Target-controlled infusion (TCI) is based on pharmacokinetic models designed to achieve a desired drug level in the blood. TCI's predictive accuracy of plasma propofol levels at the end of surgery with major blood loss has not been well established. This prospective observational study included adult patients (BMI 20-35 kg/m2) undergoing surgery with expected blood loss ≥ 1500 mL. The study was conducted with the Schnider TCI propofol model (Alaris PK Infusion Pump, CareFusion, Switzerland). Propofol levels were assessed in steady-state at the end of anaesthesia induction (Tinitial) and before the end of surgery (Tfinal). Predicted propofol levels (CTCI) were compared to measured levels (Cblood). Twenty-one patients were included. The median estimated blood loss was 1600 mL (IQR 1000-2300), and the median fluid balance at Tfinal was + 3200 mL (IQR 2320-4715). Heart rate, mean arterial blood pressure, and blood lactate did not differ significantly between Tinitial and Tfinal. The median bispectral index (0-100) was 50 (IQR 42-54) and 49 (IQR 42-56) at the two respective time points. At Tinitial, median CTCI was 2.2 µmol/L (IQR 2-2.45) and Cblood was 2.0 µmol/L (bias 0.3 µmol/L, limits of agreement - 1.1 to 1.3, p = 0.33). CTCI and Cblood at Tfinal were 2.0 µmol/L (IQR 1.6-2.2) and 1 µmol/L (IQR 0.8-1.4), respectively (bias 0.6 µmol/L, limits of agreement - 0.89 to 1.4, p < 0.0001). Propofol TCI allows clinically unproblematic conduct of general anaesthesia. In cases of major blood loss, the probability of propofol TCI overestimating plasma levels increases.Trial registration German Clinical Trials Register (DRKS; DRKS00009312).
Keywords: Automated drug delivery devices; Propofol administration; Target controlled infusion.