Individual requirements for sedatives vary. For example, there is considerable interindividual variability in the propofol effect-site concentration (Ce) at loss of responsiveness.1–3 Consequently, standardized dosing provides inadequate medication for some patients while proving excessive in others.4 Effective titration of drug effect in individual patients is necessary to avoid potential adverse events due to inappropriate dosing.5,6
Patients given non-opioid sedatives are unlikely to experience respiratory and/or hemodynamic complications from doses that do not cause a loss of responsiveness. We have evaluated responsiveness during propofol sedation both clinically and with an automated responsiveness monitor (ARM), a novel negative feedback system for individual titration of propofol sedation.2,7 The system incorporates a handset approximately the size of a mobile phone that consists of a conveniently located thumb switch and vibrator. It is linked to a single “ear bud” headphone worn by the subject. A computerized voice asks the participant to push the button at intervals. The voice request is accompanied by vibration of the handset. The voice repeats the request four times over a 10-s period, becoming louder and more insistent with each repetition. The handset vibrates during questioning, and the vibration becomes progressively more intense with each query over the 10-s period. Pressing the button in response to any of the four queries presented in this 10-s window is considered an evidence of responsiveness and stops both the queries and vibration.
In previous studies, we showed that failure to respond to the ARM precedes potentially serious sedation-related adverse events associated with loss of responsiveness, such as apnea and hypotension, and that the ARM was not susceptible to false-positive responses.2,7 It remains unknown, however, whether loss and return of response to the ARM occurs at similar sedation levels, independent of the scheme used to titrate the drug effect. The issue is important because time invariance is one of the basic assumptions of pharmacokinetics.8 As a corollary, a clinical surrogate measure of sedation must be highly predictable. Predictability in this context means that the dynamics of a behavioral titration instrument like the ARM must be reliable and have a reasonably limited intra- and interindividual variability.
We therefore tested the hypothesis that loss and return of response to the ARM occur at similar sedation levels in individual subjects during propofol sedation. We tested our theory by increasing the propofol Ce until loss of response to the ARM was observed. After loss of response, we compared two methods of decreasing the propofol Ce: a fixed percentage decrease and a linear deramping of the Ce until the ARM response returned. Bispectral Index (BIS) of the electroencephalogram (EEG) and Ce information were used independently, as well as in combination, to characterize ARM response dynamics.
With approval of the University of California, San Francisco Human Studies Committees and written informed consent, we evaluated 21 healthy volunteers of both genders between April and May of 2000. All volunteers received propofol until they stopped responding to the ARM. Ten volunteers were studied using a fixed percentage protocol and 11 others were studied using a linear deramping protocol, as described below. Age was restricted to 20–45 yr. Volunteers fasted at least 8 h before the trial.
All standard anesthetic monitors including oscillometric blood pressure, electrocardiogram, end-tidal CO2 through a sealed anesthesia mask, and pulse oximetry (Spo2) were applied to the participating volunteers. Electrodes to capture the BIS of the EEG (A-2000 monitor, BIS 3.3 algorithm, system revision 1.07, Aspect Medical Systems, Newton, MA) were applied to the forehead according to the manufacturer’s instructions. The BIS recording began with a 2-min period of quiet relaxation with the volunteer’s eyes closed. Resistance of the BIS sensors was maintained at <5 kω throughout the study period.
A 20-gauge catheter was inserted at the antecubital fossa on the dominant arm for the propofol infusion, whereas a 20-gauge catheter was inserted into the radial artery in the contralateral arm for blood sampling. Normothermia was maintained with forced-air warming. Volunteers breathed supplemental oxygen via a sealed anesthesia mask to maintain a Spo2 more than 92%.
The ARM apparatus was strapped loosely to the dominant hand. The volunteers were trained with the ARM handset and headphone for 10–15 min before the first sedation trial. The volume of the query was adjusted to a level that they were able to hear easily. We confirmed that the volunteers responded promptly to the ARM apparatus during this prestudy period.
We used a target-controlled infusion (TCI) drug delivery system according to the method of Shafer and Gregg9 to target propofol Ce using the covariate-adjusted propofol kinetic model reported by Schnider et al.10 with a ke0 of 0.46/min.11 The performance of the system was previously evaluated under pseudosteady-state conditions.12 The drug delivery system consisted of a Harvard 22 (Harvard Clinical Technology, South Natick, MA) electronic syringe pump, which could be commanded by a host system (Pentium II 450 MHz microprocessor-based system) through an RS232 serial communication port. A customized software platform, written by Scott Laboratories (Lubbock, TX), was used to drive the pump.
Fixed Percentage Protocol
Ten volunteers were studied six times each during administration of propofol using an infusion “ramp” with a slope adjusted to change the propofol Ce with a rate of 0.6 μg · mL−1 · min−1. The infusion ramp was maintained until the volunteers lost response to the ARM apparatus, because this is determined by a nonresponse to a 10-s-long query period. Subsequently, propofol Ce was decreased by one of six randomly ordered percentages: 20%, 30%, 40%, 50%, 60%, and 70%. The target propofol Ce was maintained stable at this level for 6 min (Ce plateau) before the infusion was stopped. Sedation was evaluated at 1-min intervals during this steady-state period. After a recovery period of at least 15 min, and a reduction in the predicted Ce to ≤0.5 μg/mL, the next trial began, again consisting of an infusion ramp followed by a decrease of a different percentage. Each participant experienced each percentage decrease (Fig. 1).
A set of 11 different volunteers was studied three times each during administration of propofol using an infusion ramp with a slope adjusted to change propofol Ce at a rate of 0.5 μg · mL−1 · min−1. The infusion was maintained until the volunteers lost response to the ARM apparatus, because this is determined by a nonresponse during a 10-s query period. Subsequently, the Ce of propofol was decreased by one of three randomly ordered slopes adjusted to change Ce by 0.1, 0.2, or 0.3 μg · mL−1 · min−1. The Ce of propofol was decreased until three successive positive responses to the ARM apparatus occurred. After the third ARM response, the propofol Ce was maintained at that level for 6 min (Ce plateau) before the infusion was stopped. At the end of each steady-state period, the sedation level was evaluated. As described above, after a recovery period of at least 15 min and a reduction in the predicted Ce to ≤0.5 μg/mL, the next trial began to test a different deramping rate. Each participant experienced each deramping rate decrease in propofol Ce (Fig. 1).
Demographic and morphometric characteristics of the volunteers were recorded. All standard physiologic values were downloaded and recorded to an automated data acquisition system for off-line analysis. These included heart rate, arterial blood pressure, respiratory rate, end-tidal CO2, and Spo2. BIS and all standard anesthetic monitoring data, except for blood pressure, were recorded at 15-s intervals.
Each ARM test lasted for 10 s, contained four individual queries, and was repeated with a rate of four tests per minute, i.e., after each completed (four queries) ARM test there was a 5-s interval before the initiation of the next one. Failure to press the button in response to all four queries that were presented during a 10-s ARM test was considered a nonresponse. On the contrary, a positive response to any of the four queries was considered a positive response to the ARM. Both the positive and the negative responses to each ARM test were recorded approximately 5 s after the end of each test, i.e., just before the initiation of the next test. Thus, the resolution of each ARM test was approximately 15 s. A detailed diagram in Figure 1 depicts the timing characteristics of the ARM test.
An arterial blood sample for propofol determination was obtained at the beginning and the end of each Ce plateau in both protocols to document the presence of steady state. The samples were analyzed using a high performance liquid chromatography assay modified from the method of Plummer.13 This method has a coefficient variation of 4.1% at a propofol plasma level of 2 μg/mL.
Sedation was assessed clinically using the Observer’s Assessment of Alertness/Sedation score (OAA/S). The OAA/S score consists of four components. As described by Chernik et al.,14 we summed the component scores. The presence of consciousness was defined as an OAA/S score higher than 10 (of 20). The score was applied every 1 min at the Ce plateau during the fixed percentage protocol, and only once at the end of each Ce plateau during the deramping protocol. An attempt was made to evaluate sedation at the end of an ARM test so as to minimize any interference with the ARM function.
Demographic and morphometric data were averaged across volunteers and presented for each protocol separately.
We have previously shown1 that when propofol Ce is increasing at a rate between 0.1 and 0.9 μg · mL−1 · min−1, a ke0 of 0.17/min (tpeak = 2.7 min) more accurately reflects the plasma-effect-site equilibration than the previously reported value of 0.46/min (tpeak = 1.7 min).11 Predicted Ce reported here are thus based on a ke0 of 0.17/min.
The assumption of steady-state concentration at each plateau (Ce plateau) was tested using the Bland and Altman method.15 The difference in the measured plasma concentration (Cp) between the beginning (Cp-start) and the end (Cp-end) of each concentration plateau was presented as a function of the Cp-start and Cp-end average. Differences between the two protocols were assessed using unpaired t-test.
The accuracy of the TCI system was evaluated by calculating the median performance error (MDPE) and the median absolute performance (MDAPE) error for each protocol separately, as previously proposed.16 First, for each blood sample the performance error (PE) was calculated as:
where Cm and Cp are the measured and predicted plasma propofol concentrations, respectively. Subsequently, the MDPE and the MDAPE were calculated for each subject separately. The median (range) values for MDPE and MDAPE were reported for each protocol separately. In addition, MDPE and MDAPE values at the beginning and the end of the Ce plateau were presented in a graph for each protocol separately.
Logistic regression was used to estimate the probability of response (squeezing) to the ARM device as a function of the predicted propofol Ce. Each response to the ARM was given a score of 1, and each nonresponse to the ARM was given a score of 0. The probability of responding to ARM (PCe) was then calculated as:
where Ce50, ARM is the predicted propofol Ce associated with a 50% probability for response to the ARM and γCe, ARM is the steepness of the Ce versus probability relationship (also termed the “Hill coefficient”). The parameters Ce50, ARM and γCe, ARM were estimated using nonlinear mixed effects modeling (NONMEM V, GloboMax LLC, Hanover, MD). Interindividual variability was permitted and assumed to be log-normally distributed. Residual intraindividual error was assumed to be additive. All data from both protocols (7185 data points) were used for this analysis. The effect of protocol type (fixed percentage or deramping) on the Ce50, ARM was tested by permitting different Ce50, ARM values when the protocol type was added to the model as a covariate. A decrease in the objective function of the complete model more than 3.84 points indicated a significant effect of the protocol on the ARM response dynamics.
The same analysis, as above, was used to estimate the probability of a response to the ARM as a function of BIS (PBIS). In this analysis, instead of the BIS value the BIS difference from the baseline, determined as BIS effect (BISeffect = 100 − BIS), was used. The BISeffect50, ARM, indicating the BISeffect associated with a 50% probability for response to ARM and the γBISeffect, ARM, were estimated using NONMEM. The effect of the protocol type on BISeffect50, ARM was evaluated by incorporating the protocol type into the model as a covariate.
The combined effect of propofol Ce and BISeffect on the probability for ARM response (PBIS/Ce) was also evaluated. An independent variable, IND, was calculated as:
IND = 100 − BIS + Ce × θ
and logistic regression was used to model the probability of response to ARM (PBIS/Ce) as a function of IND:
The combined effect of propofol Ce and BISeffect, as expressed above, was considered significant if the minimum objective function (−2 log likelihood, −2LL) of the model decreased by at least 3.84 points for each parameter added to the model.
Finally, in the complete model, the combined effect of propofol Ce and BISeffect was evaluated after expressing the probability of ARM response (PBIS/Ce) as the sum of the probability fractions derived from the effects of propofol Ce (PCe) and BISeffect (PBIS):
PBIS/Ce = FBIS × PBIS + (1 − FBIS) × PCe,
where FBIS is the fraction of the combined probability PBIS/Ce that is determined by BISeffect. The combined effect of propofol Ce and BISeffect, expressed as above, was considered significant if it produced a decrease by 3.84 points in the −2LL of the model, for each added parameter.
The probability of response to the ARM was modeled as a function of the sedation level (OAA/S score) during the Ce plateau periods in both protocols, using logistic regression. The OAA/S50, ARM and γOAA/S, ARM were also estimated by NONMEM. A total of 461 data points were used for this analysis.
Bootstrap resampling with replacement was used to determine 95% confidence intervals for the P50 estimates of Ce50, BISeffect 50, and OAA/S50. One thousand bootstrap samples (simple random samples of size 21 with replacement) were created from each of the three (i.e., Ce50, BISeffect 50, and OAA/S50) samples originally estimated by NONMEM, as described above. Confidence limits for each P50 were taken as the 2.5th and 97.5th percentiles of the respective bootstrap sample distribution.
The ability of each tested parameter or combination of parameters to predict the observed response to the ARM device was investigated by calculating the average of probabilities (with and without rounding each probability to the closest integer, i.e., 1 or 0) for all the individual observations for each of the tested parameters, or combination of parameters, separately.
The median (range) propofol Ce at the first loss and first recovery of ARM response for each of the individual sedation trials (i.e., fixed percentages or deramping rates) were presented for each protocol and each individual volunteer, separately. In addition, the intra- and interindividual variability, calculated as the coefficient of variation (%), in the Ce and BIS at the first loss and first recovery of ARM response were presented for each protocol separately in a tabular and graph formats.
The volunteers participating in the fixed percentage protocol (n = 10) were 32.4 ± 7.4 yr old, weighed 69.0 ± 7.6 kg, and were 169.7 ± 8.1 cm tall. The volunteers participating in the deramping protocol (n = 11) were 33.4 ± 5.8 yr old, weighed 69.9 ± 12 kg, and were 172.4 ± 9.5 cm tall. All physiology remained within normal limits during both sedation protocols.
The measured arterial propofol Cp at the beginning (Cp-start) and the end (Cp-end) of the Ce plateau differed only by 0.05 ± 0.39 μg/mL (Fig. 2). Arterial propofol Cp obtained at steady state was used to assess the performance of the TCI. The results of this analysis are presented in Table 1 and Figure 2. Timing of the sampling has not been shown to be a confounding factor regarding the performance of the TCI system.
The Ce50 of propofol for responding to the ARM was 1.73 (95% confidence interval: 1.55–2.10) μg/mL, whereas the BISeffect50 was 24.9 (23.0–28.7), corresponding to a BIS value of approximately 75 (Table 2, Fig. 3). The OAA/S score associated with a 50% probability for ARM response was 12.5/20 (12.0–13.4). Figure 3 presents the probability of responding to the ARM as a function of Ce, BISeffect, and OAA/S score.
Propofol Ce and BIS independently predicted the observed ARM response with a probability of 0.82 and 0.84, respectively, whereas their combination increased that probability to 0.85. The model that defined the probability of the observed ARM responses as the sum of the fractional probabilities determined by propofol Ce and BIS (PBIS/Ce = PCe+ PBIS, Table 3, row D) demonstrated the lowest minimum objective function (-2LL = 4796.38), when compared with other models that used only Ce (PCe = 0.82, −2LL = 5895.57, Table 3, row A) or BIS (PBIS = 0.84, −2LL = 5474.31, Table 3, row B), as predictor variables, and the model that combined Ce and BIS in the form of an independent variable IND = 100 − BIS + Ce × θ (PBIS/Ce = 0.84, −2LL = 5101.49, Table 3, row C). In the PBIS/Ce fractional probability model, the contribution of PBIS to the overall probability was estimated to be 63%.
Table 4 presents the intra- and interindividual variability, regarding the propofol Ce and BIS when the first loss and recovery of ARM response occurred. Figure 4 presents the actual Ce values and its intraindividual variability at the above end points.
Using an effect-site TCI system, we have shown that ARM can titrate propofol sedation in a reproducible manner over time even in nonsteady-state conditions. Loss and return of response to the ARM occurred at similar Ce of propofol in individual subjects and with a reasonable interindividual variability during the fixed percentage and deramping protocols. Interestingly, when BIS and propofol Ce were independently used to characterize ARM response, the latter was not influenced by the protocol titration scheme.
Our TCI system performed well during the fixed percentage protocol, and its performance was similar to what we have previously demonstrated,12 using the same kinetic model for propofol, developed by Schnider et al.,10 and subsequently validated by Doufas et al.1 The system did not perform as well during the deramping protocol but was still in an acceptable range. The replacement of our original ke0 value of 0.46/min10 with the value of 0.17/min is justified by two arguments: (a) the ke0 of 0.17/min has been estimated in a study of a young healthy population1 with similar characteristics as our present participants, using similar infusion designs as in the current trial, and (b) the ke0 value of 0.17/min was derived from a study,1 which prospectively validated the propofol kinetic set developed by Schnider et al.10
In one study, adequately characterized sedation end points like loss and recovery of consciousness (defined by the ability to respond to verbal command) occurred at similar propofol Ce in each subject, despite the presence of a large interindividual variability.3 In this study, responsiveness to the ARM apparatus demonstrated similarly predictable dynamics with a propofol Ce50, ARM of 1.73 μg/mL (95% confidence interval: 1.55–2.10, intraindividual variability = 19%). This Ce value is comparable with those we have found previously in pseudosteady-state (1.6 μg/mL)2 and nonsteady-state (1.76 μg/mL)1 conditions and are approximately 0.5–1 μg/mL less than the Ce at loss of response to verbal command2,3,7,17,18 or tactile stimulation.1 Furthermore, the BIS50 for ARM response was much higher than the BIS level previously associated with loss of response to verbal7,18,19 or tactile1 stimulation. BIS is highly correlated with propofol Ce,20 and it has been shown to predict clinical sedation (OAA/S score) and loss of responsiveness comparably well21 with, if not slightly better20 than, the Ce. In addition, the P50 of the OAA/S score for responding to the ARM was 12.5/20 well above the threshold for loss of consciousness (10/20). This OAA/S50, ARM value is very similar to what we demonstrated previously under pseudosteady-state conditions,2 and it was not influenced by the applied protocol scheme.
Thus, this study supports our previous finding that loss of response to the ARM tends to precede loss of consciousness during propofol-only sedation and might be used as a titration instrument when the actual loss of responsiveness is not a desired end point. Although it is difficult to compare the arousing potential between an ARM and a human-based (OAA/S scale) sedation assessment, the results of the present and previous2 trials support a preponderance of the latter. Nevertheless, we cannot exclude the possibility of an interaction between these two types of stimuli during our trial.
Propofol Ce and BIS predicted the observed ARM responses with a probability of 0.82 and 0.84, respectively. This probability increased to 0.85 when information from both Ce and BIS (fractional probabilities 0.37:0.63) were simultaneously used to predict ARM response. This was highly statistically significant, in regard to model improvement (Table 3), and reflects the importance of combining BIS and Ce information when attempting to characterize ARM response as an independent sedation end point. This result of our modeling approach provides two important insights: (a) it strengthens the evidence that the ARM relays information about a real drug effect, and (b) it questions the interchangeable use of Ce and BIS, two highly correlated but different, pharmacodynamic quantities, in characterizing drug effect.
Targeting the Ce, rather than the Cp propofol concentration, has been associated with fewer hemodynamic and respiratory consequences.22,23 However, a relatively rapid increase of Ce always entails the risk of concentration overshoot and oversedation when the desired end point is reached, with potential adverse physiological consequences. An almost reflexive, preferably clinical, monitoring system is necessary to prevent oversedation and enhance safety when the sedative effect is progressing quickly. Patient-maintained sedation systems that use a Ce-driven propofol TCI have managed to provide safe sedation mainly by pursuing a slow, stepwise increase in the Ce.24 Nonetheless, a certain number of patients who used patient-maintained sedation were able to deliberately oversedate themselves, reaching a potentially unsafe sedation depth.25 As expected, the use of a Ce ramp (0.5–0.6 μg · mL−1 · min−1) in our study not only increased the speed of sedation induction but also resulted in a relative Ce overshoot after discontinuation of the infusion when loss of response to the ARM occurred. This Ce overshoot led to oversedation in certain instances, which was never associated with severe adverse effects, such as apnea or hypotension.
We conclude that loss and return of response to the ARM occurs at similar sedation levels in individuals, even though there is considerable variability among individuals. Reproducible ARM dynamics compares favorably with clinical and EEG sedation end points and suggest that the ARM can be used as an independent instrumental guide of propofol effect. However, the wider applicability of the ARM in clinical settings, which are usually compounded by multiple stimulating events and/or drug effects, remains to be tested.
The authors greatly appreciate the assistance of Randy Hickle, MD, Brett L. Moore, BS, and Jason Derouen, BS (all from Scott Laboratories, Inc., Lubbock, TX).
1. Doufas AG, Bakhshandeh M, Bjorksten AR, Shafer SL, Sessler DI. Induction speed is not a determinant of propofol pharmacodynamics. Anesthesiology 2004;101:1112–21
2. Doufas AG, Bakhshandeh M, Bjorksten AR, Greif R, Sessler DI. Automated responsiveness test (ART) predicts loss of consciousness and adverse physiologic responses during propofol conscious sedation. Anesthesiology 2001;94:585–92
3. Iwakiri H, Nishihara N, Nagata O, Matsukawa T, Ozaki M, Sessler DI. Individual effect-site concentrations of propofol are similar at loss of consciousness and at awakening. Anesth Analg 2005;100:107–10
4. Smith I, Monk TG, White PF, Ding Y. Propofol infusion during regional anesthesia: sedative, amnestic, and anxiolytic properties. Anesth Analg 1994;79:313–9
5. Bailey PL, Pace NL, Ashburn MA, Moll JW, East KA, Stanley TH. Frequent hypoxemia and apnea after sedation with midazolam and fentanyl. Anesthesiology 1990;73:826–30
6. Malviya S, Voepel-Lewis T, Tait AR. Adverse events and risk factors associated with the sedation of children by nonanesthesiologists. Anesth Analg 1997;85:1207–13
7. Doufas AG, Bakhshandeh M, Haugh GS, Bjorksten AR, Greif R, Sessler DI. Automated responsiveness test and bispectral index monitoring during propofol and propofol/N2
O sedation. Acta Anaesthesiol Scand 2003;47:951–7
8. Schwilden H, Schuttler J. Target controlled anaesthetic drug dosing. Handb Exp Pharmacol 2008;425–50
9. Shafer SL, Gregg KM. Algorithms to rapidly achieve and maintain stable drug concentrations at the site of drug effect with a computer-controlled infusion pump. J Pharmacokinet Biopharm 1992;20:147–69
10. Schnider TW, Minto CF, Gambus PL, Andresen C, Goodale DB, Shafer SL, Youngs EJ. The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology 1998;88:1170–82
11. Schnider TW, Minto CF, Shafer SL, Gambus PL, Andresen C, Goodale DB, Youngs EJ. The influence of age on propofol pharmacodynamics. Anesthesiology 1999;90:1502–16
12. Doufas AG, Bakhshandeh M, Bjorksten AR, Greif R, Sessler DI. A new system to target the effect-site during propofol sedation. Acta Anaesthesiol Scand 2003;47:944–50
13. Plummer GF. Improved method for the determination of propofol in blood by high-performance liquid chromatography with fluorescence detection. J Chromatogr 1987;421:171–6
14. Chernik DA, Gillings D, Laine H, Hendler J, Silver JM, Davidson AB, Schwam EM, Siegel JL. Validity and reliability of the Observer’s Assessment of Alertness/Sedation Scale: study with intravenous midazolam. J Clin Psychopharmacol 1990;10:244–51
15. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10
16. Varvel JR, Donoho DL, Shafer SL. Measuring the predictive performance of computer-controlled infusion pumps. J Pharmacokinet Biopharm 1992;20:63–94
17. Chortkoff BS, Eger EI II, Crankshaw DP, Gonsowski CT, Dutton RC, Ionescu P. Concentrations of desflurane and propofol that suppress response to command in humans. Anesth Analg 1995;81:737–43
18. Milne SE, Troy A, Irwin MG, Kenny GN. Relationship between bispectral index, auditory evoked potential index and effect-site EC50
for propofol at two clinical end-points. Br J Anaesth 2003;90:127–31
19. Irwin MG, Hui TW, Milne SE, Kenny GN. Propofol effective concentration 50 and its relationship to bispectral index. Anaesthesia 2002;57:242–8
20. Struys MM, Jensen EW, Smith W, Smith NT, Rampil I, Dumortier FJ, Mestach C, Mortier EP. Performance of the ARX-derived auditory evoked potential index as an indicator of anesthetic depth: a comparison with bispectral index and hemodynamic measures during propofol administration. Anesthesiology 2002;96:803–16
21. Struys MM, Vereecke H, Moerman A, Jensen EW, Verhaeghen D, De Neve N, Dumortier FJ, Mortier EP. Ability of the bispectral index, autoregressive modelling with exogenous input-derived auditory evoked potentials, and predicted propofol concentrations to measure patient responsiveness during anesthesia with propofol and remifentanil. Anesthesiology 2003;99:802–12
22. Kazama T, Ikeda K, Morita K, Kikura M, Doi M, Ikeda T, Kurita T, Nakajima Y. Comparison of the effect-site k(eO)s of propofol for blood pressure and EEG bispectral index in elderly and younger patients. Anesthesiology 1999;90:1517–27
23. Struys MM, De Smet T, Depoorter B, Versichelen LF, Mortier EP, Dumortier FJ, Shafer SL, Rolly G. Comparison of plasma compartment versus two methods for effect compartment—controlled target-controlled infusion for propofol. Anesthesiology 2000;92:399–406
24. Chapman RM, Anderson K, Green J, Leitch JA, Gambhir S, Kenny GN. Evaluation of a new effect-site controlled, patient-maintained sedation system in dental patients. Anaesthesia 2006;61:345–9
© 2009 International Anesthesia Research Society
25. Murdoch JA, Grant SA, Kenny GN. Safety of patient-maintained propofol sedation using a target-controlled system in healthy volunteers. Br J Anaesth 2000;85:299–301