High vancomycin dosage regimens required by intensive care unit patients cotreated with drugs to improve haemodynamics following cardiac surgical procedures.

The aim of this study was to evaluate retrospectively the importance of a Bayesian pharmacokinetic approach for predicting vancomycin concentrations to individualize its dosing regimen in 18 critically ill patients admitted to intensive care units following cardiothoracic surgery. The possible influence of some coadministered drugs with important haemodynamic effects (dopamine, dobutamine, frusemide) on vancomycin pharmacokinetics was assessed. Vancomycin serum concentrations were measured by fluorescence polarization immunoassay. Vancomycin dosage regimens predicted by the Bayesian method (D(a)) were compared retrospectively with Moellering's nomogram-based dosages (D(M)) to assess possible major differences in vancomycin dosing. D(a) values were similar to D(M) in 10 patients (D(a) approximately D(M) group) (20.52 +/- 8.40 mg/kg/day versus 18.81 +/- 7.24 mg/kg; P = 0.15), whereas much higher dosages were required in the other eight patients (D(a) >> D(M) group) (26.78 +/- 3.01 mg/kg/day versus 18.95 +/- 3.41 mg/kg/day; P < 0.0001) despite no major difference in attained vancomycin steady-state trough concentration (C(min ss)) (9.22 +/- 1. 33 mg/L versus 8.99 +/- 1.26 mg/L; = 0.75) or estimated creatinine clearance (1.23 +/- 0.49 mL/min/kg versus 1.21 +/- 0.24 mL/min/kg; P = 0.95) being found between the two groups. The ratio between D(a) and D(M) was significantly higher in the D(a) >> D(M) group than in the D(a) approximately D(M) group (1.44 +/- 0.18 versus 1.10 +/- 0. 21; P < 0.01). In four D(a) >> D(M) patients the withdrawal of cotreatment with haemodynamically active drugs was followed by a sudden substantial increase in the vancomycin C(min ss) (13.30 +/- 1. 13 mg/L versus 8.79 +/- 0.87 mg/L; P < 0.01), despite no major change in bodyweight or estimated creatinine clearance being observed. We postulate that these drugs with important haemodynamic effects may enhance vancomycin clearance by inducing an improvement in cardiac output and/or renal blood flow, and/or by interacting with the renal anion transport system, and thus by causing an increased glomerular filtration rate and renal tubular secretion. Given the wide simultaneous use of vancomycin and dopamine and/or dobutamine and/or frusemide in patients admitted to intensive care units, clinicians must be aware of possible subtherapeutic serum vancomycin concentrations when these drugs are coadministered. Therefore, therapeutic drug monitoring (TDM) for the pharmacokinetic optimization of vancomycin therapy is strongly recommended in these situations.