Secondary Logo

Journal Logo

Original Article

Use of a target-controlled infusion system for propofol does not improve subjective assessment of anaesthetic depth by inexperienced anaesthesiologists*

Rehberg, B.*; Ryll, C.*; Hadzidiakos, D.*; Baars, J.*

Author Information
European Journal of Anaesthesiology: November 2007 - Volume 24 - Issue 11 - p 920-926
doi: 10.1017/S0265021507000907
  • Free



Adjusting anaesthetic concentration to clinical needs is a difficult task to learn for anaesthesiology trainees, especially during total intravenous (i.v.) anaesthesia (TIVA). In a previous study [1], we have shown that inexperienced anaesthesiologists have more difficulties in assessing ‘anaesthetic depth’ during TIVA than experienced ones, leading to longer recovery times on the one hand and to a higher incidence of overly light anaesthesia on the other.

Target-controlled infusion (TCI) systems are a tool used to obtain more constant plasma levels of i.v. anaesthetics. There is evidence that TCI outperforms manually controlled infusion (MCI) regarding safety and clinical performance [2-6], probably by limiting the variability of the infusion regimen [7]. Since ‘depth of anaesthesia’ is a function of anaesthetic concentration, the calculated plasma concentrations supplied by TCI systems may also help in the assessment of anaesthetic depth by anaesthesiologists.

Therefore, we hypothesized that anaesthesiologists at the beginning of their training (<1 yr of clinical experience in anaesthesiology) might profit from TCI systems, leading to shorter and less variable recovery times and a better subjective assessment of anaesthetic depth. Since there is no gold standard to measure anaesthetic depth, we used the EEG-derived bispectral index (BIS) as a surrogate parameter of the underlying anaesthetic depth against which the subjective assessment was compared. In addition, we used the percentage of intraoperative data points with BIS values less than 40 and above 60 as a measure of too deep or too shallow ‘anaesthetic depth’, respectively.

To test these hypotheses, we performed a randomized study to compare subjective assessment of anaesthetic depth by inexperienced anaesthesiologists during total i.v. propofol/fentanyl anaesthesia performed either as conventional MCI or as TCI.

Materials and methods

Following Local Ethics Committee approval and written informed consent, 92 ASA I-III patients scheduled for elective minor urological or gynaecological surgery (laparoscopic, transurethral or breast surgery) under general anaesthesia were enrolled in the study. Excluded were patients scheduled for surgery which might interfere with cerebral function (e.g. carotid endarterectomy, cardiopulmonary bypass, neurosurgery), patients suffering from neurological or psychiatric disease, severe cardiac and vascular symptoms, and patients needing a rapid-sequence induction. In addition, patients were excluded when they were transferred postoperatively to the ICU or transferred intubated to the recovery unit.

Midazolam (0.1 mg kg−1) was administered orally 30 min before anaesthesia induction. Six anaesthesiologists with <1 yr of practical experience in anaesthesiology at the beginning of the study period participated in the study. All anaesthesiologists were supervised by an attending physician. However, the attending physician refrained from interference with propofol dosing unless patient safety was considered to be compromised. Prior to the study, anaesthesiologists had no experience with a TCI system, but were introduced to the concept of TCI at the beginning of the study by a short lecture and during two supervised anaesthetic cases.

Participating anaesthesiologists were allowed to perform total i.v. propofol anaesthesia supplemented by fentanyl but without nitrous oxide. No restrictions were made to drugs used for other purposes such as prevention of nausea and vomiting, hypo- or hypertension or to control heart rate (HR).

Ninety-two cases were prospectively randomized in two groups: 46 cases were performed as MCI with conventional syringe pumps. No restriction was specified concerning drug dosing. Propofol infusion and fentanyl boluses were titrated against clinical signs. However, the standards of our department suggest a propofol dosage of a bolus dose of 1-3 mg kg−1 for induction and a propofol infusion in the range of 4-8 mg kg−1 h−1 (66-120 μg kg−1min−1) for maintenance.

Forty-six cases were performed as TCI with a computerized TCI system (Base Primea®; Fresenius Medical Care, Brezins, France) and effect-site targeting. The TCI system was set to use Schnider's pharmacokinetic model for propofol [8], and a keo of 0.456 min−1 [9]. Anaesthesiologists were instructed to use an effect-site concentration (Ce Propofol) of 7-10 μg mL−1 for induction of anaesthesia and to reduce the concentration after intubation or laryngeal mask airway (LMA) insertion as appropriate for the level of surgical other stimuli in the range of 2.5-4.5 μg mL−1. Anaesthesiologists were also informed that the calculated time of return to consciousness (time from stopping propofol until return of consciousness (ROC)) is displayed on the TCI-pump monitor, based on a propofol concentration at ROC of 1.5 μg mL−1. Neuromuscular blockers were given only during anaesthesia induction to facilitate intubation, and no further intraoperative doses were given.

Prior to anaesthesia, self-adhesive electrodes for monitoring BIS (BIS XP version 3.1, Aspect Medical Systems; Newton, MA, USA) were attached to the patient's forehead. The BIS monitor was placed away from the anaesthesiologists' workplace. Following the connection of standard monitoring (electrocardiogram, HR, blood pressure (BP), pulse oximetry) anaesthesia was induced by propofol and fentanyl. Fentanyl was given as a bolus of 1-3 μg kg−1 approximately 1 min prior to the propofol bolus (MCI) or start of the propofol infusion (TCI). All patients were mechanically ventilated by an anaesthesia machine (Julian/Primus, Dräger, Lübeck, Germany).

Anaesthesiologists, who were blinded towards the BIS, were asked to assess anaesthetic depth subjectively using a 10-point numeric scale (1 = very deep anaesthesia, 10 = awake) [1]. Anaesthesiologists were given no clues how to define the scales or which anaesthetic endpoint (e.g. hypnosis, analgesia, immobility) the term anaesthesia should comprise. Assessment was performed and recorded in intervals of 2 min from injection of propofol until intubation (or LMA insertion), and at intervals of 5 min from intubation until start of surgery (skin incision). A typical case (TCI group) is shown in Figure 1. During surgery, assessment was performed every 10 min until end of the surgery (end of skin closure). From this point on, assessment was performed at intervals of 2 min until ROC and extubation. Intubation was defined as either endotracheal tube or LMA insertion. ROC was defined as eye opening and reaction to simple verbal commands.

Figure 1.
Figure 1.:
Example case (TCI group) showing the intervals of assessment of subjective depth of anaesthesia (open squares). The continuous line indicates the BIS and the dashed line the calculated propofol concentration in the effect compartment. TCI: target-controlled infusion; BIS: bispectral index.

At each time point, subjective assessment of anaesthetic depth, BIS, HR, BP, drug use (e.g. opioids, sympathomimetics) and special events (e.g. patient's movements) were recorded.

Time of surgery was measured as the time from skin incision to end of skin closure and recovery time was measured as the time from the end of surgery (end of skin closure) until ROC.

Postoperatively (12-24 h after the end of surgery), patients were interviewed for intraoperative awareness using a structured interview. Questions concerned the last memory before induction and first memory after ROC, memory of events in between and dreams during anaesthesia.


The association between subjective assessment and BIS was calculated using the prediction probability PK introduced by Smith and colleagues [10], using a macro for EXCEL (Microsoft; Redmond, WA, USA) kindly supplied by the author. Prediction probability is a non-parametric correlation measure, indicating the probability that a parameter (here the subjective assessment) correctly predicts anaesthetic depth (here the BIS). A Pk value of 1.0 indicates perfect prediction, whereas a value of 0.5 indicates that the predictive value of the parameter is not better than chance alone. Pk values were compared using the paired-data jacknife analysis. All other calculations were made using standard statistical software (Prism 3.0; Graphpad Software, San Diego, CA, USA). Values of P < 0.0.5 were considered significant.

To compare effect-site concentrations of propofol during anaesthesia using TCI and MCI, drug administration was recorded carefully. Effect-site concentrations of propofol in the MCI group were simulated after surgery using the STANPUMP program written by S. Shafer (freely available at and the Schnider [8] parameter set.

Effect-site concentrations of fentanyl were also recalculated via STANPUMP from the recorded dosing history, using the pharmacokinetic parameter set of Shafer and colleagues [11].

To detect periods of potentially light anaesthesia during surgery, we analysed the BIS during anaesthesia. The percentage of intraoperative BIS values >60 and <40 was calculated.


Ninety-two patients scheduled for elective surgery were randomized to receiving TIVA either as manually or target-controlled propofol infusion. Six patients were transferred unplanned to the ICU for surgical reasons and were therefore excluded after randomization. Data from the remaining 86 patients (42 in the TCI and 44 in the MCI group) were included in the analysis. The flow of patients is shown in Figure 2.

Figure 2.
Figure 2.:
Flow chart showing the progress of participants.

Six anaesthesiologists with <1 yr experience in anaesthesia performed both types of TIVA during the study. Mean duration of anaesthesia experience did not differ between the two groups (χ2-test).

The patient and perioperative characteristics are shown in Table 1. Between both groups (MCI vs. TCI), there was no significant difference for age, ASA status, duration of surgery and duration of anaesthesia (t-test).

Table 1
Table 1:
Patient characteristics.

Mean±SD fentanyl dose for induction was 170±80 μg in the MCI and 170±60 μg in the TCI group. Intraoperatively, a mean±SD fentanyl dose of 160±80 μg in the MCI and 170±10 μg in the TCI group were given, with no statistical difference between the two groups. Recalculation of the fentanyl concentration at ROC yielded median (25-75% percentile) effect compartment concentrations of 0.70 (0.50-0.94) ng mL−1 in the MCI and 0.68 (0.49-1.02) ng mL−1 in the TCI group.

Median calculated effect-site concentrations of propofol at different endpoints during anaesthesia and mean propofol concentration during surgery are shown in Table 2.

Table 2
Table 2:
Effect-site concentrations of Propofol (Ce Propofol) in μg mL−1 at different endpoints during anaesthesia.

Effect-site concentration at intubation (tracheal tube or LMA insertion) was significantly higher in the TCI group compared to MCI (U-test, P = 0.04). However, after splitting into subgroups (intubation vs. insertion of laryngeal mask) effect-site concentrations of propofol at intubation were not significantly different in the two groups. Median (25-75% quartiles) propofol effect-site concentrations in the MCI group were 4.3 (3.9-4.8) μg mL−1 at the time of intubation and 7.7 (6.3-9.6) μg mL−1 at LMA insertion. In the TCI group, the corresponding concentrations were 6.7 (3.8-7.8) μg mL−1 at the time of intubation and 8.8 (6.9-9.8) μg mL−1 at LMA insertion.

In the MCI group there was an overshoot in propofol effect-site concentrations during the first minutes after start of propofol administration. It should be noted that the mean time to intubation was different: MCI group - 4 min, TCI group - 5 min after the start of propofol. At all other time points, concentrations were not different in both groups (Table 2).

At the end of anaesthesia, the recovery time defined as the time from end of surgery (last skin closure) until ROC varied considerably in both groups (Fig. 3). The median recovery times (25-75% percentiles) were 15 (10-20) min for the MCI group and 16 (11-22) min for the TCI group and not significantly different. Median effect-site concentration of propofol at ROC was also not different (Table 2).

Figure 3.
Figure 3.:
Recovery times measured as time from end of surgery (skin closure) until ROC (open squares - MCI group, filled triangles - TCI group). Black bars indicate median values. ROC: reutrn of conciousness; MCI: manually controlled infusion; TCI: target-controlled infusion.

The median BIS (25-75% percentile) at intubation was 29 (25-37) in the MCI group, and 27 (22-39) in the TCI group. The median BIS averaged over the total period of surgery was 31 (25-36) in the MCI and 31 (25-39) in the TCI group. During surgery, the BIS was less than 40 for 74.5% of the time in the MCI group and for 68.8% in the TCI group. Compared to the TCI group, the MCI group showed a significantly higher percentage of BIS values above 60 (2.5% vs. 5.1%, χ2 test). Although BIS > 60 is considered a risk for intraoperative awareness, none of our patients expressed explicit memory for this period of time in the postoperative interview.

The association of BIS values with the subjective assessment of anaesthetic depth on the numerical sale is shown in Figure 4a for the MCI group and Figure 4b for the TCI group. This association between the BIS and the subjective assessment of anaesthetic depth was used to calculate the prediction probability Pk of the subjective assessment predicting the BIS as a surrogate parameter of underlying true anaesthetic depth.

Figure 4.
Figure 4.:
Distribution of BIS values at different ratings of subjectively assessed anaesthetic depth on an arbitrary numerical scale ranging from 10 - awake to 1 - deepest anaesthesia. (a) Assessment in the MCI group and (b) in the TCI group. Black bars indicate median values. BIS: bispectral index; MCI: manually controlled infusion; TCI: target-controlled infusion.

Judging from the Pk values, the association between the subjective assessment of anaesthetic depth and the EEG parameter does not show any significant difference between the two groups (MCI group: Pk = 0.80, TCI group: Pk =0.81).


The results of this study show that the use of a TCI system by anaesthesiology trainees does not lead to a reduction in recovery times and better assessment of anaesthetic depth. We had hypothesized that the association of subjective assessment of anaesthetic depth with the BIS (judged by the PK value) would improve by using the TCI system, since BIS is strongly dependent on propofol concentration. Implicitly, we had assumed that the anaesthesiologists would use the information of the propofol concentration to guide their anaesthetic depth assessment.

The negative finding of our study may thus be caused by either the inherent variability of the concentration - BIS relation, or by the failure of the participating anaesthesiologists to make use of the information. In the first case, the study results would suggest that TCI systems are useless for clinical practice, and in the second case the conclusion would be that the use of TCI systems needs training to be effective. It has been pointed out before [3] that training is required to make full use of a TCI system. However, trained anaesthesiologists are better able to assess anaesthetic depth even without the use of a TCI system, as we demonstrated in our previous study [1].

In our study, propofol administration with a TCI system was very similar to the MCI. Mean calculated concentration of propofol during surgery did not differ between the two groups. However, the percentage of BIS values above 60 was significantly smaller in the TCI group, confirming other studies that TCI reduces periods of light anaesthesia [12,13]. BIS values above 60 are associated with the risk of awareness, [14] and thus TCI can be considered to be contributing to patient safety in this respect. Surprisingly, the percentage of BIS values above 60 was much lower than expected from a previous study [1].

Analysis of recorded BIS values also revealed that in both groups BIS values were mostly below the recommended lower value of 40. Although this may have been inadvertent due to lack of experience or fear of too light anaesthesia, Iselin-Chaves and colleagues [15] have recently shown evidence that implicit memory formation persists even during adequate anaesthesia (BIS values 40-60) but not in deep anaesthesia (BIS under 40). Although it is unknown what consequences implicit memory formation has on patient outcome, deeper anaesthesia may be preferable in this respect. On the other hand, a recent study has found higher mortality associated with BIS levels below 45 [16], but the methodology of this study has been harshly criticized [17].

The rather deep anaesthesia levels found in this study may have been partially influenced by the result of our previous study, where we found that especially anaesthetics performed by inexperienced anaesthesiologists had a high percentage of time of ‘too light’ anaesthesia [1]. The discussion of those results at our department may have led to an ‘overshoot’ in the direction of deep anaesthesia in both TCI and MCI groups of this study, as also indicated by the much longer recovery times despite a similar duration of anaesthesia in both studies. The anaesthesia in both groups was characterized by relatively high concentrations of propofol and relatively small concentrations of fentanyl. However, bolus doses of fentanyl given shortly before the end of surgery may have masked a potential difference in depth of anaesthesia between both groups and thus influenced recovery time.

In summary, we have shown that a TCI system is not helpful to inexperienced anaesthesiologists as a tool to improve assessment of anaesthetic depth. However, apart from a possible usefulness to anaesthesiologists with more experience, a TCI system may reduce episodes of too light anaesthesia by providing a more stable propofol concentration.


We thank all participating colleagues and nurses for their support, and Dr Thomas Bouillon (University Hospital, Bern, Switzerland) for helpful discussions of the manuscript. The study was supported solely by departmental funding. The TCI system was kindly supplied by Fresenius-Kabi, Bad Homburg, Germany.


1. Hadzidiakos D, Nowak A, Laudahn N, Baars J, Herold K, Rehberg B. Subjective assessment of depth of anaesthesia by experienced and inexperienced anaesthetists. Eur J Anaesthesiol 2006: 1-8.
2. Servin FS. TCI compared with manually controlled infusion of propofol: a multicentre study. Anaesthesia 1998; 53(Suppl 1): 82-86.
3. Ecoffey C, Viviand X, Billard V et al. Target controlled infusion (TCI) anesthesia using propofol. Assessment of training and practice in the operating room. Ann Fr Anesth Reanim 2001; 20: 228-245.
4. Passot S, Servin F, Allary R et al. Target-controlled versus manually-controlled infusion of propofol for direct laryngoscopy and bronchoscopy. Anesth Analg 2002; 94: 1212-1216, table.
5. Passot S, Servin F, Pascal J, Charret F, Auboyer C, Molliex S. A comparison of target- and manually controlled infusion propofol and etomidate/desflurane anesthesia in elderly patients undergoing hip fracture surgery. Anesth Analg 2005; 100: 1338-1342, table.
6. Russell D, Wilkes MP, Hunter SC, Glen JB, Hutton P, Kenny GN. Manual compared with target-controlled infusion of propofol. Br J Anaesth 1995; 75: 562-566.
7. Hu C, Horstman DJ, Shafer SL. Variability of target-controlled infusion is less than the variability after bolus injection. Anesthesiology 2005; 102: 639-645.
8. Schnider TW, Minto CF, Gambus PL et al. The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology 1998; 88: 1170-1182.
9. Schnider TW, Minto CF, Shafer SL et al. The influence of age on propofol pharmacodynamics. Anesthesiology 1999; 90: 1502-1516.
10. Smith WD, Dutton RC, Smith NT. Measuring the performance of anesthetic depth indicators. Anesthesiology 1996; 84: 38-51.
11. Shafer SL, Varvel JR, Aziz N, Scott JC. Pharmacokinetics of fentanyl administered by computer-controlled infusion pump. Anesthesiology 1990; 73: 1091-1102.
12. Breslin DS, Mirakhur RK, Reid JE, Kyle A. Manual versus target-controlled infusions of propofol. Anaesthesia 2004; 59: 1059-1063.
13. Struys M, Versichelen L, Rolly G. Propofol target-controlled infusion in clinical practice. Acta Anaesthesiol Belg 1997; 48: 207-211.
14. Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 836-847.
15. Iselin-Chaves IA, Willems SJ, Jermann FC, Forster A, Adam SR, Van der LM. Investigation of implicit memory during isoflurane anesthesia for elective surgery using the process dissociation procedure. Anesthesiology 2005; 103: 925-933.
16. Monk TG, Saini V, Weldon BC, Sigl JC. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg 2005; 100: 4-10.
17. Drummond JC, Patel PM. Editorial Board reproached for publication of BIS-mortality correlation. Anesth Analg 2005; 101: 1238-1239.


© 2007 European Society of Anaesthesiology