Using estimated creatinine clearance for individualizing drug therapy: a reassessment.

Estimates of renal function are frequently used to design individual dosing regimens. The accuracy of these estimates naturally influences their ability to predict certain pharmacokinetic parameters and appropriate drug dosages. Creatinine clearance is the most widely used estimate of renal function. Many formulas have been developed to provide a quick, relatively accurate prediction of creatinine clearance and, supposedly, the glomerular filtration rate (GFR). However, an understanding of the limitations associated with creatinine clearance estimations raises questions concerning their reliability as an aid in individualizing drug therapy. Many factors such as disease states, age, diet, analytical variations, and drug interactions affect the relationship between estimates and measures of creatinine clearance and GFR. As a result, creatinine clearance estimates using these formulas are often poor reflections of measured creatinine clearance or GFR. Also, studies comparing measured creatinine clearance with more accurate methods of assessing renal function (i.e., inulin) reveal errors that are often exaggerated as renal function declines. Therefore, estimated creatinine clearance is twice removed from the associated pharmacokinetic parameter. Despite these limitations, no other clinically relevant and convenient assessment of renal function is available. The authors recommend that the appropriate caveats be considered when using these tools clinically. For drugs with narrow therapeutic indices, estimates of creatinine clearance should only be used to establish initial dosing regimens, with subsequent therapy based on parameters generated from concentration determinations.