- Email: [email protected]
Manual compared with target-controlled infusion of propofol
British Journal of Anaesthesia 1995; 75: 562–566
Manual compared with target-controlled infusion of propofol D. RUSSELL, M. P. WILKES, S. C. HUNTER, J. B. GLEN, P. HUTTON AND G. N. C. KENNY Summary We studied 160 ASA l–ll patients, anaesthetized with propofol by infusion, using either a manually controlled or target-controlled infusion system. Patients were anaesthetized by eight consultant anaesthetists who had little or no previous experience of the use of propofol by infusion. In addition to propofol, patients received temazepam premedication, a single dose of fentanyl and 67 % nitrous oxide in oxygen. Each consultant anaesthetized 10 patients in sequential fashion with each system. Use of the target-controlled infusion resulted in more rapid induction of anaesthesia and allowed earlier insertion of a laryngeal mask airway. There was a tendency towards less movement in response to the initial surgical stimulus and significantly less movement during the remainder of surgery. Significantly more propofol was administered during both induction and maintenance of anaesthesia with the target-controlled system. This was associated with significantly increased end-tidal carbon dioxide measurements during the middle period of maintenance only, but recovery from anaesthesia was not significantly prolonged in the target-controlled group. With the exception of a clinically insignificant difference in heart rate, haemodynamic variables were similar in the two groups. Six of the eight anaesthetists found the target-controlled system easier to use, and seven would use the target-controlled system in preference to a manually controlled infusion. Anaesthetists without prior experience of propofol infusion anaesthesia quickly became familiar with both manual and target-controlled techniques, and expressed a clear preference for the target-controlled system. (Br. J. Anaesth. 1995; 75: 562–566) Key words Anaesthetics i.v., propofol. Anaesthetic techniques, i.v. infusion. Equipment, computers. Equipment, infusion systems.
Propofol is an i.v. anaesthetic agent suitable for induction and maintenance of general anaesthesia. When used in this fashion, the aim is to infuse propofol at an appropriate rate, with supplementary boluses as required, to achieve an effective blood propofol concentration appropriate to the degree of surgical stimulation. It is important to avoid excessive administration which may be associated with adverse cardiorespiratory effects and delayed recovery. Stokes and Hutton [1] studied 80 patients premedicated with temazepam undergoing body surface
surgery. They compared induction with propofol administered at four different infusion rates from an Ohmeda 9000 syringe pump [2] until verbal contact with the patient was lost, and recommended that to achieve induction, propofol should be infused at a rate of 600 ml h91. They then maintained anaesthesia using a manually controlled infusion of propofol set initially at 6 mg kg91 h91 and thereafter adjusted the infusion rate such that movement in response to surgical stimulation was eliminated. Fentanyl 1.5 g kg91 was administered 5 min before induction, and patients breathed spontaneously from a Mapleson A system delivering 67 % nitrous oxide in oxygen. In a subsequent study, Gill, Lewis and Reilly [3] infused propofol at 600 ml h91 to achieve induction, and then compared maintenance at either a fixed rate of 6 mg kg91 h91 or a “stepdown” regimen of 10 mg kg91 h91 for 10 min, 8 mg kg91 h91 for 10 min and 6 mg kg91 h91 thereafter. Patients in both groups required additional incremental doses of propofol 20 mg in order to maintain adequate anaesthesia, but both methods were considered satisfactory. White and Kenny [4, 5] developed a portable computerized infusion system which allows the anaesthetist to select any desired target blood concentration of propofol at any time. The system was based on the series II Psion handheld computer interfaced with an Ohmeda 9000 syringe driver via an RS232 linkage but has since been superseded by a backbar integrated with the syringe driver. The software for the administration of propofol (Diprifusor; Zeneca Ltd) consists of a pharmacokinetic model, a set of specific pharmacokinetic variables for propofol and infusion control algorithms. To achieve induction, the patient’s weight and age are entered, and then a target blood propofol concentration is selected, depending on age, ASA status and concomitant opioid administration. Thereafter, the control unit commands the infusion pump to deliver a rapid zero order infusion at a rate of 1200 ml h91 until the pharmacokinetic model in the target-controlled system predicts that the desired target concentration has been achieved. It then D. RUSSELL, FRCA, University Department of Anaesthesia, Royal Infirmary, Glasgow G31 2ER. M. P. WILKES, FRCA, P. HUTTON, BSC, PHD, FRCA, University Department of Anaesthesia, Queen Elizabeth Hospital, Birmingham B15 2TH. S. C. HUNTER, J. B. GLEN, BVMS, PHD, DVA, Clinical Research Group, Zeneca Pharmaceuticals, Macclesfield, Cheshire SK10 4TG. G. N. C. KENNY, BSC (HONS), FRCA, MD, Glasgow University Department of Anaesthesia, HCI Hospital, Beardmore Street, Clydebank, Glasgow G81 4HX. Accepted for publication: June 2, 1995. Correspondence to G.N.C.K.
Manual vs target-controlled propofol infusion provides a variable rate infusion as required to maintain that target concentration. The anaesthetist has at all times the option of selecting a higher or lower target concentration, with the rate of delivery of propofol by the syringe pump being controlled automatically. If a higher target concentration is requested, this results in administration of another bolus followed by infusion at an increased rate, while a request for a lower target concentration results in temporary discontinuation of infusion followed by resumption at a lower rate. The target blood propofol concentration is therefore titrated upwards and downwards in response to clinical signs, thus increasing or decreasing the blood and effector site concentrations, in the same way that the inspired concentration of traditional inhalation agents such as halothane may be increased or decreased to achieve alterations in end-tidal and brain partial pressures. When introduced into clinical practice [6], the target-controlled infusion system proved to be popular, mainly because it was easy to use and provided a high level of confidence regarding the predictability of anaesthetic effects. Furthermore, in a study of patients receiving anaesthesia for cardiac surgery, Alvis and colleagues [7] demonstrated improved cardiovascular stability from the use of a computer-assisted fentanyl infusion compared with a manual technique. The aims of the present study were to compare the quality of anaesthesia achieved by manual infusion of propofol as described by Stokes and Hutton [1] with that achieved by the target-controlled infusion system and to compare the ease of use, pattern of propofol administration and time to recovery from anaesthesia. We also sought to examine the rate at which anaesthetists became familiar and confident with one or other technique.
Patients and methods Eight consultant anaesthetists from two centres (four from each), unfamiliar with the administration of propofol by infusion, attended a workshop where they received instruction and practical demonstration of the manual and target-controlled systems. Comprehensive written instructions in the rationale of each technique and practical aspects of the systems were provided. With the target-controlled system it was suggested that the initial target concentration should be between 4 and 8 g ml91 depending on the degree of hypnosis present immediately before induction, with titration thereafter according to response. For a 70-kg patient, the selection of a target propofol concentration of 6 g ml91 would result in the administration of approximately 110 mg over 33 s, followed by a slowly decreasing infusion of 156 ml h91. With the manual system the infusion rate was 600 ml h91 until loss of consciousness (loss of verbal contact), followed immediately by an initial maintenance infusion rate of 6 mg kg91 h91 (42 ml h91 for a 70-kg patient) and a variable infusion rate thereafter of up to 15 mg kg91 h91. Additional propofol was infused at 600 ml h91 as judged appropriate to achieve induction, or if it was felt that anaesthetic depth was inadequate. To allow exam-
563 ination of the rate at which anaesthetists became familiar with these techniques, each consultant anaesthetized 10 patients in sequential fashion with each system. Thus individual patients were not allocated randomly to one or other system, but the order in which each anaesthetist used both systems was randomized. After obtaining informed written consent and Ethics Committee approval, we studied 160 ASA I–II patients, aged 18 yr or more, requiring anaesthesia lasting 20–60 min for a variety of different surgical procedures involving skin incision. Exclusion criteria included gross obesity, a history of drug or alcohol abuse, and concomitant drug therapy likely to influence the course of anaesthesia, the haemodynamic response to surgical stimuli or the insertion of a laryngeal mask. One of the authors (D.R. or M.W.) was present during each anaesthetic for data collection purposes only and did not become involved in the conduct of anaesthesia. Data recorded included cardiorespiratory variables, timing of events, presence or absence of movement during surgery and the nature of all alterations to the manual or target-controlled system. Patients were premedicated with oral temazepam 20 mg approximately 1 h before induction of anaesthesia. After venous cannulation, baseline measurements of heart rate and non-invasive arterial pressure were recorded, after which fentanyl 1.5 g kg91 was administered i.v. Two minutes later, after a short period of preoxygenation, induction of anaesthesia was commenced using either manual or target control. Patients were asked to count out aloud during induction and the time to achieve induction, defined as loss of verbal control, was recorded. A laryngeal mask airway (LMA) was inserted as soon as considered appropriate, and thereafter patients breathed 67 % nitrous oxide in oxygen from a Mapleson A circuit, with a fresh gas flow of 70–100 ml91 kg91 min91. Ventilation was assisted if required until return of spontaneous respiration. Additional analgesia was not given unless deemed absolutely necessary, and propofol administration was adjusted throughout by alterations to either the target concentration or manual infusion system as appropriate. On completion of surgery, administration of nitrous oxide and propofol was discontinued, and the time to recovery (eye opening spontaneously or to command) and orientation (recall of date of birth) were recorded. For each end-point, a mean value was calculated for each group of 10 patients for each anaesthetist. These mean values (n ⫽ 8 for each system) were then compared using analysis of variance and of covariance. The chi-square test was used for qualitative data. P : 0.05 was taken throughout as reflecting statistical significance.
Results The two groups of patients were similar in age, weight, sex and preinduction arterial pressure and heart rate (table 1), and the majority was ASA I. They underwent a wide range of surgical procedures,
564
British Journal of Anaesthesia
Table 1 Mean (SD or range) patient characteristics and induction and preinduction arterial pressure (systolic (SAP) and diastolic (DAP)) and heart rate (HR) values in the manually controlled and target-controlled (TCI) infusion groups. *P : 0.05, **P : 0.01 (analysis of variance/covariance) Variable
Manual
Age (yr) Weight (kg) SAP before induction (mm Hg) DAP before induction (mm Hg) HR before induction (beat min91) Induction time (s) Time to insert LMA (s) Propofol to insert LMA (mg) SAP 2 min after LMA (mm Hg) DAP 2 min after LMA (mm Hg) HR 2 min after LMA (beat min91)
41.5 (16–77) 40.4 (18–83) 68.3 (3.13) 71.2 (3.53) 133.2 (5.51) 135.1 (7.03)
TCI
80.8 (2.83)
81.2 (3.00)
76.9 (3.57)
77.3 (4.66)
75 (18.8)** 132 (16.8)* 160 (23.5)*
55 (10.6)** 114 (17.4)* 201 (42.5)*
110.4 (9.23)
111.3 (6.62)
65.5 (7.81)
64.2 (6.87)
70.9 (3.83)
72.3 (4.19)
Table 2 Mean (SD) haemodynamic variables (systolic (SAP) and diastolic (DAP) arterial pressure and heart rate (HR)) during the early, middle and late maintenance periods in the manually controlled and target-controlled (TCI) infusion groups. *P : 0.05 (analysis of variance) Manual Preincision: SAP (mm Hg) DAP (mm Hg) HR (beat min91) Early (1–5 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91) Middle (10–30 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91) Late (930 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91)
TCI
105.2 (6.38) 107.6 (5.37) 60.6 (6.42) 61.7 (6.75) 67.0 (4.57)* 72.6 (5.55)* 109.8 (5.85) 109.6 (5.71) 64.5 (4.60) 63.5 (5.37) 68.4 (3.53)* 74.6 (4.55)* 113.2 (5.96) 113.2 (4.49) 65.8 (4.87) 64.2 (4.74) 69.8 (3.49)* 75.5 (4.04)* 116.8 (9.46) 68.3 (8.59) 70.4 (8.01)
Table 3 Mean (SD) end tidal carbon dioxide concentration (kPa) immediately before incision and during maintenance in the manually controlled and target-controlled (TCI) infusion groups. Data were recorded during spontaneous respiration only. *P : 0.05 (analysis of variance)
114.4 (8.00) 61.6 (7.49) 74.4 (9.95)
the most common of which were arthroscopy, varicose vein surgery and hernia repair. Data for one patient in the target-controlled system group, erroneously recruited despite receiving atenolol for hypertension, were excluded from analysis. Induction of anaesthesia and insertion of the LMA were achieved significantly more rapidly in the target-controlled group, at the expense of significantly increased propofol consumption (table 1). The frequency of apnoea (exceeding 20 s) during induction was similar, occurring in 55 of 79 (69.6 %) patients in the target-controlled group and in 53 of 80 (66.2 %) in the manual group. Three patients in the target-controlled group were withdrawn from the study after induction of anaesthesia because of technical problems. The targetcontrolled systems used in the study were capable of operating using integral batteries while disconnected from mains electricity. On three occasions the batteries became discharged during transfer of the
Preincision Early (1–5 min after incision) Middle (10–30 min after incision) Late (⬎30 min after incision)
Manual
TCI
6.5 (0.38) 6.4 (0.57) 6.5 (0.47)* 6.3 (0.62)
6.6 (0.76) 6.8 (0.74) 7.1 (0.69)* 6.9 (0.78)
patient from the anaesthetic room to theatre, and anaesthesia was maintained subsequently with inhalation agents. There was more movement in response to the initial surgical incision in the manual group (23/80 (28.8 %)) compared with the target-controlled group (15/76 (19.7 %)), but this failed to reach statistical significance (P ⫽ 0.19). The initial surgical incision resulted in similar, clinically insignificant, haemodynamic responses in both groups (table 2). During maintenance of anaesthesia there were no statistically significant differences in arterial pressure between the groups (table 2). Mean heart rate was significantly greater before incision in the targetcontrolled group, and remained so during the early (1–5 min after incision) and middle (10–30 min) maintenance periods. However, this was not considered to be clinically significant. Mean end-tidal carbon dioxide concentration tended to be higher in the target-controlled group, and the difference was significant during the middle maintenance period (table 3). There was a tendency for more apnoea during maintenance in the targetcontrolled group, with 30 of 80 (37.5 %) patients in the manual group requiring intermittent positive pressure ventilation (IPPV) for a mean period of 5.7 min (range 1–32 min), compared with 35 of 76 (46 %) in the target-controlled group who required IPPV for 9.1 min (1–27 min). However, the frequency of apnoea was not significantly different between the two groups (P ⫽ 0.28, chi-square test). Movement during maintenance occurred significantly more frequently (P ⫽ 0.02) in the manual group, in 21 of 80 (26.2 %) patients, compared with nine of 76 (11.8 %) in the target-controlled group. Movement was said to be “marked”, causing temporary interruption to surgery on four occasions in the manual group compared with one in the target-controlled group. The amount of propofol infused to achieve induction (insertion of the LMA) was subtracted from the total amount infused by the end of the procedure, to determine the amount of propofol used to maintain anaesthesia. The mean maintenance infusion rate (in mg kg91 h91) was then calculated by dividing this by the duration of maintenance and body weight. The mean (SD) overall infusion rate during maintenance was significantly greater in the target-controlled group (P ⫽ 0.001, ANOVA), at 13.2 (4.77) mg kg91 h91, compared with 8.2 (2.32) mg kg91 h91 in the manual group. The duration of propofol administration was similar in the two groups, and despite the signifi-
Manual vs target-controlled propofol infusion
565
Table 4 Mean (SD) duration of propofol administration and times to recovery (eyes opening spontaneously or to speech) and orientation (recall of date of birth) from the end of surgery in the manually controlled and target-controlled (TCI) infusion groups
Duration of propofol administration (min) Recovery (min) Orientation (min)
Manual
TCI
40.8 (5.82)
37.2 (8.19)
6.2 (3.48) 8.0 (2.66)
8.5 (6.49) 10.5 (6.37)
Table 5 Ease of use: preferences expressed by the eight anaesthetists regarding the variables shown
Ease of set-up Ease of setting target or infusion rate Ease of adjusting depth of anaesthesia Portability Reliability Ease of use Preferred choice of infusion technique
Manual
TCI
No preference
3 0
1 6
4 2
1
4
3
4 2 1 0
0 1 6 7
4 5 1 1
cantly greater amount administered during both induction and maintenance in the target-controlled group, the times to recovery (eyes opening spontaneously or to speech) and orientation (ability to recall date of birth) from the end of surgery were not significantly prolonged compared with the manual group (table 4). Although no additional systemic analgesics were administered during anaesthesia, similar numbers of patients in both groups (15 patients in the target-controlled group and 16 in the manual group) received bupivacaine by wound infiltration-nerve block, which by reducing the stimulus from the site of operation may have increased their recovery time. Most anaesthetists felt confident after anaesthetizing two or three patients with either technique, and at the end of the study each completed a questionnaire on the systems (table 5). Six of the eight anaesthetists found the target-controlled system easier to use, and seven would choose to use it in preference to manual control using the Ohmeda 9000 syringe driver.
Discussion INDUCTION OF ANAESTHESIA
We found that use of the target-controlled system resulted in more rapid induction of anaesthesia and allowed earlier insertion of an LMA compared with the manual system. With the target-controlled system, induction of anaesthesia is achieved more rapidly with higher initial target concentrations [8] leading to faster attainment of an adequate effector site concentration. Furthermore, the targetcontrolled system is designed to maintain a steady blood propofol concentration until a higher or lower target concentration is selected, and this may not have been achieved with the manual regimen used.
During the first minute of induction the mean target concentration was 7.5 g ml91 (range 4– 12 g ml91), which for a 72-kg patient (the mean weight in the target-controlled group) represents a bolus of 130 mg, followed by a slowly reducing infusion of approximately 200 ml h91. At the end of the second minute of induction, the mean target was 7.8 g ml91 (range 4–15 g ml91). This is just within the 4–8 g ml91 initial target concentration suggested at the workshop, but it is clear that on occasions target concentrations which were considerably outside this range were used. This may reflect a desire on the part of the anaesthetist to achieve a more rapid induction or may indicate a degree of inter-patient variability in terms of response to propofol. RESPONSE TO SURGICAL STIMULUS
The presence or absence of movement in response to surgical stimulus is dependent on the degree of stimulus involved and the concentration of propofol in the brain, which in turn is related to the blood concentration of propofol. The time between insertion of the LMA and surgical incision is usually a period of reduced stimulus and, depending on the duration of this period, it is often appropriate to allow the blood propofol concentration to decrease. While this is likely to occur automatically with manual infusion rates of 6 mg kg91 h91, a conscious decision to reduce the target concentration must be made if this is considered appropriate. Administration of 67 % nitrous oxide in oxygen further reduces propofol requirements [9], but nevertheless it is important to ensure that the blood propofol concentration is increased a few minutes before incision to minimize the occurrence of movement in response to surgical stimuli both at the beginning of surgery and later during the procedure. Some of the anaesthetists left the target-controlled system unchanged between induction and incision, while some altered the target concentration as described above. The mean target concentration immediately before incision was 7.0 g ml91, and the fact that movement during surgery occurred significantly less frequently in the target-controlled group suggests that higher blood propofol concentrations were achieved using the target-controlled system. This is borne out by the significant difference in propofol consumption between the two groups during maintenance, which may be interpreted in several ways. The flexibility of the target-controlled system may have encouraged a greater degree of titration such that a deeper level of anaesthesia was achieved more easily than with the manual system, with a more appropriate amount of propofol being administered. Conversely, use of the manual system may have discouraged excessive propofol administration. RESPIRATORY SYSTEM
Similar numbers of patients in both groups became apnoeic during induction, and apnoea recorded during maintenance was almost always the result of apnoea persisting from induction. This was probably
566 caused by the bolus dose of fentanyl 1.5 g kg91 sadministered before induction, which is higher than would normally be used in clinical practice. This dose of fentanyl was chosen to avoid further doses of analgesia being required during maintenance of anaesthesia, the aim being to test the adequacy of control of anaesthesia using propofol alone. The statistically significant increase in mean endtidal carbon dioxide concentration in the targetcontrolled group compared with the manually controlled group during the middle maintenance period was probably caused by increased blood propofol concentrations. CARDIOVASCULAR VARIABLES
There were no clinically significant differences in arterial pressure or heart rate at any time during the study. In particular, use of the target-controlled system, which was associated with significantly increased propofol administration during induction and maintenance, did not result in significant cardiovascular depression. RECOVERY FROM ANAESTHESIA
Duration of propofol administration was similar in the two groups and the increased propofol administration rate during maintenance in the targetcontrolled group did not result in significantly prolonged recovery. Most patients in both groups made a prompt, clear-headed recovery, although some required additional fentanyl and longer acting analgesics on awakening. EASE OF USE
All the anaesthetists quickly became familiar and confident with both techniques, with no evidence of a learning effect using the quality of anaesthesia scoring system proposed by Glen [10]. At the end of the study most expressed an overall preference for the target-controlled system. None considered setting the infusion rate on the Ohmeda 9000 syringe pump easier than selecting the target concentration on the target-controlled system, although four considered the syringe pump more portable than the target-controlled system. The overall preference for
British Journal of Anaesthesia the target-controlled system may reflect increased confidence regarding predictability of anaesthetic effect. It may also be related to the fact that when making the transition to i.v. anaesthesia, with the target-controlled system traditional anaesthetic skills are all that are required. Target propofol concentrations can be increased and decreased according to response in the same way that anaesthetists are familiar with varying the inspired concentration of inhalation agents.
Acknowledgements We gratefully acknowledge the assistance of Dr D. Howes, Dr P. Jackson, Dr M. J. McNeill, Dr P. Milns, Dr J. A. Patrick, Dr G. A. Sutherland, Dr J. Watt and Dr A. Winter in the conduct of this study. The study was supported financially by Zeneca.
References 1. Stokes DN, Hutton P. Rate dependent induction phenomena with propofol: implications for the relative potency of intravenous anesthetics. Anesthesia and Analgesia 1991; 72: 578–583. 2. Stokes DN, Peacock JE, Lewis R, Hutton P. The Ohmeda 9000 syringe pump. The first of a new generation of syringe drivers. Anaesthesia 1990; 45: 1062–1066. 3. Gill SS, Lewis RP, Reilly CS. Maintenance of anaesthesia with propofol—a comparative study of a stepdown infusion of propofol and a low dose infusion supplemented by incremental doses. European Journal of Anaesthesiology 1992; 9: 203–207. 4. White M, Kenny GNC. Intravenous propofol using a computerised infusion system. Anaesthesia 1990; 45: 204–209. 5. Kenny GNC, White M. A portable computerised infusion system for propofol. Anaesthesia 1990; 45: 692–693. 6. Taylor I, White M, Kenny GNC. Assessment of the value and pattern of use of a target controlled propofol infusion system. International Journal of Clinical Monitoring and Computing 1993; 10: 175–180. 7. Alvis JM, Reves JG, Govier AV, Menkhaus PG, Hanling CE, Spain JS, Bradley E. Computer-assisted continuous infusions of fentanyl during cardiac anesthesia: comparison with a manual method. Anesthesiology 1985; 63: 41–49. 8. Chaudhri S, White M, Kenny GNC. Induction of anaesthesia with propofol using a target controlled infusion system. Anaesthesia 1992; 47: 551–553. 9. Davidson JAH, MacLeod AD, Howie JC, White M, Kenny GNC. Effective concentration 50 for propofol with and without 67 % nitrous oxide. Acta Anaesthesiologica Scandinavica 1993; 37: 458–464. 10. Glen JB. Quality of anaesthesia during spontaneous respiration: a proposed scoring system. Anaesthesia 1991; 46: 1081–1082.
Manual compared with target-controlled infusion of propofol D. RUSSELL, M. P. WILKES, S. C. HUNTER, J. B. GLEN, P. HUTTON AND G. N. C. KENNY Summary We studied 160 ASA l–ll patients, anaesthetized with propofol by infusion, using either a manually controlled or target-controlled infusion system. Patients were anaesthetized by eight consultant anaesthetists who had little or no previous experience of the use of propofol by infusion. In addition to propofol, patients received temazepam premedication, a single dose of fentanyl and 67 % nitrous oxide in oxygen. Each consultant anaesthetized 10 patients in sequential fashion with each system. Use of the target-controlled infusion resulted in more rapid induction of anaesthesia and allowed earlier insertion of a laryngeal mask airway. There was a tendency towards less movement in response to the initial surgical stimulus and significantly less movement during the remainder of surgery. Significantly more propofol was administered during both induction and maintenance of anaesthesia with the target-controlled system. This was associated with significantly increased end-tidal carbon dioxide measurements during the middle period of maintenance only, but recovery from anaesthesia was not significantly prolonged in the target-controlled group. With the exception of a clinically insignificant difference in heart rate, haemodynamic variables were similar in the two groups. Six of the eight anaesthetists found the target-controlled system easier to use, and seven would use the target-controlled system in preference to a manually controlled infusion. Anaesthetists without prior experience of propofol infusion anaesthesia quickly became familiar with both manual and target-controlled techniques, and expressed a clear preference for the target-controlled system. (Br. J. Anaesth. 1995; 75: 562–566) Key words Anaesthetics i.v., propofol. Anaesthetic techniques, i.v. infusion. Equipment, computers. Equipment, infusion systems.
Propofol is an i.v. anaesthetic agent suitable for induction and maintenance of general anaesthesia. When used in this fashion, the aim is to infuse propofol at an appropriate rate, with supplementary boluses as required, to achieve an effective blood propofol concentration appropriate to the degree of surgical stimulation. It is important to avoid excessive administration which may be associated with adverse cardiorespiratory effects and delayed recovery. Stokes and Hutton [1] studied 80 patients premedicated with temazepam undergoing body surface
surgery. They compared induction with propofol administered at four different infusion rates from an Ohmeda 9000 syringe pump [2] until verbal contact with the patient was lost, and recommended that to achieve induction, propofol should be infused at a rate of 600 ml h91. They then maintained anaesthesia using a manually controlled infusion of propofol set initially at 6 mg kg91 h91 and thereafter adjusted the infusion rate such that movement in response to surgical stimulation was eliminated. Fentanyl 1.5 g kg91 was administered 5 min before induction, and patients breathed spontaneously from a Mapleson A system delivering 67 % nitrous oxide in oxygen. In a subsequent study, Gill, Lewis and Reilly [3] infused propofol at 600 ml h91 to achieve induction, and then compared maintenance at either a fixed rate of 6 mg kg91 h91 or a “stepdown” regimen of 10 mg kg91 h91 for 10 min, 8 mg kg91 h91 for 10 min and 6 mg kg91 h91 thereafter. Patients in both groups required additional incremental doses of propofol 20 mg in order to maintain adequate anaesthesia, but both methods were considered satisfactory. White and Kenny [4, 5] developed a portable computerized infusion system which allows the anaesthetist to select any desired target blood concentration of propofol at any time. The system was based on the series II Psion handheld computer interfaced with an Ohmeda 9000 syringe driver via an RS232 linkage but has since been superseded by a backbar integrated with the syringe driver. The software for the administration of propofol (Diprifusor; Zeneca Ltd) consists of a pharmacokinetic model, a set of specific pharmacokinetic variables for propofol and infusion control algorithms. To achieve induction, the patient’s weight and age are entered, and then a target blood propofol concentration is selected, depending on age, ASA status and concomitant opioid administration. Thereafter, the control unit commands the infusion pump to deliver a rapid zero order infusion at a rate of 1200 ml h91 until the pharmacokinetic model in the target-controlled system predicts that the desired target concentration has been achieved. It then D. RUSSELL, FRCA, University Department of Anaesthesia, Royal Infirmary, Glasgow G31 2ER. M. P. WILKES, FRCA, P. HUTTON, BSC, PHD, FRCA, University Department of Anaesthesia, Queen Elizabeth Hospital, Birmingham B15 2TH. S. C. HUNTER, J. B. GLEN, BVMS, PHD, DVA, Clinical Research Group, Zeneca Pharmaceuticals, Macclesfield, Cheshire SK10 4TG. G. N. C. KENNY, BSC (HONS), FRCA, MD, Glasgow University Department of Anaesthesia, HCI Hospital, Beardmore Street, Clydebank, Glasgow G81 4HX. Accepted for publication: June 2, 1995. Correspondence to G.N.C.K.
Manual vs target-controlled propofol infusion provides a variable rate infusion as required to maintain that target concentration. The anaesthetist has at all times the option of selecting a higher or lower target concentration, with the rate of delivery of propofol by the syringe pump being controlled automatically. If a higher target concentration is requested, this results in administration of another bolus followed by infusion at an increased rate, while a request for a lower target concentration results in temporary discontinuation of infusion followed by resumption at a lower rate. The target blood propofol concentration is therefore titrated upwards and downwards in response to clinical signs, thus increasing or decreasing the blood and effector site concentrations, in the same way that the inspired concentration of traditional inhalation agents such as halothane may be increased or decreased to achieve alterations in end-tidal and brain partial pressures. When introduced into clinical practice [6], the target-controlled infusion system proved to be popular, mainly because it was easy to use and provided a high level of confidence regarding the predictability of anaesthetic effects. Furthermore, in a study of patients receiving anaesthesia for cardiac surgery, Alvis and colleagues [7] demonstrated improved cardiovascular stability from the use of a computer-assisted fentanyl infusion compared with a manual technique. The aims of the present study were to compare the quality of anaesthesia achieved by manual infusion of propofol as described by Stokes and Hutton [1] with that achieved by the target-controlled infusion system and to compare the ease of use, pattern of propofol administration and time to recovery from anaesthesia. We also sought to examine the rate at which anaesthetists became familiar and confident with one or other technique.
Patients and methods Eight consultant anaesthetists from two centres (four from each), unfamiliar with the administration of propofol by infusion, attended a workshop where they received instruction and practical demonstration of the manual and target-controlled systems. Comprehensive written instructions in the rationale of each technique and practical aspects of the systems were provided. With the target-controlled system it was suggested that the initial target concentration should be between 4 and 8 g ml91 depending on the degree of hypnosis present immediately before induction, with titration thereafter according to response. For a 70-kg patient, the selection of a target propofol concentration of 6 g ml91 would result in the administration of approximately 110 mg over 33 s, followed by a slowly decreasing infusion of 156 ml h91. With the manual system the infusion rate was 600 ml h91 until loss of consciousness (loss of verbal contact), followed immediately by an initial maintenance infusion rate of 6 mg kg91 h91 (42 ml h91 for a 70-kg patient) and a variable infusion rate thereafter of up to 15 mg kg91 h91. Additional propofol was infused at 600 ml h91 as judged appropriate to achieve induction, or if it was felt that anaesthetic depth was inadequate. To allow exam-
563 ination of the rate at which anaesthetists became familiar with these techniques, each consultant anaesthetized 10 patients in sequential fashion with each system. Thus individual patients were not allocated randomly to one or other system, but the order in which each anaesthetist used both systems was randomized. After obtaining informed written consent and Ethics Committee approval, we studied 160 ASA I–II patients, aged 18 yr or more, requiring anaesthesia lasting 20–60 min for a variety of different surgical procedures involving skin incision. Exclusion criteria included gross obesity, a history of drug or alcohol abuse, and concomitant drug therapy likely to influence the course of anaesthesia, the haemodynamic response to surgical stimuli or the insertion of a laryngeal mask. One of the authors (D.R. or M.W.) was present during each anaesthetic for data collection purposes only and did not become involved in the conduct of anaesthesia. Data recorded included cardiorespiratory variables, timing of events, presence or absence of movement during surgery and the nature of all alterations to the manual or target-controlled system. Patients were premedicated with oral temazepam 20 mg approximately 1 h before induction of anaesthesia. After venous cannulation, baseline measurements of heart rate and non-invasive arterial pressure were recorded, after which fentanyl 1.5 g kg91 was administered i.v. Two minutes later, after a short period of preoxygenation, induction of anaesthesia was commenced using either manual or target control. Patients were asked to count out aloud during induction and the time to achieve induction, defined as loss of verbal control, was recorded. A laryngeal mask airway (LMA) was inserted as soon as considered appropriate, and thereafter patients breathed 67 % nitrous oxide in oxygen from a Mapleson A circuit, with a fresh gas flow of 70–100 ml91 kg91 min91. Ventilation was assisted if required until return of spontaneous respiration. Additional analgesia was not given unless deemed absolutely necessary, and propofol administration was adjusted throughout by alterations to either the target concentration or manual infusion system as appropriate. On completion of surgery, administration of nitrous oxide and propofol was discontinued, and the time to recovery (eye opening spontaneously or to command) and orientation (recall of date of birth) were recorded. For each end-point, a mean value was calculated for each group of 10 patients for each anaesthetist. These mean values (n ⫽ 8 for each system) were then compared using analysis of variance and of covariance. The chi-square test was used for qualitative data. P : 0.05 was taken throughout as reflecting statistical significance.
Results The two groups of patients were similar in age, weight, sex and preinduction arterial pressure and heart rate (table 1), and the majority was ASA I. They underwent a wide range of surgical procedures,
564
British Journal of Anaesthesia
Table 1 Mean (SD or range) patient characteristics and induction and preinduction arterial pressure (systolic (SAP) and diastolic (DAP)) and heart rate (HR) values in the manually controlled and target-controlled (TCI) infusion groups. *P : 0.05, **P : 0.01 (analysis of variance/covariance) Variable
Manual
Age (yr) Weight (kg) SAP before induction (mm Hg) DAP before induction (mm Hg) HR before induction (beat min91) Induction time (s) Time to insert LMA (s) Propofol to insert LMA (mg) SAP 2 min after LMA (mm Hg) DAP 2 min after LMA (mm Hg) HR 2 min after LMA (beat min91)
41.5 (16–77) 40.4 (18–83) 68.3 (3.13) 71.2 (3.53) 133.2 (5.51) 135.1 (7.03)
TCI
80.8 (2.83)
81.2 (3.00)
76.9 (3.57)
77.3 (4.66)
75 (18.8)** 132 (16.8)* 160 (23.5)*
55 (10.6)** 114 (17.4)* 201 (42.5)*
110.4 (9.23)
111.3 (6.62)
65.5 (7.81)
64.2 (6.87)
70.9 (3.83)
72.3 (4.19)
Table 2 Mean (SD) haemodynamic variables (systolic (SAP) and diastolic (DAP) arterial pressure and heart rate (HR)) during the early, middle and late maintenance periods in the manually controlled and target-controlled (TCI) infusion groups. *P : 0.05 (analysis of variance) Manual Preincision: SAP (mm Hg) DAP (mm Hg) HR (beat min91) Early (1–5 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91) Middle (10–30 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91) Late (930 min after incision) SAP (mm Hg) DAP (mm Hg) HR (beat min91)
TCI
105.2 (6.38) 107.6 (5.37) 60.6 (6.42) 61.7 (6.75) 67.0 (4.57)* 72.6 (5.55)* 109.8 (5.85) 109.6 (5.71) 64.5 (4.60) 63.5 (5.37) 68.4 (3.53)* 74.6 (4.55)* 113.2 (5.96) 113.2 (4.49) 65.8 (4.87) 64.2 (4.74) 69.8 (3.49)* 75.5 (4.04)* 116.8 (9.46) 68.3 (8.59) 70.4 (8.01)
Table 3 Mean (SD) end tidal carbon dioxide concentration (kPa) immediately before incision and during maintenance in the manually controlled and target-controlled (TCI) infusion groups. Data were recorded during spontaneous respiration only. *P : 0.05 (analysis of variance)
114.4 (8.00) 61.6 (7.49) 74.4 (9.95)
the most common of which were arthroscopy, varicose vein surgery and hernia repair. Data for one patient in the target-controlled system group, erroneously recruited despite receiving atenolol for hypertension, were excluded from analysis. Induction of anaesthesia and insertion of the LMA were achieved significantly more rapidly in the target-controlled group, at the expense of significantly increased propofol consumption (table 1). The frequency of apnoea (exceeding 20 s) during induction was similar, occurring in 55 of 79 (69.6 %) patients in the target-controlled group and in 53 of 80 (66.2 %) in the manual group. Three patients in the target-controlled group were withdrawn from the study after induction of anaesthesia because of technical problems. The targetcontrolled systems used in the study were capable of operating using integral batteries while disconnected from mains electricity. On three occasions the batteries became discharged during transfer of the
Preincision Early (1–5 min after incision) Middle (10–30 min after incision) Late (⬎30 min after incision)
Manual
TCI
6.5 (0.38) 6.4 (0.57) 6.5 (0.47)* 6.3 (0.62)
6.6 (0.76) 6.8 (0.74) 7.1 (0.69)* 6.9 (0.78)
patient from the anaesthetic room to theatre, and anaesthesia was maintained subsequently with inhalation agents. There was more movement in response to the initial surgical incision in the manual group (23/80 (28.8 %)) compared with the target-controlled group (15/76 (19.7 %)), but this failed to reach statistical significance (P ⫽ 0.19). The initial surgical incision resulted in similar, clinically insignificant, haemodynamic responses in both groups (table 2). During maintenance of anaesthesia there were no statistically significant differences in arterial pressure between the groups (table 2). Mean heart rate was significantly greater before incision in the targetcontrolled group, and remained so during the early (1–5 min after incision) and middle (10–30 min) maintenance periods. However, this was not considered to be clinically significant. Mean end-tidal carbon dioxide concentration tended to be higher in the target-controlled group, and the difference was significant during the middle maintenance period (table 3). There was a tendency for more apnoea during maintenance in the targetcontrolled group, with 30 of 80 (37.5 %) patients in the manual group requiring intermittent positive pressure ventilation (IPPV) for a mean period of 5.7 min (range 1–32 min), compared with 35 of 76 (46 %) in the target-controlled group who required IPPV for 9.1 min (1–27 min). However, the frequency of apnoea was not significantly different between the two groups (P ⫽ 0.28, chi-square test). Movement during maintenance occurred significantly more frequently (P ⫽ 0.02) in the manual group, in 21 of 80 (26.2 %) patients, compared with nine of 76 (11.8 %) in the target-controlled group. Movement was said to be “marked”, causing temporary interruption to surgery on four occasions in the manual group compared with one in the target-controlled group. The amount of propofol infused to achieve induction (insertion of the LMA) was subtracted from the total amount infused by the end of the procedure, to determine the amount of propofol used to maintain anaesthesia. The mean maintenance infusion rate (in mg kg91 h91) was then calculated by dividing this by the duration of maintenance and body weight. The mean (SD) overall infusion rate during maintenance was significantly greater in the target-controlled group (P ⫽ 0.001, ANOVA), at 13.2 (4.77) mg kg91 h91, compared with 8.2 (2.32) mg kg91 h91 in the manual group. The duration of propofol administration was similar in the two groups, and despite the signifi-
Manual vs target-controlled propofol infusion
565
Table 4 Mean (SD) duration of propofol administration and times to recovery (eyes opening spontaneously or to speech) and orientation (recall of date of birth) from the end of surgery in the manually controlled and target-controlled (TCI) infusion groups
Duration of propofol administration (min) Recovery (min) Orientation (min)
Manual
TCI
40.8 (5.82)
37.2 (8.19)
6.2 (3.48) 8.0 (2.66)
8.5 (6.49) 10.5 (6.37)
Table 5 Ease of use: preferences expressed by the eight anaesthetists regarding the variables shown
Ease of set-up Ease of setting target or infusion rate Ease of adjusting depth of anaesthesia Portability Reliability Ease of use Preferred choice of infusion technique
Manual
TCI
No preference
3 0
1 6
4 2
1
4
3
4 2 1 0
0 1 6 7
4 5 1 1
cantly greater amount administered during both induction and maintenance in the target-controlled group, the times to recovery (eyes opening spontaneously or to speech) and orientation (ability to recall date of birth) from the end of surgery were not significantly prolonged compared with the manual group (table 4). Although no additional systemic analgesics were administered during anaesthesia, similar numbers of patients in both groups (15 patients in the target-controlled group and 16 in the manual group) received bupivacaine by wound infiltration-nerve block, which by reducing the stimulus from the site of operation may have increased their recovery time. Most anaesthetists felt confident after anaesthetizing two or three patients with either technique, and at the end of the study each completed a questionnaire on the systems (table 5). Six of the eight anaesthetists found the target-controlled system easier to use, and seven would choose to use it in preference to manual control using the Ohmeda 9000 syringe driver.
Discussion INDUCTION OF ANAESTHESIA
We found that use of the target-controlled system resulted in more rapid induction of anaesthesia and allowed earlier insertion of an LMA compared with the manual system. With the target-controlled system, induction of anaesthesia is achieved more rapidly with higher initial target concentrations [8] leading to faster attainment of an adequate effector site concentration. Furthermore, the targetcontrolled system is designed to maintain a steady blood propofol concentration until a higher or lower target concentration is selected, and this may not have been achieved with the manual regimen used.
During the first minute of induction the mean target concentration was 7.5 g ml91 (range 4– 12 g ml91), which for a 72-kg patient (the mean weight in the target-controlled group) represents a bolus of 130 mg, followed by a slowly reducing infusion of approximately 200 ml h91. At the end of the second minute of induction, the mean target was 7.8 g ml91 (range 4–15 g ml91). This is just within the 4–8 g ml91 initial target concentration suggested at the workshop, but it is clear that on occasions target concentrations which were considerably outside this range were used. This may reflect a desire on the part of the anaesthetist to achieve a more rapid induction or may indicate a degree of inter-patient variability in terms of response to propofol. RESPONSE TO SURGICAL STIMULUS
The presence or absence of movement in response to surgical stimulus is dependent on the degree of stimulus involved and the concentration of propofol in the brain, which in turn is related to the blood concentration of propofol. The time between insertion of the LMA and surgical incision is usually a period of reduced stimulus and, depending on the duration of this period, it is often appropriate to allow the blood propofol concentration to decrease. While this is likely to occur automatically with manual infusion rates of 6 mg kg91 h91, a conscious decision to reduce the target concentration must be made if this is considered appropriate. Administration of 67 % nitrous oxide in oxygen further reduces propofol requirements [9], but nevertheless it is important to ensure that the blood propofol concentration is increased a few minutes before incision to minimize the occurrence of movement in response to surgical stimuli both at the beginning of surgery and later during the procedure. Some of the anaesthetists left the target-controlled system unchanged between induction and incision, while some altered the target concentration as described above. The mean target concentration immediately before incision was 7.0 g ml91, and the fact that movement during surgery occurred significantly less frequently in the target-controlled group suggests that higher blood propofol concentrations were achieved using the target-controlled system. This is borne out by the significant difference in propofol consumption between the two groups during maintenance, which may be interpreted in several ways. The flexibility of the target-controlled system may have encouraged a greater degree of titration such that a deeper level of anaesthesia was achieved more easily than with the manual system, with a more appropriate amount of propofol being administered. Conversely, use of the manual system may have discouraged excessive propofol administration. RESPIRATORY SYSTEM
Similar numbers of patients in both groups became apnoeic during induction, and apnoea recorded during maintenance was almost always the result of apnoea persisting from induction. This was probably
566 caused by the bolus dose of fentanyl 1.5 g kg91 sadministered before induction, which is higher than would normally be used in clinical practice. This dose of fentanyl was chosen to avoid further doses of analgesia being required during maintenance of anaesthesia, the aim being to test the adequacy of control of anaesthesia using propofol alone. The statistically significant increase in mean endtidal carbon dioxide concentration in the targetcontrolled group compared with the manually controlled group during the middle maintenance period was probably caused by increased blood propofol concentrations. CARDIOVASCULAR VARIABLES
There were no clinically significant differences in arterial pressure or heart rate at any time during the study. In particular, use of the target-controlled system, which was associated with significantly increased propofol administration during induction and maintenance, did not result in significant cardiovascular depression. RECOVERY FROM ANAESTHESIA
Duration of propofol administration was similar in the two groups and the increased propofol administration rate during maintenance in the targetcontrolled group did not result in significantly prolonged recovery. Most patients in both groups made a prompt, clear-headed recovery, although some required additional fentanyl and longer acting analgesics on awakening. EASE OF USE
All the anaesthetists quickly became familiar and confident with both techniques, with no evidence of a learning effect using the quality of anaesthesia scoring system proposed by Glen [10]. At the end of the study most expressed an overall preference for the target-controlled system. None considered setting the infusion rate on the Ohmeda 9000 syringe pump easier than selecting the target concentration on the target-controlled system, although four considered the syringe pump more portable than the target-controlled system. The overall preference for
British Journal of Anaesthesia the target-controlled system may reflect increased confidence regarding predictability of anaesthetic effect. It may also be related to the fact that when making the transition to i.v. anaesthesia, with the target-controlled system traditional anaesthetic skills are all that are required. Target propofol concentrations can be increased and decreased according to response in the same way that anaesthetists are familiar with varying the inspired concentration of inhalation agents.
Acknowledgements We gratefully acknowledge the assistance of Dr D. Howes, Dr P. Jackson, Dr M. J. McNeill, Dr P. Milns, Dr J. A. Patrick, Dr G. A. Sutherland, Dr J. Watt and Dr A. Winter in the conduct of this study. The study was supported financially by Zeneca.
References 1. Stokes DN, Hutton P. Rate dependent induction phenomena with propofol: implications for the relative potency of intravenous anesthetics. Anesthesia and Analgesia 1991; 72: 578–583. 2. Stokes DN, Peacock JE, Lewis R, Hutton P. The Ohmeda 9000 syringe pump. The first of a new generation of syringe drivers. Anaesthesia 1990; 45: 1062–1066. 3. Gill SS, Lewis RP, Reilly CS. Maintenance of anaesthesia with propofol—a comparative study of a stepdown infusion of propofol and a low dose infusion supplemented by incremental doses. European Journal of Anaesthesiology 1992; 9: 203–207. 4. White M, Kenny GNC. Intravenous propofol using a computerised infusion system. Anaesthesia 1990; 45: 204–209. 5. Kenny GNC, White M. A portable computerised infusion system for propofol. Anaesthesia 1990; 45: 692–693. 6. Taylor I, White M, Kenny GNC. Assessment of the value and pattern of use of a target controlled propofol infusion system. International Journal of Clinical Monitoring and Computing 1993; 10: 175–180. 7. Alvis JM, Reves JG, Govier AV, Menkhaus PG, Hanling CE, Spain JS, Bradley E. Computer-assisted continuous infusions of fentanyl during cardiac anesthesia: comparison with a manual method. Anesthesiology 1985; 63: 41–49. 8. Chaudhri S, White M, Kenny GNC. Induction of anaesthesia with propofol using a target controlled infusion system. Anaesthesia 1992; 47: 551–553. 9. Davidson JAH, MacLeod AD, Howie JC, White M, Kenny GNC. Effective concentration 50 for propofol with and without 67 % nitrous oxide. Acta Anaesthesiologica Scandinavica 1993; 37: 458–464. 10. Glen JB. Quality of anaesthesia during spontaneous respiration: a proposed scoring system. Anaesthesia 1991; 46: 1081–1082.