Enhancing script safety for kidney patients
Thank you for highlighting the important issues around estimated glomerular filtration rate (eGFR) reporting and drug prescribing. However, your news item (News, 6 September) contains some inaccuracies and misconceptions, which need to be clarified.
• eGFR reporting was recommended by the national service framework for renal services and has allowed the identification of previously unrecognised kidney disease.
Although not a perfect estimate of GFR, the nationally recommended approach to eGFR reporting performs better than either measured creatinine clearance (by
24-hour urine collection) or estimated creatinine clearance (by Cockcroft-Gault formula) in patients with chronic kidney disease.
All three measures perform better than serum creatinine, which previously led to systematic under-diagnosis of kidney disease in patients with low muscle mass, such as older patients and women.
• It is necessary for prescribers to recognise the presence of kidney disease to prompt them to stop or avoid nephrotoxic drugs, and to modify the dose of other renally excreted drugs.
This is particularly likely to be relevant in the elderly, who suffer a disproportionate number of adverse drug reactions.
• eGFR is adjusted for body surface area (BSA) and is reported as mL/min/1.73m2. GFR varies with size, and adjustment for BSA allows the reporting of a single reference range. Renal drug excretion, however, relates more closely to absolute GFR (that is, not adjusted for BSA). For most patients and with most drugs, the impact of adjustment for BSA is not clinically relevant (see graph, above right).
At the extremes of BSA it would be sensible to calculate the absolute GFR (divide eGFR by 1.73 and multiply by the actual BSA, or use the Cockroft-Gault formula).
For drugs with a low therapeutic ratio in renal failure, guidance by therapeutic effect or drug levels may be appropriate. In certain specialised circumstances it may be necessary to measure GFR (for example by isotopic studies).
• The Cockcroft-Gault formula has traditionally been used to guide prescribing, and is usually expressed in mL/min. There are a number of problematic issues around its use.
Unlike the nationally recommended approach to eGFR reporting, Cockcroft-Gault takes no account of the large variation in serum creatinine assays across the country.
These variations can lead to an estimated creatinine clearance ranging from +20% to -15% in clinically important situations. Furthermore, should prescribers use actual weight or ideal weight, and how should ideal weight be estimated?
In most clinical settings weight is unlikely to be measured accurately.
The changes to the BNF (edition 54) go some way to address these issues. I understand further changes are planned for the next edition of the BNF to address individual drugs.
These will build on the recent changes to address our joint goal of enhancing the safety of prescribing for kidney patients in the UK.
From Dr Donal O'Donoghue, national clinical director, Kidney Care