Propofol administered by a manual infusion regimen

Transcription

1 British Journal of Anaesthesia 995; 74: CLINICAL INVESTIGATIONS Propofol administered by a manual infusion regimen J. W. SEAR AND J. B. GLEN Summary We have evaluated the clinical utility and blood propofol concentrations produced with two different infusion regimens for propofol, given to supplement 67% nitrous oxide-morphine anaesthesia. Patients received a standardized three-step infusion of propofol based either on body weight (weight-corrected ) or on a mean body weight of 70 kg (standard dose ). Both s showed similar cardiovascular stability and recovery times. In the 48 patients studied, isoflurane was used as a supplement in nine (two in the weight-corrected ). Apparent steady state blood propofol concentrations were 3.4 (SD 0.69) ig mh in the weight-corrected and 3.46 (0.79) ug ml" in the standard dose. These results suggest that for patients weighing kg body weight, a standard dose infusion regimen may be a suitable starting point. In routine clinical practice, the need for isoflurane supplementation may be avoided by subsequent titration of the infusion rate according to clinical response. Computer simulation of the actual infusion rates used in each patient has allowed retrospective comparison of the predictive performance of different pharmacokinetic descriptors for propofol. The variables described by Tackley and colleagues provided a more accurate prediction of the measured blood propofol concentration than did the variable set reported by Gepts and colleagues. (Br. J. Anaesth. 995; 74: ) Key words Anaesthetics i.v., propofol. Anaesthetic techniques, i.v. infusion. There is considerable variability in the reported data for the disposition of propofol by infusion as a supplement to nitrous oxide for maintenance of anaesthesia. Clearance estimates vary between mean values of.3 and 2.09 litre min" [-3]. As a result, the plasma or blood propofol concentration produced by a given infusion regimen might be expected to vary over a two- or three-fold range. In clinical practice many infusion regimens are based on body weight. Shafer and colleagues reported a significant positive correlation between systemic clearance and body weight []. Taken in conjunction with the variability in clearance, this may result in even greater differences between predicted and measured blood propofol concentrations. Spelina and colleagues determined the EC 95 blood propofol concentration (that concentration needed to suppress movement to the initial incision in 95 % of the population during nitrous oxide-propofol anaesthesia) as 3.39 ug ml" [4]. Roberts and co-workers have previously described a manually controlled three-stage infusion regimen aimed at achieving a predicted concentration of 3 ug ml" [5]. However, to date, the accuracy of this regimen has not been assessed nor has the influence of body weight on the blood propofol concentration produced. The present study was therefore designed to compare two dosing strategies for propofol, one based on body weight and the other based on an assumed "mean" body weight of 70 kg. Assessments were made to examine clinical adequacy, haemodynamic effects, recovery times and the variability in measured blood propofol concentrations produced with each technique. The information on the actual infusion rates used in each patient and the measured blood propofol concentrations allowed retrospective evaluation of the accuracy of different pharmacokinetic data sets in predicting the concentrations achieved and of the influence of body weight and propofol dose on predictive accuracy. This assessment of predictive performance may be useful in selecting an optimum set of pharmacokinetic input variables for incorporation in computer-controlled infusion devices which aim to achieve a targeted blood propofol concentration. Patients and methods After obtaining informed consent and Ethics Committee approval, we studied 48 patients (4 female, ASA I or II, aged 6-60 yr, weights kg) undergoing abdominal surgery. After premedication with temazepam 20 mg orally h before surgery, anaesthesia was induced with propofol, the trachea J. W. SEAR, MA, BSC, PHD, FRCA, Nuffield Department of Anaesthetics, University of Oxford, John RadclifFe Hospital, Headington, Oxford. J. B. GLEN, BVMS, PHD, DVA, Medical Affairs Department, Zeneca Pharmaceuticals, Macdesfield, Cheshire. Accepted for publication: October 25, 994. Correspondence to J. W. S. Part of this paper was presented at a workshop meeting " Focus on infusion" and has been published in abstract form as "Should propofoi infusion rate be related to body weight?" In: Prys- Roberts C, ed. Focus on Infusion: Intravenous Anaesthesia. London: Current Medical Literature Ltd, 994; 00-0.

2 Propofol administered by manual infusion regimen 363 intubated after administration of vecuronium 0. mg kg", and anaesthesia maintained with an infusion of propofol and nitrous oxide. Additional analgesia was provided by a single bolus dose of morphine 0 mg i.v. before surgery. The electrocardiogram, heart rate and arterial pressure were monitored continuously. At the end of surgery, the infusion of propofol and nitrous oxide were discontinued and residual neuromuscular block antagonized with neostigmine and glycopyrronium. Patients were stratified according to body weight into one of four bands: kg, kg, kg and kg and recruitment continued until equal numbers were obtained in each band. After stratification, patients were allocated randomly to receive either an infusion of propofol using a regimen based on body weight or a standard infusion regimen based on an assumed body weight of 70 kg (standard regimen). WEIGHT-CORRECTED GROUPS Propofol mg kg" was administered for induction of anaesthesia followed by an infusion of 0 mg kg" h" for 0 min, 8 mg kg" h" for the next 0 min and 6 mg kg" h" thereafter. This regimen of Roberts and colleagues [5] was designed on the basis of the kinetic variables reported by Tackley and colleagues [6]. STANDARD REGIMEN Propofol 70 mg was given for induction of anaesthesia; 700 mg h" for the first 0 min, 560 mg h~ ] for the next 0 min and 420 mg h" thereafter. Venous blood samples were obtained for measurement of propofol concentrations at incision; at 35,40 and 45 min after the start of the propofol infusion; at the end of the infusion; and at recovery (eyes open to command). Additional samples were obtained at times of clinically inadequate anaesthesia (systolic arterial pressure > 5% above awake value; heart rate > 90 beat min" ; movement to the initial incision; autonomic responses sweating, tears, reacting pupils), and at such times, the propofol infusion was supplemented by % isoflurane. Inadequate neuromuscular block (as assessed by hiccuping or evidence of unsatisfactory surgical conditions) was treated by increments of vecuronium -2 mg. However, no increments were given before the surgical incision to allow us to assess somatic response to inadequate anaesthesia. Whole blood concentrations of propofol were measured by HPLC according to Plummer [7]. The inter-assay coefficient of variation for paired samples was % over the concentration range -6 ug ml" and the limit of detection was 0 ng ml". STATISTICS Data are shown as mean (SD) except where otherwise stated. Comparison of haemodynamic data against preinduction values was made using the Wilcoxon Table Pharmacokinetic variables for propofol for a 70 kg patient determined by Tackley and colleagues [6] and Gepts and colleagues [8]. K, = Initial volume of distribution; Cl = systemic clearance; k l0, k n, k 2i, k n and & 3 = microconstants describing a three-compartment mammillary model Variable K, (litre) * 0 (min" ) k i2 (min" ) k 2l (min" ) * 3 (min" ) k it (min" ) Cl (litre min ) Tackley [6] Gepts [8] matched pairs signed ranks test, and between s by the Mann-Whitney U test; the incidences of side effects between s by the chi-square test with Yates' correction, and Fisher exact tests; blood propofol concentrations and recovery data by the Mann-Whitney U test; and performance error (PE) and absolute performance error (APE) by Student's unpaired r-test. DATA ANALYSIS Computer simulation of the propofol dose regimen used in each patient (in mg kg" and mg kg h" ) provided information on the predicted blood propofol concentrations at those times when actual concentrations were measured. Simulations were conducted with pharmacokinetic input variables described by Tackley and colleagues [6] and a separate set advocated by Gepts and colleagues [8]. As these two publications describe several different subsets of rate constants, the actual values used in the present study are shown in table. The accuracy with which the infusion regimen achieved the predicted drug concentration, derived by computer simulation, was assessed by calculation of performance errors. For a given estimate of blood concentration, the performance error (PE) was calculated as: PE = (C M -C P )/C P xl00% where C M = measured blood concentration of propofol and C P = predicted concentration according to the kinetic model used [9, 0]. The individual errors of all patient samples were then summated and the mean (SD) PE and APE calculated. Patient bias (systematic error above or below the predicted concentration (%)) was denned by PE and inaccuracy (size of the typical miss) by APE. When the 95 % confidence intervals of the population mean PE included zero, it was concluded that no significant bias existed. Results Anaesthesia and surgery were uneventful in all patients. The duration of anaesthesia was min and total doses of propofol (excluding the initial bolus dose) were mg. Both s showed similar cardiovascular stability (table 2). However, the increases in systolic and

3 364 British Journal of Anaesthesia Table 2 Mean (SD) haemodynamic variables (systolic and diastolic arterial pressures, and heart rate) during anaesthesia. *P < 0.05, **P < 0.0 compared with preinduction values; \P < 0.0, f-j-p < 0.0 compared with preincision values Systolic pressure (mm Hg) Preinduction At incision 5 min postincision 35 min infusion 45 min infusion End of infusion Diastolic pressure (mm Preinduction At incision 5 min postincision 35 min infusion 45 min infusion End of infusion Heart rate (beat min" ) Preinduction At incision 5 min postincision 35 min infusion 45 min infusion End of infusion Weight-corrected Table 3 Perioperativeadverse effects in the two s n Pain on injection Bradycardia/absent beats (intraoperative) Postop. bradycardia Hypotension (SAP < 80 mm Hg) (intraoperative) 30. (4.3) 99.4 (2.9)** 07.4 (9.9)**ft 98.5 (3.7)** 99.9 (4.0)** 03. (2.8)** Hg) 76. (7.5) 59.(7.4)** 68.9 (7.5)**f 59.3 (8.3)** 59.3 (9.2)** 62.5(8.)** 77.6(4.) 68.3 (0.7) 63.0 (0.0)**f 6.6(8.3)** 60.2 (8.3)** 60.3 (8.0)** Weight-corrected Standard dose 28.3 (2.5) 00.9(8.)** 0.0(20.)** 05.4 (6.4)** 08.2 (20.4)** 0.7(20.3)** 77.4 (8.4) 6.6(.)** 68.7(5.)* 65.5 (3.4)* 66.8(4.)* 69.0 (4.0) 77.3 (7.4) 69.7 (2.0) 67.8 (3.0) 65.0 (2.7)* 64.2 (0.8)** 65.4 (0.9) Standard dose diastolic arterial pressures, and heart rate to surgical incision were statistically significant in the weightcorrected (P < 0.0) but not in the standard dose. In the latter there were no haemodynamic differences between data recorded for the patients in the lightest band (weights kg) and those in the heaviest band (weights kg) (table 2). The incidence of perioperative adverse events was low and there was no difference between the two s (table 3). CONCENTRATION-EFFECT RELATIONSHIPS Five patients in the standard dose (two in kg weight band and one each in the other bands) responded to the initial surgical incision either by somatic movement (n = ) or by increases in arterial pressure or heart rate. These responses were treated with an increment of vecuronium or isoflurane, respectively. Blood propofol concentrations in these patients varied between 2.32 and 4.4 ig ml". Propofol concentrations in nonresponders varied between 2.87 and 6.50 ug ml" in the weight-corrected and between 2.75 and 5.2 ug ml" in the standard dose (ns) During surgery six patients had clinical signs of inadequate anaesthesia (one in the weight-corrected at a propofol concentration of 2.60 ug ml" and five in the standard dose, two in the kg band and one in each of the other three weight bands). Propofol concentrations associated with inadequate anaesthesia in the standard dose varied between.6 and 3.85 ug ml" (median 3.0 ug ml" ). Nine patients in the weight-corrected and eight in the standard dose received intraoperative increments of vecuronium. Propofol concentrations associated with clinically adequate anaesthesia during surgery were ug ml" in the weight-corrected and ug ml" in the standard dose (ns). The apparent steady state propofol concentrations (calculated as mean concentration values between 35 and 45 min) were 3.42 (SD 0.69) ug ml" in the weight-corrected s and 3.46 (0.79) ug ml" in the standard dose (ns). However, when the latter was divided into the four weight bands, the mean value in the kg was significantly greater than the means in the other three weight bands (P < 0.05) and the mean in the weight-corrected (P < 0.025) (fig. ). In the standard dose there was a weak negative correlation between the apparent steady state concentration and body weight (r 2 = 0.227, P < 0.05). Recovery from anaesthesia was prompt in both s. The median times from discontinuing nitrous oxide to eyes opening to command were 0 min in the weight-corrected infusion s (range 2-22 min) and 7 (4-5) min in the standard dose (ns). Blood propofol concentrations at recovery were similar for the two treatment regimens (weight-corrected :.7 (SD 0.38) (range ) ug ml" (K = 23); standard dose :.3 (0.39) ( ) ug ml" (n = 9)). After operation no patient had any recall of anaesthesia or surgery. PREDICTIVE ACCURACY OF KINETIC MODELS The 48 patients provided 230 samples of intraoperative data. Predicted propofol concentrations (based on Tackley and colleagues [6]; table ) were compared with measured values. Individual patient values of the median PE ranged from 34.8% to 76.3 %. The mean (SD) of all sample PE values was 22. (32.0) % for the weight-corrected infusion and 7.4 (3.3) % for the control (ns). The mean values for individual APE were 29. (25.7)% and 28.9(2.0)%, respectively (ns). For both s, the two results were significantly different from zero (P < 0.0). The infusion profile used in the weight-corrected provided a single set of predicted propofol concentrations against which measured values were compared. In the standard dose s the predicted values were calculated based on the actual dose given to each patient, and this differed depending on the body weight of each patient. Thus in lighter patients, a relatively greater dose was given in terms of mg kg", and predicted propofol concentrations were

4 Propofol administered by manual infusion regimen i- E ^ 3 - o c o I 2 o Q. O < > W-C * i : i i S-D s Figure Apparent steady state blood propofol concentrations for the weight-corrected (W-C) dosing (n = 25) and the standard dose (S-D), shown individually for the four weight bands (n = 6 per, except for kg, where n = 5). *P < 0.05 compared with the standard dose, kg weight band. Table 4 Population means (SD) of individual sample PE and APE estimates in the two treatment s (based on the kinetic variables described by Tackley and colleagues [6]). W-C = Weight-corrected ; S-D = standard dose, n = number of blood propofol concentrations per. Weight ranges indicate weight bands for standard dose subs Group («) Mean SD P (between s) Prediction errors of populations W-C (20) S-D (0) kg (27) vs kg (< 0.02) kg (29) vs kg (< 0.00); kg (< 0.02) kg (28) kg (26) Absolute prediction errors of populations W-C S-D kg vs kg (< 0.0) kg vs kg (< 0.05); kg (< 0.00); kg (< 0.00) kg kg greater than those obtained in heavier patients. The standard dose s thus provide an opportunity to examine the influence of propofol dose on PE; the mean dose per kilogram body weight was lowest in the kg and greatest in the kg. The mean (SD) PE and APE values of individual samples are shown in table 4. Within the four weight bands of the standard dose, the smallest PE and APE values were in the kg body weight. There were also significant differences between the weight-corrected infusion regimen and some weight bands of the standard dosing (for PE: vs kg (P < 0.05) and vs kg (P< 0.00); for APE: vs kg (P < 0.02) and kg (P < 0.00)). The magnitude of the median PE varied with the duration of the infusion. During the first 30 min of infusion, the median value of the individual sample PE was 24.6% (95% CI %); 6.3% 365 ( %) during min of infusion; and 0.7 % (- 0.2 to 6.3 %) for samples taken after more than 60 min of infusion (the median values in the first two time periods were significantly different from zero (P<0.0)). To assess if the measured values could be predicted more accurately with a different pharmacokinetic model, predicted values were also calculated using the kinetic variable estimates of Gepts and colleagues [8] (table ). The corresponding population mean PE value for all 48 patients was 38.0 (SD 28.8) % and the mean APE 40.8 (27.2) %. These values were significantly different (P < 0.0) from the corresponding values for all 48 patients when obtained using the kinetic variable estimates of Tackley and co-workers [6] (2.6 (24.) % and 25. (20.2) %, respectively). Discussion A principal objective of the study was to compare the variability in blood propofol concentrations obtained with two manually controlled infusion regimens and therefore the propofol infusion was not titrated to the requirements of each individual patient. Despite these restrictions, adequate clinical anaesthesia was achieved in 39 of 48 patients and isoflurane supplementation was used in the remainder. The cardiovascular responses to surgical incision and surgery were minimal, and were similar to those seen in patients (of comparable age and ASA status) receiving either volatile or opioid supplementation to nitrous oxide anaesthesia, or other hypnotic-nitrous oxide techniques for maintenance of anaesthesia [, 2]. The infusion regimen described by Roberts and colleagues [5] was calculated as a manual approximation to a complex infusion scheme based on Vaughan and Tucker [3], with a target concentration of between 2.5 and 3.0ugml"'. After the initial bolus dose and two fast infusion phases, we measured individual blood propofol concentrations during the maintenance infusion phase (6 mg kg" h" ), which ranged between.63 and 5.55 ug ml" when the infusion rate was administered on a per kilogram body weight basis. This represents a threefold variability and reflects both kinetic variability and the effects of abdominal surgery on liver blood flow and hence propofol clearance [4, 5]. Another possible interacting factor might have been the addition of isoflurane in those patients exhibiting inadequate anaesthesia. However, most studies suggest that isoflurane has little effect on liver blood flow at concentrations less than MAC [6]. Although most i.v. drug regimens have been described on the basis of dose per kilogram body weight, there are few data showing that weight is an important factor in dosing to achieve a desired drug concentration. Thus in this study we also compared propofol concentrations achieved when given on the basis of weight-adjusted dose or on the basis of the mean weight of the population range (i.e. 70 kg). In this latter, with a fixed dosing regimen, mean blood propofol concentrations at min varied between.32 and 5.28 ug ml". They were

5 366 British Journal of Anaesthesia significantly greater in the kg weight band compared with the heavier weight s (Fig. ), although apparent steady state propofol concentrations in the two s as a whole were similar. Thus in most patients dosing could be simplified to a standard regimen of an initial 7-ml bolus dose, followed by 70 ml h" and 56 ml h", respectively, for 0 min each and a maintenance rate of 42 ml h". This recommendation may appear in apparent disagreement with the preliminary results of Raftery and co-workers [7] who demonstrated that dose requirements for propofol TIVA should be calculated as mgkg~ - 75 h~'. However, their study was not supported by blood propofol concentration data. In a more recent abstract, Schuttler and Ihmsen [8] examined data from 85 patients and volunteers, and using population kinetic analysis showed a linear effect of weight on clearance. However, they also found a 53 % inter-individual variability in clearance in their large sample. In our study, the correlation coefficient (r 2 = 0.227) indicates that clearance was influenced by weight, but there are also other more important factors which account for 77% of the association. There were no significant differences in blood propofol concentrations between 35 and 45 min in the weight-corrected and the three s (60-89 kg) receiving the standard dose regimen. However, when assessed in terms of bias and accuracy, there were significant between- differences. The lowest mean PE and APE values were in the kg weight, and there was a significant difference in APE between the kg and the other three weight bands of the standard dose regimen. Thus we would suggest that a standardized regimen may be a useful starting strategy for i.v. infusion anaesthesia with propofol, as we have demonstrated both kinetic and dynamic variability with regard to dosing requirements. As has been shown by the earlier studies of Sear and colleagues [9] and Shafer and colleagues [], there is no absolute propofol concentration associated with adequate anaesthesia for both surgical incision and intra-abdominal surgery, and hence the infusion rate should always be titrated according to clinical response. Several other studies have examined the predictability of infusion regimens in terms of measured and predicted blood or plasma propofol concentrations. For example, Schuttler and co-workers [20] found that the measured to predicted ratio for propofol was 0.88 when comparing 508 samples obtained from 20 patients receiving alfentaniloxygen-enriched air-propofol anaesthesia. For a predicted propofol concentration of 2.5 ug ml", measured concentrations varied between and 3.5 ug ml"". Marsh and co-workers have described the use of a computer-driven infusion system for children [2] based on the infusion kinetics of Gepts and colleagues [8]. This produced a median PE (bias) of 8.5% (caused partly by the larger volume of distribution of propofol in the paediatric patient [22, 23]) and a median APE of 25.4 %. Using a computer-driven infusion pump based on the pharmacokinetic model of Gepts and colleagues [8], Glass and colleagues found a mean PE of 5 % and a median APE of 29% for propofol given to adult patients [24]. Our own data give values of population means for the individual PE and APE values of 2.6% and 25.%, respectively. Smaller estimates of both PE might be achieved by examining fewer patients with more samples per patient the observed magnitude of the present estimates reflecting closely interindividual kinetic variability (±30%) in surgical patients. The effect of duration of infusion on the median PE (i.e. the manual infusion regimens had a worse performance during the first 45 min of infusion) can be attributed in part to several definable factors, such as the poor delineation of the initial volume of distribution and the distribution phase rate constant, a, by Kay and colleagues [25]. This is the result of venous sampling and inadequate early sampling. Early infusion drug concentrations are dependent on the accuracy of the initial volume of distribution while later drug concentrations are dependent more on systemic drug clearance. For other drugs, Raemer and co-workers found that the median APE for alfentanil was between 52 and 55%; for fentanyl [26], Glass and colleagues showed that the median PE was 4 % and median APE 2 % when using a computer-assisted infusion system [27], while Shafer and co-workers reported median APE values of 3-59% for fentanyl [28]. Recently, Buhrer and colleagues have reported population median PE and APE estimates of 5 % and 6%, respectively for thiopentone by infusion to volunteers [29], and Theil and colleagues median PE and APE estimates of 4 % and 28 %, respectively, for fentanyl, and 9 % and 25 % for midazolam during computer-assisted continuous infusions for cardiac surgery [30]. The pharmacokinetic variables of Gepts and colleagues [8] did not predict the measured values more accurately than those of Tackley and colleagues [6]. This may be related to the use of arterial samples for the derivation of the variables in the former study, while venous sampling was used by Tackley and colleagues and in the present study. To select an optimum set of pharmacokinetic variables for computer-controlled administration of propofol, further studies are required which assess short-term accuracy in response to changing infusion rates of propofol. Acknowledgement We are grateful to Mr E. J. Douglas, Safety of Medicines Department, Zeneca Pharmaceuticals, for conducting the propofol analyses. References. Shafer A, Doze VA, Shafer SL, White PF. Pharmacokinetics and pharmacodynamics of propofol infusions during general anesthesia. Anesthesiology 988; 69: Cockshott ID, Douglas EJ, Prys-Roberts C, Turtle M, Coates DP. The pharmacokinetics of propofol during and after intravenous infusion in man. European Journal of Anaesthesiology 990; 7:

6 Propofol administered by manual infusion regimen Morgan DJ, Campbell GA, Crankshaw DP. Pharmacokinetics of propofol when given by intravenous infusion. British Journal of Clinical Pharmacology 990; 30: Spelina KR, Coates DP, Monk CR, Prys-Roberts C, Norley I, Turtle MJ. Dose requirements of propofol by infusion during nitrous oxide anaesthesia in man. I: Patients premedicated with morphine. British Journal of Anaesthesia 986; 58: Roberts FL, Dixon J, Lewis GTR, Tackley RM, Prys- Roberts C. Induction and maintenance of propofol anaesthesia. A manual infusion scheme. Anaesthesia 988; 43 (Suppl.): Tackley RM, Lewis GTR, Prys-Roberts C, Boaden RW, Dixon J, Harvey JT. Computer controlled infusion of propofol. British Journal of Anaesthesia 989; 62: Plummer GF. An improved method for the determination of propofol (ICI 35868) in blood. Journal of Chromatography : Biomedical Application 987; 42: Gepts E, Camu F, Cockshott ID, Douglas EJ. Disposition of propofol administered as constant rate intravenous infusions in humans. Anesthesia and Analgesia 987; 66: Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. Journal of Pharmacokinetics and Biopharmaceutics 98; 9: Varvel JR, Donoho DL, Shafer SL. Measuring the predictive performance of computer-controlled infusion pumps. Journal of Pharmacokinetics and Biopharmaceutics 992; 20: Prys-Roberts C. Cardiovascular effects of continuous intravenous anaesthesia compared with those of inhalational anaesthesia. Ada Anaesthesiologica Scandinavica 982; 26: (Suppl. 75): Sear JW. Intravenous anaesthetics. In: Foex P, ed. Anaesthesia for the Compromised Heart. Bailliere's Clinical Anaesthesiology 989; 3: Vaughan DP, Tucker GT. General theory for rapidly establishing steady-state drug concentrations using two consecutive constant rate infusions. European Journal of Clinical Pharmacology 975; 9: Gelman SI. Disturbances of hepatic blood flow during anesthesia and surgery. Archives of Surgery 976; : Cowan RE, Jackson BT, Grainger SL, Thompson RPH. Effects of anesthetic agents and abdominal surgery on liver blood flow. Hepatology 99; 4: Gelman S, Fowler KC, Smith LR. Liver circulation and function during isoflurane and halothane anesthesia. Anesthesiology 984; 6: Raftery S, Browne B, Enever G, Richards MJ, Prys-Roberts C. Total i.v. anaesthesia with propofol and alfentanil: influence of age and weight on dose requirements. British Journal of Anaesthesia 99; 67: 28P. 8. Schuttler J, Ihmsen H. Population pharmacokinetics of propofol. Anesthesiology 993; 79: A Sear JW, Shaw A, Wolf A, Kay NH. Infusions of propofol to supplement nitrous oxide-oxygen for the maintenance of anaesthesia. A comparison with halothane. Anaesthesia 988; 43: (Suppl.): Schuttler J, Kloos S, Schwilden H, Stoeckel H. Total intravenous anaesthesia with propofol and alfentanil by computer-assisted infusion. Anaesthesia 988; 43 (Suppl. 2): Marsh B, White M, Morton N, Kenny GNC. Pharmacokinetic model driven infusion of propofol in children. British Journal of Anaesthesia 99; 67: Saint-Maurice C, Cockshott ID, Douglas EJ, Richard MO, Harmey JL. Pharmacokinetics of propofol in young children after a single dose. British Journal of Anaesthesia 989; 63: Vandermeersch E, Van Hemelrijck J, Byttebier G, Van Aken H. Pharmacokinetics of propofol during continuous infusion for paediatric anaesthesia. Acta Anaesthesiologica Belgica 989; 40: Glass PS, Goodman DK, Ginsberg B, Reves JG, Jacobs JR. Accuracy of pharmacokinetic model-driven infusion of propofol. Anesthesiology 989; 7: A Kay NH, Sear JW, Uppington J, Cockshott ID, Douglas EJ. Disposition of propofol in patients undergoing surgery: a comparison in men and women. British Journal of Anaesthesia 986; 58: 423-^ Raemer DB, Buschman A, Varvel JR, Philip BK, Johnson MD, Stein DA, Shafer SL. The prospective use of population pharmacokinetics in a computer-driven infusion system for alfentanil. Anesthesiology 990; 73: Glass PSA, Jacobs JR, Smith LR, Ginsberg B, Quill TJ, Bai SA, Reves JG. Pharmacokinetic model-driven infusion of fentanyl: assessment of accuracy. Anesthesiology 990; 73: Shafer SL, Varvel JR, Aziz N, Scott JC. Pharmacokinetics of fentanyl administered by computer-controlled infusion pump. Anesthesiology 990; 73: Buhrer M, Maitre PO, Hung OR, Ebling WF, Shafer SL, Stanski DR. Thiopental pharmacodynamics. I. Defining the pseudo-steady-state concentration-eeg effect relationship. Anesthesiology 992; 77: Theil DR, Stanley TE, White WD, Goodman DK, Glass PSA, Bai SA, Jacobs JR, Reves JG. Midazolam and fentanyl continuous infusion anesthesia for cardiac surgery; a comparison for computer-assisted versus manual infusion systems. Journal of Cardiothoracic and Vascular Anesthesia 993;7: