Remifentanil pharmacodynamics during conscious sedation using algometry: a more clinically relevant pharmacodynamical model.

BACKGROUND: The Minto remifentanil pharmacokinetic/pharmacodynamic (PK/PD) model is used in target-controlled infusion (TCI) devices. The endpoint used to calculate the PD parameters, including the ke0 , was the electroencephalogram (EEG), which only changes at high remifentanil concentrations. As the ke0 should adequately predict the time course of drug effects at clinically relevant concentrations, we evaluated the temporal agreement between effect-site concentrations estimated with the Minto model and pressure pain thresholds during conscious sedation.

METHODS: We enrolled 100 patients scheduled for gynaecological surgery. The first group (35 subjects) received an effect site targeted remifentanil infusion (target 1.5 ng ml-1 ); the second group (35 subjects) received the same infusion and a 1 mg bolus of midazolam to evaluate anxiolytic effects; the control group (30 subjects) received a saline infusion. Algometry and vital signs were measured at different time points.

RESULTS: The Minto model predicted stable effect-site concentrations within 1.5 min of starting the infusion. Haemodynamic variables stabilised within 5 min, whereas there was a significant increase in pressure pain threshold for up to 15 min in both remifentanil groups. Midazolam had no effect on pressure pain threshold. A PD model based on algometry and Minto PK model was developed.

CONCLUSIONS: Our results demonstrate the limitation of the Minto PD model at low target remifentanil concentrations, with a discrepancy in the time course between EEG and pressure pain threshold changes. Clinicians should be aware that the time course of onset of analgesic effects is slower than the estimates of the Minto model. Investigators should consider using algometry data in future opioid PD modelling studies.