cleared drugs, it has been observed that “what is essential for therapeutic decisions is knowing that a patient’s renal function is impaired and about to what extent, rather than the exact glomerular filtration rate”.6 Since estimates of renal function will never preclude drug concentration monitoring, modest improvements in accuracy of an estimated parameter are of limited benefit. I recommend the continued use of the Cockcroft–Gault equation for estimating doses of renally cleared drugs.