Clinical Validation Of The Eleveld Propofol Pharmacokinetic-pharmacodynamic Model Used In General Anaesthesia

Background: Target-controlled infusion (TCI) systems are used to facilitate anesthetic drug administration. Current systems use pharmacokinetic models developed for specific patient groups. Their use in patients with different characteristics to that of the development population is therefore discouraged. Recently, Eleveld et al1 developed a propofol PK-PD model for broad application. The aim of the current study was to validate it in a broad range of patients, from children to the elderly, and in obese adults. Methods: The medical ethics committee of the University Medical Center Groningen (reference METc2018/216) approved the study. Four groups of 25 patients undergoing an elective surgical procedure were included. Eligibility criteria for the four groups were: children (age ≥ 3 years and <18 years), adults (age 18 - <70 years; body mass index (BMI) <30 kg/m2), elderly patients (age ≥ 70 years) and obese adults (age 18 - <70 years; BMI ≥30 kg/m2). Arterial blood samples were collected at 5, 10, 20, 30, 40 and 60 minutes after start of propofol TCI, and every 30 minutes thereafter until end of surgery or 10 samples total had been collected. Bilateral bispectral index (BIS) was recorded and anesthesiologists were instructed to adjust the target propofol concentration to maintain the BIS in the range 40-60. Predictive performance of the model was assessed using the Varvel criteria2. Results: For PK predictive accuracy, the Eleveld model showed acceptable bias (<±20%) in children, adults and the obese (-4.4%, -14.1%, -14.1% respectively) but there was some bias (-27%) for the elderly. Precision was acceptable (<30%) for all groups. None of the other PK models tested (Schnider, Marsh, Cortinez, Paedfusor and Kataria) performed better than the Eleveld model at a significance level of P<0.01. For BIS, bias and precision was smaller than ±2 and 10 BIS units, respectively, for all groups. The median (5-95 percentile) observed BIS during the procedure was 46 (33-63) and the target concentration was 3.1 (2.2-4.2) µg/ml which was 102% (85-138) of the age-adjusted concentration for 50% drug effect (Ce50). Conclusion: The Eleveld PK-PD model is sufficiently accurate to achieve clinically relevant BIS values using effect-site target concentrations close to the individual Ce50. There is some bias in PK predictions for the elderly, but this does not appear to be a limitation for clinical practic