Variability of predicted propofol concentrations and measured sevoflurane concentrations during general anaesthesia: a single-centre retrospective cohort study.

BACKGROUND: Variability is high in predicted propofol concentrations during clinical anaesthesia titrated by target-controlled infusion (TCI) to maintain a processed EEG parameter (bispectral index [BIS]) within a specified range. We have shown that the potential for improving the pharmacokinetic model is minimal. The drug titration paradox revealed that titration challenges the classical relationship between drug dose and effect in both individuals and the population. We hypothesised that dynamic factors during surgery beyond the static genetic, epigenetic, and other factors such as age, height, and weight affect the necessary dose. We compared the variability of measured end-tidal sevoflurane concentrations with predicted effect-site propofol concentrations when titrated to a BIS range of 40-60, with the hypothesis that the variability in measured sevoflurane concentrations would not be less than the variability in estimated propofol concentrations. METHODS: Clinical data from 2280 surgical procedures >1 h in duration were included in the analysis. Anaesthesia with sevoflurane or propofol was based on an institutional protocol. The titration performance for both drugs was assessed by comparing BIS values 30 min after skin incision. The variability of the required concentrations at the same time point was calculated and compared. RESULTS: The achieved 30-min post-incision BIS ranges were not significantly different for sevoflurane or propofol TCI (30 [99% CI: 28-33] and 31 [99% CI: 27-36], respectively). The variability of sevoflurane concentrations was not significantly different from measured predicted propofol concentrations during BIS-guided anaesthesia (normalized concentration range of 0.89 [99% CI: 0.78-0.99] and 0.93 [99% CI 0.87-1.02). CONCLUSIONS: Improvements in prediction accuracy of pharmacokinetic models beyond that of those already in clinical use are unlikely to reduce variability in target anaesthetic concentrations across patients in clinical practice.

The drug titration paradox: more drug does not correlate with more effect in individual clinical data.

BACKGROUND: A fundamental concept in pharmacology is that increasing dose increases drug effect. This is the basis of anaesthetic titration: the dose is increased when increased drug effect is desired and decreased when reduced drug effect is desired. In the setting of titration, the correlation of doses and observed drug effects can be negative, for example increasing dose reduces drug effect. We have termed this the drug titration paradox. We hypothesised that this could be explained, at least in part, by intrasubject variability. If the drug titration paradox is simply an artifact of pooling population data, then a mixed-effects analysis that accounts for interindividual variability in drug sensitivity should 'flip' the observed correlation, such that increasing dose increases drug effect. METHODS: We tested whether a mixed-effects analysis could correctly reveal the underlying pharmacology using previously published data obtained during automatic feedback control of mean arterial pressure (MAP) with alfentanil (effect site concentration, CeAlf) during surgery. The relationship between MAP and CeAlf was explored with linear regression and a linear mixed-effects model. RESULTS: A linear mixed-effects model did not identify the correct underlying pharmacology because of the presence of the titration paradox in the individual data. CONCLUSIONS: The relationship between drug dose and drug effect must be determined under carefully controlled experimental conditions. In routine care, where the effect is profoundly influenced by varying clinical conditions and drugs are titrated to achieve the desired effect, it is nearly impossible to draw meaningful conclusions about the relationship between dose and effect.

The Drug Titration Paradox: Correlation of More Drug With Less Effect in Clinical Data.

While analyzing clinical data where an anesthetic was titrated based on an objective measure of drug effect, we observed paradoxically that greater effect was associated with lesser dose. With this study we sought to find a mathematical explanation for this negative correlation between dose and effect, to confirm its existence with additional clinical data, and to explore it further with Monte Carlo simulations. Automatically recorded dosing and effect data from more than 9,000 patients was available for the analysis. The anesthetics propofol and sevoflurane and the catecholamine norepinephrine were titrated to defined effect targets, i.e., the processed electroencephalogram (Bispectral Index, BIS) and the blood pressure. A proportional control titration algorithm was developed for the simulations. We prove by deduction that the average dose-effect relationship during titration to the targeted effect will associate lower doses with greater effects. The finding of negative correlations between propofol and BIS, sevoflurane and BIS, and norepinephrine and mean arterial pressure confirmed the titration paradox. Monte Carlo simulations revealed two additional factors that contribute to the paradox. During stepwise titration toward a target effect, the slope of the dose-effect data for the population will be "reversed," i.e., the correlation between dose and effect will not be positive, but will be negative, and will be "horizontal" when the titration is "perfect." The titration paradox must be considered whenever data from clinical titration (flexible dose) studies are interpreted. Such data should not be used naively for the development of dosing guidelines.

Relationship Between Propofol Target Concentrations, Bispectral Index, and Patient Covariates During Anesthesia.

BACKGROUND: Internationally, propofol is commonly titrated by target-controlled infusion (TCI) to maintain a processed electroencephalographic (EEG) parameter (eg, bispectral index [BIS]) within a specified range. The overall variability in propofol target effect-site concentrations (CeT) necessary to maintain adequate anesthesia in real-world conditions is poorly characterized, as are the patient demographic factors that contribute to this variability. This study explored these issues, hypothesizing that the variability in covariate-adjusted propofol target concentrations during BIS-controlled anesthesia would be substantial and that most of the remaining interpatient variability in drug response would be due to random effects, thus suggesting that the opportunity to improve on the Schnider model with further demographic data is limited. METHODS: With ethics committee approval and a waiver of informed consent, a deidentified, high-resolution, intraoperative database consisting of propofol target concentrations, BIS values, and vital signs from 13,239 patients was mined to identify patients who underwent general endotracheal anesthesia using propofol (titrated to BIS), fentanyl, remifentanil, and rocuronium that lasted at least 1 hour. The propofol target concentrations and BIS values 30 minutes after incision (CeT30 and BIS30) were considered representative of stable intraoperative conditions. The data were plotted and analyzed by descriptive statistics. Confidence intervals were computed using a bootstrap method. A linear model was fit to the data to test for correlation with factors of interest (eg, age and weight). RESULTS: A total of 4584 patients met inclusion criteria and were entered into the analysis. Of the propofol target concentrations, 95% were between 1.5 and 3.5 µg·mL-1. Higher BIS30 values were correlated with higher propofol concentrations. Except for age, all the patient-related variables analyzed entered the regression model linearly. Only 10.2% of the variability in CeT30 was explained by the patient factors of age and weight combined. CONCLUSIONS: Our hypothesis was confirmed. The variability in covariate-adjusted propofol CeT30 titrated to BIS in real-world conditions is considerable, and only a small portion of the remaining variability in drug response is explained by patient demographic factors. This finding may have important implications for the development of new pharmacokinetic (PK) models for propofol TCI.

Disposition of Remifentanil in Obesity: A New Pharmacokinetic Model Incorporating the Influence of Body Mass.

BACKGROUND: The influence of obesity on the pharmacokinetic (PK) behavior of remifentanil is incompletely understood. The aim of the current investigation was to develop a new population PK model for remifentanil that would adequately characterize the influence of body weight (among other covariates, e.g., age) on the disposition of remifentanil in the general adult population. We hypothesized that age and various indices of body mass would be important covariates in the new model. METHODS: Nine previously published data sets containing 4,455 blood concentration measurements from 229 subjects were merged. A new PK model was built using nonlinear mixed-effects modeling. Satisfactory model performance was assessed graphically and numerically; an internal, boot-strapping validation procedure was performed to determine the CIs of the model. RESULTS: Body weight, fat-free body mass, and age (but not body mass index) exhibited significant covariate effects on certain three-compartment model parameters. Visual and numerical assessments of model performance were satisfactory. The bootstrap procedure showed satisfactory CIs on all of the model parameters. CONCLUSIONS: A new model estimated from a large, diverse data set provides the PK foundation for remifentanil dosing calculations in adult obese and elderly patients. It is suitable for use in target-controlled infusion systems and pharmacologic simulation.

The Safety of Target-Controlled Infusions.

Target-controlled infusion (TCI) technology has been available in most countries worldwide for clinical use in anesthesia for approximately 2 decades. This infusion mode uses pharmacokinetic models to calculate infusion rates necessary to reach and maintain the desired drug concentration. TCI is computationally more complex than traditional modes of drug administration. The primary difference between TCI and conventional infusions is that TCI decreases the infusion rate at regular intervals to account for the uptake of drug into saturable compartments. Although the calculated infusion rates are consistent with manually controlled infusion rates, there are concerns that TCI administration of IV anesthetics could introduce unique safety concerns. After approximately 2 decades of clinical use, it is appropriate to assess the safety of TCI. Our aim in this article was to describe safety-relevant issues related to TCI, which should have emerged after its use in millions of patients. We collected information from published medical literature, TCI manufacturers, and publicly available governmental Web sites to find evidence of safety issues with the clinical use of TCI. Although many case reports emphasize that IV anesthesia is technically more demanding than inhaled anesthesia, including human errors associated with setting up IV infusions, no data suggest that a TCI mode of drug delivery introduces unique safety issues other than selecting the wrong pharmacokinetic model. This is analogous to the risk of selecting the wrong drug with current infusion pumps. We found no evidence that TCI is not at least as safe as anesthetic administration using constant rate infusions.

Using the time of maximum effect site concentration to combine pharmacokinetics and pharmacodynamics.

BACKGROUND: To simulate the time course of drug effect, it is sometimes necessary to combine the pharmacodynamic parameters from an integrated pharmacodynamic-pharmacodynamic study (e.g., volumes, clearances, k(e0) [the effect site equilibration rate constant], C(50) [the steady state plasma concentration associated with 50% maximum effect], and the Hill coefficient) with pharmacokinetic parameters from a different study (e.g., a study examining a different age group or sampling over longer periods of time). Pharmacokinetic-pharmacodynamic parameters form an interlocked vector that describes the relationship between input (dose) and output (effect). Unintended consequences may result if individual elements of this vector (e.g., k(e0)) are combined with pharmacokinetic parameters from a different study. The authors propose an alternative methodology to rationally combine the results of separate pharmacokinetic and pharmacodynamic studies, based on t(peak), the time of peak effect after bolus injection. METHODS: The naive approach to combining separate pharmacokinetic and pharmacodynamic studies is to simply take the k(e0) from the pharmacodynamic study and apply it naively to the pharmacokinetic study of interest. In the t(peak) approach, k(e0) is recalculated using the pharmacokinetics of interest to yield the correct time of peak effect. The authors proposed that the t(peak) method would yield better predictions of the time course of drug effect than the naive approach. They tested this hypothesis in three simulations: thiopental, remifentanil, and propofol. RESULTS: In each set of simulations, the t(peak) method better approximated the postulated "true" time course of drug effect than the naive method. CONCLUSIONS: T(peak) is a useful pharmacodynamic parameter and can be used to link separate pharmacokinetic and pharmacodynamic studies. This addresses a common difficulty in clinical pharmacology simulation and control problems, where there is usually a wide choice of pharmacokinetic models but only one or two published pharmacokinetic-pharmacodynamic models. The results will be immediately applicable to target-controlled anesthetic infusion systems, where linkage of separate pharmacokinetic and pharmacodynamic parameters into a single model is inherent in several target-controlled infusion designs.

Efficient trial design for eliciting a pharmacokinetic-pharmacodynamic model-based response surface describing the interaction between two intravenous anesthetic drugs.

BACKGROUND: The authors published a pharmacokinetic- pharmacodynamic model for two drugs based on response surface methodology. Because of the complexity of the model, they performed a simulation study to answer two questions about use of the model: (1) which study design would be most satisfactory; and (2) how many patients would need to be studied to adequately describe an entire response surface. METHODS: Data were simulated using realistic variability for two hypothetical intravenous anesthetic drugs that interact synergistically and that could be given by computer-controlled infusion. Three trial designs were simulated, one that made a series of parallel slices of the response surface, one that crisscrossed the response surface, and one that made a series of radial slices across the surface. Series of 5, 10, 20, and 40 "subjects" were simulated. A pooled data approach was used to assess the ability of the various trial designs and numbers of subjects to adequately identify the interaction response surface and estimate the original response surface. RESULTS: The crisscross design was shown to be the most robust in terms of its ability to both discriminate the correct order of the interaction term and to discriminate the original response surface using the least number of patients. Twenty subjects would be required to adequately define a surface using the crisscross study design, and 40 subjects would be required using the other trial designs. CONCLUSIONS: The results showed that a number of trial designs would be viable, but a design that crossed the surface in a crisscross fashion would give the most robust result with the least patients.