Predictive Performance of Vancomycin Trough Concentrations in Patients With Diabetes With Microalbuminuria.

Background: The performance of a population pharmacokinetic model in predicting trough concentrations after the initial vancomycin dose was evaluated in patients with albuminuria compared with patients who did not have albuminuria.

Methods: Data were collected from 52 patients infected with methicillin-resistant Staphylococcus aureus (excluding patients undergoing dialysis and acute kidney injury) and treated with vancomycin. The data included urinary albumin concentration.

Results: The calculated mean prediction error and mean absolute error for the serum trough concentrations of vancomycin (with 95% confidence intervals) were 4.65 (4.13-5.17) and 6.1 (5.65-6.51), respectively, in microalbuminuria and 0.33 (-0.2 to 0.86) and 4.02 (3.59-4.45), respectively, in those without. There was no significant difference observed in serum creatinine concentration, age, weight, estimation of vancomycin trough concentration in serum, and actual trough concentration of vancomycin in serum between individuals with microalbuminuria and those without albuminuria.

Conclusions: Microalbuminuria in patients with diabetes is a marker of the difference between predicted vancomycin trough concentrations and actual vancomycin trough concentrations.