Propofol infusion for induction and maintenance of anaesthesia in patients with end-stage renal disease

Transcription

1 British Journal of Anaesthesia 1998; 81: Propofol infusion for induction and maintenance of anaesthesia in patients with end-stage renal disease B. ICKX, I. D. COCKSHOTT, L. BYTTEBIER, L. DE PAUW, A. VANDESTEENE AND A.A. D HOLLANDER Summary We gave investigated the pharmacokinetics and pharmacodynamics of propofol in 11 patients with end-stage renal disease (ESRD) compared with nine healthy patients during and after a manually controlled three-stage infusion of propofol 21, 12 and 6 mg kg 1 h 1 lasting a minimum of 2 h. Mean total body clearance was not reduced significantly in the ESRD group (30.66 (SD 8.47 ml kg 1 min 1 ) compared with the control group (33.75 (7.8) ml kg 1 h 1 ). ESRD patients exhibited a greater, but not statistically significant, volume of distribution at steady state compared with patients in the control group (11.25 (8.86) vs 5.79 (2.14) litre kg 1, respectively). Elimination half-life values were unchanged by renal failure. Mean times to induction of anaesthesia were similar in both groups: 177 (SD 57) and 167 (58) s for the ESRD and control groups, respectively. Waking time after cessation of propofol infusion was significantly shorter in the ESRD group (474 (156) s) compared with the control group (714 (240) s) (P0.05). Mean plasma concentrations on waking were similar. We conclude that the pharmacokinetic and pharmacodynamic profiles of propofol after infusion were not markedly affected by renal failure. (Br. J. Anaesth. 1998; 81: ). Keywords: anaesthetics i.v., propofol; anaesthetic techniques, i.v. infusion; kidney, failure; pharmacokinetics, propofol Selection of a safe and effective drug regimen for patients with renal impairment can be difficult. Data obtained from studies in subjects with normal renal function cannot with confidence be used to design a drug administration regimen for patients with endstage renal disease (ESRD). Renal failure may influence hepatic drug metabolism either by inducing or inhibiting hepatic enzymes and also by altering protein binding or protein denaturation. 1 2 During disposition may be affected by changes in body fluid distribution and circulatory volume in these dialysed patients. This can induce changes in hepatic blood flow which in turn can alter production and elimination of metabolites. 1 Propofol is used widely as an i.v. anaesthetic agent for both induction and maintenance of general anaesthesia. In healthy patients, propofol has a relatively large volume of distribution and a high total body clearance. 3 5 The drug is extensively biotransformed in the liver and its metabolites appear to have no pharmacological activity. 6 Since the first report of Morcos and Payne on the disposition of a bolus dose of propofol in four patients with chronic renal failure, 7 Kirvelä and colleagues have concluded that propofol may be used safely for induction of general anaesthesia in patients with renal impairment. 8 Limited information has been obtained on the use of a continuous infusion of propofol in patients with renal impairment for maintenance of general anaesthesia. 9 We have investigated the safety of propofol infusion for induction and maintenance of anaesthesia in uraemic patients, and whether or not its pharmacokinetics and pharmacodynamics were similar in a group of healthy patients anaesthetized in a similar way. Patients and methods The study was approved by the Institutional Ethics Committee of the University Hospital and all patients gave informed consent. We enrolled 11 patients, ASA III or IV, with creatinine clearance rates of less than 10 ml min 1, presenting for arteriovenous fistulae surgery (ESRD group) and 10 patients, ASA I, with normal renal function (group C) undergoing surgery of an expected duration of at least 120 min. These patients did not suffer from cardiac or hepatic dysfunction or obesity (body weight not exceeding 20% of ideal). Patients characteristics and laboratory data are summarized in table 1. Concurrent therapy was received only by patients in the ESRD group (table 2). No patient was receiving cytochrome P450 inducers or inhibitors. All patients were premedicated with diazepam 10 mg orally, 1 h before anaesthesia, except for one patient in the ESRD group who received bromazepam. Electrocardiogram, arterial pressure (by sphygmomanometry measured on a lower limb in the ESRD group) and oxygen saturation (by pulse oximetry) were monitored continuously. Before B. ICKX, MD, L. BARVAIS, MD, A. VANDESTEENE, MD, PHD, A. A. D HOLLANDER, MD, PHD, (Department of Anaesthesiology); L. DE PAUW, MD (Department of Nephrology); CUB Erasme, 808 route de Lennik, 1070 Bruxelles, Belgium. I. D. COCKSHOTT, PHD, Zeneca Pharmaceuticals, Mereside Alderley Park, Macclesfield, Cheshire, SK10 4TG. G. BYTTEBIER, MSC, Zeneca Pharmaceuticals, Biometrics Unit, Medical Department, Schaessestraat, 15, B-9070 Destelbergen, Belgium. Accepted for publication: July 27, Correspondence to B. I.

2 Propofol in patients with end-stage renal disease 855 Table 1 Patient characteristics and preoperative laboratory values (mean (SD)). NRNormal range Table 2 Antihypertensive medication received by patients in the ESRD group Patient No. Control ESRD P Age (yr) ns Height (cm) (10) 164 (5.7) ns Weight (kg) 72.6 (10) 62.3 (7.9) 0.01 Sex (M/F) 4/5 5/6 ns Haemoglobin (g litre 1 ) (NR ) 14.2 (1.6) 9.9 (2) 0.01 Creatinine (mg dl 1 ) (NR ) 0.86 (0.12) 17.6 (25.8) 0.01 Albumin (g dl 1 ) (NR 3.5 5) 4.38 (0.61) 3.71 (0.53) 0.05 AST (iu litre 1 ) (NR 5 35) 16 (4) 14 (7) ns ALT (iu litre 1 ) (NR 5 35) 16 (5) 15 (10) ns Creatinine clearance (ml min 1 ) (NR ) (20) 7.3 (3.3) 0.01 Medication 1 Metoprolol Atenolol 5 6 Amiodarone 7 Nifedipine, Metoprolol 8 Nifedipine, Metoprolol 9 10 Metoprolol 11 Metoprolol inductin of anaesthesia, a venous catheter was inserted in the contralateral arm to that undergoing surgery, for infusion of 5% dextrose at a rate of 3 ml kg 1 h 1 and for drug administration until the end of anaesthesia. A catheter was also inserted into the radial artery of the same arm for blood sampling and continuous monitoring of arterial pressure. The lungs were preoxygenated with 100% oxygen and after administration of fentanyl 0.1 mg i.v., anaesthesia was induced and maintained using a three-step infusion regimen of propofol, delivered by a calibrated syringe pump (Ohmeda): 21 mg kg 1 h 1 for 5 min, followed by 12 mg kg 1 h 1 for 10 min and 6 mg kg 1 h 1 until the end of the procedure. The final rate was not altered throughout the study and no supplementary doses were given. A bolus dose of vecuronium 0.1 mg kg 1 was given at loss of consciousness to facilitate tracheal intubation. Ventilation was maintained with 60% nitrous oxide in oxygen and end-tidal carbon dioxide concentration was kept constant at kpa by adjusting ventilation. Bolus doses of fentanyl 0.05 mg were administered when heart rate or mean arterial pressure increased to more than 30% of preoperative values. If arterial pressure decreased by more than 20% of mean pressure, ephedrine 5 mg i.v. was administered. Dextran , 6% up to a maximum of 200 ml, was given to ESRD patients at the surgeon s request. The propofol infusion was maintained for at least 120 min according to the length of the procedure, and duration of infusion was noted. At the end of the surgical procedure, neuromuscular block was antagonized when four twitches of the train-of-four response were present at the adductor pollicis muscle. Nitrous oxide was stopped simultaneously with the propofol infusion. CLINICAL ASSESSMENT Time to induction of anaesthesia was defined as the time from the beginning of infusion of propofol to the absence of response on verbal stimulation. Recovery was assessed by measurement of the time from the end of infusion to when patients opened their eyes and answered standard questions correctly. The total amount of fentanyl was noted as was the time since the last dose to cessation of the infusion and amount of ephedrine and additional fluids given. PHARMACOKINETICS Arterial blood samples (2 ml) for measurement of propofol concentrations in whole blood were obtained before induction of anaesthesia and at 1, 2, 3, 4, 5, 6, 7, 9, 11, 13, 15, 17, 19, 21, 25, 30, 40, 50, 60, 75, 105 and 120 min after initiation of infusion. For infusions continuing beyond 120 min, additional samples were collected every 15 min. After cessation of infusion, further samples were collected at 2, 4, 6, 8, 10, 20, 40, 60, 90, 120, 240, 300, 360, 420, 480 and 720 min. Additional samples were obtained when patients lost consciousness and when they opened their eyes to verbal command. Samples were cooled immediately and stored at 4C, and assays were carried out within 12 weeks. Whole blood concentrations of propofol were measured by high pressure liquid chromatography (HPLC) with fluorescence detection. 10 The assay had an inter-batch coefficient of variation of 3 12% with a limit of detection of g ml 1. As a quality control, 20% of all sample assays were replicated in a separate batch; differences between assay results were less than 10%. Bi- and triexponential functions of the following form, derived from Colburn, 11 were fitted to the data using the least squares curve fitting program MKMODEL (Holford N. MKMODEL Version 3.36 User Manual, Biosoft, Cambridge, UK, 1988): C = 3.5 * t * * λit1 λi ( t 5) Ci *(1 e )*e i = 1 λi* T2 λi*( t 15) Ci i = 1 λi* T3 λi*( t T) Ci i = * * (1 e ) *e * * (1 e ) * e where C tpropofol concentration at any time t after the start of the first infusion; 1, 2, and 3model fitted hybrid rate constants for three exponential phases; these are referred to as, and, respectively, in the tabulation and reporting of results. TT 1T 2T 3; T 1, T 2 and T 3times from the start of each infusion to the maximum times for that infusion; ttotal time from the start of the first infusion. When (t5), (t15) or (tt) is not greater than 0, than that function is set to zero. C 1, C 2 and C 3model-fitted coefficients defining the contribution to the steady state concentration of each exponential phase; 3.5, 2.0 and 1.0ratios of the three infusion rates to the terminal infusion rate (6 mg kg 1 h 1 ). In all non-linear

3 856 British Journal of Anaesthesia least squares analysis it is necessary to weight the dependent variable (concentration) data in such a way that the weighted errors for this variable are constant over the full range of the independent variable; these errors include contributions from both the assay and inter-subject variability. 12 Balant and Garrett justify this by showing that the use of equal weights can virtually ignore terminal phase data, if the data range over two or more logarithmic cycles. 12 In our study a weighting of 1/C 2 was used for all data sets. This was confirmed as optimal in a subset of patients; residuals for observed data points from the line of best fit indicated a better fit when using a weighting factor of 1/C 2 compared with 1/C. The following parameters were derived from the fitted model: ss Steady state propofol concentration ( C ) = Ci i=1 Total body clearance Cl(m)terminal infusion rate/c s Volume of central = Ci compartment( V ) 3 total dose/ 1 i=1 where C i corr.fitted coefficients corrected for the infusion duration using the method described by Loo and Riegelman. 13 Apparent volume of Clm/ distribution at equilibrium ( V γ = ) γ Model-independent parameter estimates were calculated to validate the parameters derived from the fitted model and were as follows: Total body clearance ( Cl(mi)=total dose/auc where AUC area under the propofol concentration curve extrapolated to infinity, calculated using the trapezoidal approximation: Apparent volume of AUMC total dose distribution at = 2 ss steady state ( V ) (AUC ) total dose εinfusion duration 2 ε AUC ( ) where AUMC area under the first moment of the propofol concentration curve extrapolated to infinity, calculated using the trapezoidal approximation. Mean residence time MRT = AUMC / AUC STATISTICAL ANALYSIS The number of patients required to detect a predefined difference in a pharmacokinetic parameter between groups was estimated from data in a previous study. 7 Total body clearance was considered to be the most important variable on which to base calculations. We considered it clinically important to detect a 30% reduction in control mean. With 10 patients per group and a significance level of 5%, the power of the study was estimated as 75% using a standard statistical method. 14 All data are expressed as mean (SD). The data were analysed by the Wilcoxon rank sum test for betwen-group comparisons of pharmacokinetic data and measured data. P values were calculated using PC-SAS version 6.04 corr computer program (SAS Institute Inc., Cary, NC, USA, 1990). P0.05 was considered significant. Inter-subject variability difference between groups was tested for the pharmacokinetic data using the F test of Snedecor (SAS/STAT User s Guide, Release 6.03 Edition. Cary, NC: SAS Institute Inc., 1988). Mean arterial pressure and heart rate from induction to 60 min after infusion of propofol between the two groups were compared by analysis of variance (ANOVA) for repeated measurements using the multivariate approach with the pre-induction value in the contrast statement. Results For the control group, surgery consisted of lumbar hernia (four patients), varicose vein surgery (two patients), mastectomy (one patient), rhinoplasty (two patients) and laparoscopy (one patient). Anaesthesia was uneventiful in all patients except for one in the control group. In this patient, anaesthesia was continuously too light as a result of technical problems with the infusion pump and all data from this patient were excluded from analysis. Mean body weight in the ESRD group was significantly smaller than that in the control group but there was no significant difference between groups for age or height (table 1). Patients in the ESRD group had been haemodialysed within the previous 24 h. Discomfort at the injection site was reported by one patient in each group. Mean duration of infusion was 176 (SD 60) min (range min) for the ESRD group and 152 (48) min ( min) for the control group (ns). Patients lost consciousness within 177 (57) s and 167 (58) s after the beginning of infusion in the ESRD and control groups, respectively (ns). The corresponding mean propofol concentration in the ESRD group (4.41 (1.93) g ml 1 ) was higher than that in the control group (3.71 (1.14) g m 1 ). This difference was not statistically significant. Patients with impaired renal function opened their eyes significantly earlier than control patients (474 (156) s compared with 714 (240) s, respectively) (P0.05). Blood propofol concentrations on awakening were not significantly different (0.89 (0.43) and 0.78 (0.33) g ml 1 in the ESRD and control groups, respectively). Mean time at which patients were able to answer questions correctly was significantly shorter in the ESRD group (978 (258) s) than in the control group (1266 (198) s) (P0.05). The total amount of fentanyl administered was larger in the control group (0.317 (0.178) mg) than in the ESRD group (0.215 (0.085) mg). However, this difference was not significant. Time from the last fentanyl dose before termination of the propofol infusion (59 (26) vs 64 (21) min in the control and ESRD groups, respectively) was not significantly different between groups. No major haemodynamic instability was detected in either group. A mean volume of 223 (69) ml of Dextran , 6% was infused in nine patients in the ESRD group. Two patients in both groups received ephedrine (up to a total of 20 mg for each patient in the ESRD group). In the control group, one patient received ephedrine 5 mg and one 10 mg. Heart rate did not differ between groups except at 10 and 60 min after induction of anaesthesia. There was no difference between groups in mean arterial pressure.

4 Propofol in patients with end-stage renal disease 857 Figure 1 Mean (SD) blood concentrations of propofol (logarithmic scale) during infusion of propofol in the end-stage renal failure (ESRD) and control groups. In the right corner of the graph (inset), the first 15 min after the beginning of the infusion are magnified. PHARMACOKINETICS Changes in mean blood concentration of propofol with time, during and after infusion, followed a similar pattern in both groups (figs 1, 2). After an initial peak concentration of 5 g ml 1 at 5 min after the start of infusion, the profile reflected the changes in infusion rate; changes in mean concentration from 12 min onwards were small (fig. 1). After termination of the infusion, propofol concentrations declined curvilinearly with time (fig. 2). Two- and three-compartment models were fitted to the full propofol concentration profile for each patient. These were compared by reference to the Schwarz criterion and by inspection of the residuals of the observed data from the line of best fit. 15 For three patients in each group, the preferred model was two compartment as three phases could not be defined with confidence. Individual values for the pharmacokinetic parameters are shown in table 3 for both model-dependent and model-independent analyses. To eliminate bias cause by the patient s weight, the appropriate pharmacokinetic data were analysed with adjustment for body weight. Estimates of propofol clearance were similar for both analyses. There was no significant difference between ESRD and control patients for any of the pharmacokinetic parameters. The F test of Snedecor showed no inter-subject variability difference between the groups tested for clearance but there was a significant difference in variability for the volumes of distribution (table 3). Discussion Clearance of propofol is predominantly metabolic as little or no unchanged propofol is detected in excreta 16 ; this emphasizes the importance of the liver in the clearance of propofol In patients with renal failure, hepatic drug metabolism may be increased, decreased or remain unchanged. 1 2 Our study has confirmed that uraemia requiring haemodialysis does not affect significantly the pharmacokinetics of propofol. Mean clearance in the renal failure patients (30.6 ml min 1 kg 1 ) was similar to that of healthy control patients (33.8 ml min 1 kg 1 ), whether derived by model or model-independent analysis. Moreover, mean clearance in the ESRD group was similar to that reported in other healthy patient groups receiving propofol as an infusion ( ml min 1 kg 1 ) This indicates that the infusion dose requirements of propofol in ESRD patients are similar to those in patients with normal renal function. Extrahepatic mechanisms for clearance of propofol have been thought to explain values in excess of the total blood supply to the liver. The fact that there was no significant reduction in total body clearance in the ESRD group in this study suggests that clearance by the kidney, if present in humans, is not contributing significantly to the clearance of propofol. It is interesting to note that the sheep kidney has been observed to convert a metabolite of propofol (probably propofol glucuronide) into propofol 22 ; data from our study suggest that this is unlikely to occur in humans. The medications received by the ESRD group did not appear to influence the pharmacokinetics of propofol. The volume of the central compartment, V 1, was similar in the two groups (0.13 (0.067) vs 0.12 (0.039) litre kg 1 for the ESRD and control groups, respectively) but was at the lower end of the range reported previously ( litre kg 1 ) This probably resulted from a combination of the high frequency of sampling during the early period of the infusion 23 and the use of arterial blood. Arterial blood sampling was chosen in this study because it is considered to be the most appropriate when performing pharmacokinetic modelling and pharmacodynamic assessments. Plasma expansion fluids were given in the ESRD group but not in the control group. However, the

5 858 British Journal of Anaesthesia Figure 2 Mean (SD) blood concentrations of propofol (logarithmic scale) after infusion of propofol in the end-stage renal failure (ESRD) and control groups. In the right corner of the graph (inset), the first 20 min after infusion cessation are magnified. amounts given were small in these chronically haemodialysed patients with restricted fluid intake and it is considered that the difference would be unlikely to affect the volume of distribution substantially, as intravascular volume is only a small part of the volume of distribution of propofol. The mean value for V ss in our control patients (5.8 litre kg 1 ) was somewhat higher than that reported in other studies of propofol infusions with a similar sampling duration ( litre kg ). The mean value for V ss in the ESRD group (11.3 litre kg 1 ) was higher than that in the control group Table 3 Individual pharmacokinetic data from the model and from model-independent analysis T1/2, T1/2, T1/2 initial, intermediate and terminal half-lives, respectively; C ss propofol concentration at steady state; V 1 kg 1 volume of distribution of the central compartment per kg body weight; V kg 1 volume of distribution at equilibrium per kg body weight; Cl(m) kg 1 and Cl(mi) kg 1 total body clearance per kg body weight derived, respectively, from model-dependent and model-independent analysis; V ss kg 1 volume of distribution at steady state per kg body weight; MRTmean resistance time. P values for the Wilcoxon rank sum test for between-group comparison and for the F test of Snedecor C ss Patient No. T1/2 (min) T1/2 (min) T1/2 (min) (g ml 1 ) V 1 kg 1 (litre kg 1 ) V kg 1 (litre kg 1 ) Cl(m) kg 1 Cl(mi) kg 1 (ml kg 1 min 1 ) V ss kg 1 (ml kg 1 min 1 ) (litre kg 1 ) MRT (min) Control Median Mean SD ESRD Median Mean SD P ns ns ns ns ns ns ns ns ns F test ns ns ns ns 0.02 ns ns ns

6 Propofol in patients with end-stage renal disease 859 (5.8 litre kg 1 ). This difference was not statistically significant, possibly as a consequence of the significantly larger inter-subject variability in the ESRD group (F test of Snedecor). V, V ss and the terminal half-life values for propofol in our study were smaller than those observed previously in patients with chronic renal failure. 8 9 This may be because of the longer sampling period after administration of propofol (24 h) 28 or the sampling site used. 25 Chiou has pointed out that profiles of concentration in blood or plasma from different sampling sites can lead to calculation of widely differing pharmacokinetic parameters and that venous concentration data may grossly misrepresent the true distributive and terminal phases of a drug in the body. 25 The total sampling duration was clearly inadequate to define T 1/2 with confidence in all subjects; it has been suggested that estimated values of this half-life tend to increase with sampling duration 28 and this is consistent with the very long T 1/2 value reported by Albanese and colleagues (1878 min) when sampling for 72 h after infusion. 29 However, no trend in T 1/2 values was observed over a range of infusion sampling time ( min) by Bailie and colleagues. 30 In view of the difficulties of defining T 1/2 with confidence and the difference between subjects in both groups in the number of exponential phases required to define the elimination kinetics, MRT values have also been compared between groups. There was no significant difference between groups for this parameter. The propofol concentration associated with loss of consciousness was comparable with values reported in unpremedicated patients ( g ml 1 ) ) and was therefore unexpectedly high in our group of premedicated patients. Recovery times in the ESRD patients were significantly shorter than those in the control patients. The amount of fentanyl given during the infusion was larger in the control group. This could have contributed to the difference in recovery times between groups. But this is unlikely as propofol concentrations at eyes open were similar in both groups and were in good agreement with published data (approximately 1 g ml 1 ) Vuyk and colleagues demonstrated that return of consciousness is caused mainly by a reduction in the effect site propofol concentration and less by fentanyl decay. 33 Interestingly, Nathan and colleagues also noted a trend to a shorter emergence time in their uraemic group compared with control patients. 9 There was no major haemodynamic instability justifying interruption of the infusion. This may have been because of the slow infusion rate used for induction. Peacock and colleagues demonstrated that in elderly patients, the slower the infusion rate the smaller were the decreases in arterial pressure. 34 Our infusion rate corresponded to a dose of 1.75 mg kg 1 over a 5-min period for induction of anaesthesia, which is lower than the standard induction dose used in clinical practice. Nevertheless, during anaesthesia, treatment with ephedrine to restore haemodynamic stability was more frequent in the ESRD group (total amount given 4 (8.4) mg) than in the control group (1.7 (3.5) mg) (ns). We believe that this was related to the volume status of these haemodialysed patients. In summary, the pharmacokinetics of propofol infusion were similar in ESRD and control subjects. In particular, propofol clearance, which determines the steady state concentration during infusion, was unchanged in subjects with end-stage renal failure. This suggest that the kidney is not involved substantially in the extrahepatic metabolism of propofol. The manually controlled three-stage infusion to supplement nitrous oxide resulted in a satisfactory state of surgical anaesthesia in almost all patients, whether or not renal function was impaired. In both ESRD patients and patients with normal renal function, it is important that careful dose titration is used when administrating propofol by infusion. References 1. Elston AC, Bayliss MK, Park GR. Effect of renal failure on drug metabolism by the liver. British Journal of Anaesthesia 1993; 71: Reidenberg MM. The biotransformation of drugs in renal failure. American Journal of Medecine 1977; 62: Smith I, White P, Nathanson M, Gouldson R. PROPOFOL. An update on its clinical use. Anesthesiology 1994; 81: Bryson HM, Fulton BR, Faulds D. PROPOFOL. An update of its use in anaesthesia and conscious sedation. Drugs 1995; 50: Gepts E, Trenchant A. Propofol. Anaesthetic Pharmacology Review 1995; 3: Simons PJ, Cockshott ID, Glen JB, Gordon EA, Knott S, Ruane RJ. Disposition and pharmacology of propofol glucuronide administered intravenously to animals. Xenobiotica 1992; 22: Morcos WE, Payne JP. The induction of anaesthesia with propofol ( Diprivan ) compared in normal and renal failure patients. Postgraduate Medical Journal 1985; 61: (Suppl. 3): Kirvelä M, Olkkola KT, Rosenberg PH, Yli-Hankala A, Salmela K, Lindgren L. Pharmacokinetics of propofol and haemodynamic changes during induction of anaesthesia in uraemic patients. British Journal of Anaesthesia 1992; 68: Nathan N, Debord J, Narcisse F, Dupuis JL, Lagarde M, Benevent D, Lachatre G, Feiss P. Pharmacokinetics of propofol and its conjugates after continuous infusion in normal and in renal failure patients: a preliminary study. Acta Anaesthesiologica Belgica 1993; 44: Plummer GF. Improved method for the determination of propofol in blood by high-performance liquid chromatography with fluorescence detection. Journal of Chromatography 1987; 421: Colburn WA. Simultaneous pharmacokinetic and pharmacodynamic modeling. Journal of Pharmacokinetics and Biopharmaceutics 1981; 9: Balant LP, Garrett ER. Computer use in pharmacokinetics. In: Garrett ER, Hirtz JL, eds. Drug Fate and Metabolism, vol. 4. New York: Loo JCK, Riegelman S. Assessment of pharmacokinetic constants from post-infusion blood curves obtained after intravenous infusion. Journal of Pharmacological Sciences 1970; 59: Machin D, Campbell MJ. Statistical Tables for the Design of Clinical Trials. Medical Statistics and Computing. Oxford: Blackwell Scientific Publications, 1987; 3: Schwarz G. Estimating the dimension of a model. Annals of Statistics 1978; 6: Simons PJ, Cockshott ID, Douglas EJ, Gordon EA, Hopkins K, Rowland M. Disposition in male volunteers of a subanaesthetic intravenous dose of an oil in water emulsion of 14 C- propofol. Xenobiotica 1988; 18: 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 1990; 7: Gepts E, Jonckheer K, Maes V, Sonck W, Camu F. Disposition kinetics of propofol (Diprivan) during alfentanil anaesthesia. Anaesthesia 1988; 43: (Suppl.): Shafer A, Doze VA, Shafer SL, White PF. Pharmacokinetics and pharmacodynamics of propofol infusions during general anesthesia. Anesthesiology 1988; 69: Cockshott ID, Briggs LP, Douglas EJ, White M.

7 860 British Journal of Anaesthesia Pharmacokinetics of propofol in female patients. British Journal of Anaesthesia 1987; 59: Veroli P, O Kelly B, Bertrand F, Trouvin JH, Farinotti R, Ecoffey C. Extrahepatic metabolism of propofol in man during the anhepatic phase of orthotopic liver transplanation. British Journal of Anaesthesia 1992; 68: Mather LE, Selby DG, Runciman WB, McLean CF. Propofol: assay and regional mass balance in the sheep. Xenobiotica 1989; 19: Verpoorte R. Potential effect of early blood sampling schedule on calculated pharmacokinetic parameters of drugs after intravenous administration. Journal of Pharmaceutical Sciences 1980; 69, 7: Coetzee JF, Glen JB, Wium CA, Boshoff L. Pharmacokinetic model selection for target controlled infusions of propofol. Anesthesiology 1995; 82: Chiou WL. The phenomenon and rationale of marked dependence of drug concentration on blood sampling site: implications in pharmacokinetics, pharmacodynamics, toxicology and therapeutics. Clinical Pharmacokinetics 1989; 17: Servin F, Cockshott ID, Farinotti R, Haberer JP, Winckler C, Desmonts JM. Pharmacokinetics of propofol infusions in patients with cirrhosis. British Journal of Anaesthesia 1990; 65: Servin F, Farinotti R, Haberer JP, Desmonts JM. Propofol infusion for maintenance of anesthesia in morbidly obese patients receiving nitrous oxide. Anesthesiology 1993; 78: Kirkpatrick T, Cockshott ID, Douglas EJ, Nimmo WS. Pharmacokinetics of propofol (Diprivan) in elderly patients. British Journal of Anaesthesia 1988; 60: Albanese J, Martin C, Lacarelle B, Saux P, Durand A, Gouin F. Pharmacokinetics of long-term propofol infusion used for sedation in ICU patients. Anesthesiology 1990; 73: Bailie GR, Cockshott ID, Douglas EJ, Bowles BJM. Pharmacokinetics of propofol during and after long term continuous infusion for maintenance of sedation in ICU patients. British Journal of Anaesthesia 1992; 68: Vuyk J, Engbers FHM, Lemmens HJM, Burm AGL, Vletter AA, Gladines MPR, Bovill JG. Pharmacodynamics of propofol in female patients. Anesthesiology 1992; 77: Leslie K, Sessler DI, Smith WD, Larson MD, Ozaki M, Blanchard D, Crankshaw DP. Prediction of movement during propofol/nitrous oxide anesthesia. Anesthesiology 1996; 84: Vuyk J, Martijn JM, Olofsen E, Burm AGL, Bovill JG. Propofol anesthesia and rationale opioid selection. Anesthesiology 1997; 87: Peacock JE, Lewis RP, Reilly CS, Nimmo WS. Effect of different rates of infusion of propofol for induction of anaesthesia in elderly patients. British Journal of Anaesthesia 1990; 65: