CREATININE-CLEARANCE ESTIMATES FOR PREDICTING GENTAMICIN PHARMACOKINETIC VALUES IN OBESE PATIENTS

The Cockcroft-Gault and Salazar-Corcoran equations were compared with respect to prediction of gentamicin pharmacokinetic values in obese and nonobese patients, and the results were used to formulate guidelines for calculating initial gentamicin dosages in obese patients. Creatinine clearance (CL(cr)) was estimated by applying the Cockcroft-Gault equation using total body weight (TBW), ideal body weight (IBW), and dosage weight (DW) and with Salazar-Corcoran equations using fat-free body mass (FBM) in 100 obese and 100 nonobese patients. Gentamicin pharmacokinetic values (k, CL, and t(1/2)) were estimated by using CL(cr) estimated by each method and standardized to a body surface area of 1.73 sq m. Actual pharmacokinetic values were determined by using steady-state gentamicin concentrations and a modified Sawchuk-Zaske equation; these values were compared with the predicted values. In the obese patients, pharmacokinetic values predicted from standardized CL(cr) by the Cockcroft-Gault equation using estimated DW were not significantly biased, compared with actual values; most predictions produced by the other methods were significantly biased. Predictions produced by the DW method were generally more precise than those resulting from the other methods. In nonobese patients, k values estimated by the Cockcroft-Gault equation using IBW were not significantly biased, while values obtained with all other methods were biased. All methods were biased when predicting CL and t(1/2) in nonobese patients. Significant correlations existed between standardized estimates of CL(cr) (by all methods) and pharmacokinetic values in both groups. Predictions of gentamicin k, CL, and t(1/2) were best overall when CL(cr) was estimated by the Cockcroft-Gault equation using DW, rather than by other methods. Use of DW is recommended for calculating initial gentamicin regimens in obese patients, but dosage adjustments should be based on serum drug concentrations.