Sou Hyun Lee

Aims This study aimed to evaluate the predictive performance of previously constructed cefazolin pharmacokinetic models and determine whether cefazolin administration via the target‐controlled infusion (TCI) method may be possible in clinical practice. Methods Twenty‐five gastrectomy patients receiving cefazolin as a prophylactic antibiotic were enrolled. Two grams of cefazolin was dissolved in 50 mL of normal saline to give a concentration of 40 mg mL −1 . Before skin incision, cefazolin was administered using a TCI syringe pump, and its administration continued until the end of surgery. The target total plasma concentration was set to 100 μg mL −1 . Total and unbound plasma concentrations of cefazolin were measured in three arterial blood samples collected at 30, 60 and 120 min after the start of cefazolin administration. The predictive performance of the TCI system was evaluated using four measures: inaccuracy, divergence, bias and wobble. Results Total (n = 75) and unbound (n = 75) plasma concentration measurements from 25 patients were included in the analysis. The pooled median (95% confidence interval) biases and inaccuracies were 6.3 (4.0‐8.5) and 10.5 (8.6‐12.4) for the total concentration model and −10.3 (−16.8 to −3.7) and 22.4 (18.2‐26.7) for the unbound concentration model, respectively. All unbound concentrations were above 10 μg mL −1 . Conclusion Administration of cefazolin by the TCI method showed a clinically acceptable performance. Applying the TCI method by setting the total concentration as the target concentration rather than the unbound concentration is effective in maintaining a constant target concentration of cefazolin.