Population pharmacokinetics of high-dose thiopental in patients with cerebral injuries.

Thiopental monitoring was performed in 95 critically ill patients hospitalized for neurologic damage, High-dose thiopental was infused during long-term treatment. Total dose of 333 +/- 144 mg/kg (449 +/- 185 mg/kg for females and 302 +/- 113 mg/kg for men) were given in 125 +/- 43 hours. Plasma concentration-time data were analyzed according to a population pharmacokinetic approach with an initial group of 65 patients. Clearance (CL) and central volume of distribution (Vc) were modeled alone and under the influence of demographic covariates, assuming a two-compartment open model with first-order elimination. The final population models were as follows: CL (L/hr) = 11.7.weight (kg).age (yr)/(2136 + age2) and Vc = 1.52.weight (kg) + 44.8. Mean CL and Vc mean population estimates were 8.01 L/hr (133 ml/min or 2.02 ml/min/kg) and 145 L (2.19 L/kg). The predictive performance of the population modeling and parameters was evaluated with a bayesian fitting procedure in an independent validation set of 30 patients with similar physical and clinical characteristics. There was no statistically significant bias or imprecision between measured and predicted thiopental plasma concentrations in this validation group. Moreover, there was a good adequation (r = 0.939) between individual CL values predicted from the population formula and estimated with the bayesian approach.