Predictive performance of computer-controlled infusion of remifentanil during propofol/remifentanil anaesthesia

Background. The predictive performance of the available pharmacokineticparameter sets for remifentanil, when used for target-controlledinfusion (TCI) during total i.v. anaesthesia, has not been determinedin a clinical setting. We studied the predictive performanceof five parameter sets of remifentanil when used for TCI ofremifentanil during propofol anaesthesia in surgical patients. Methods. Remifentanil concentration–time data that hadbeen collected during a previous pharmacodynamic interactionstudy in 30 female patients (ASA physical status I, aged 20–65 yr)who received a TCI of remifentanil and propofol during lowerabdominal surgery were used in this evaluation. The remifentanilconcentrations predicted by the five parameter sets were calculatedon the basis of the TCI device record of the infusion rate–timeprofile that had actually been administered to each individual.The individual and pooled bias [median performance error (MDPE)],inaccuracy [median absolute performance error (MDAPE)], divergenceand wobble of the remifentanil TCI device were determined fromthe pooled and intrasubject performance errors. Results. A total of 444 remifentanil blood samples were analysed.Blood propofol and remifentanil concentrations ranged from 0.5to 11 µg ml^–1 and 0.1 to 19.6 ng ml^–1respectively. Pooled MDPE and MDAPE of the remifentanil TCIdevice were –15 and 20% for the parameter set of Mintoand colleagues (Anesthesiology 1997; 86: 10–23), 1 and21%, –6 and 21%, and –6 and 19% for the three parametersets described by Egan and colleagues (Anesthesiology 1996;84: 821–33, Anesthesiology 1993; 79: 881–92, Anesthesiology1998; 89: 562–73), and –24 and 30% for the parameterset described by Drover and Lemmens (Anesthesiology 1998; 89:869–77). Conclusions. Remifentanil can be administered by TCI with acceptablebias and inaccuracy. The three pharmacokinetic parameter setsdescribed by Egan and colleagues resulted in the least biasand best accuracy. Br J Anaesth 2003; 90: 132–41     

Hypoalbuminaemia does not impair Diprifusor performance during sedation with propofol

Background. About 98% of plasma propofol is bound to albumin.We investigated if severe hypoalbuminaemia may affect the accuracyof a target-controlled infusion (TCI) device, the Diprifusor,during sedation in critically ill patients. Methods. Ten critically ill hypoalbuminaemic patients (<24g litre^–1) and 10 critically ill normoalbuminaemic patients(>32 g litre^–1) were included in this study. They underwentsedation with propofol, aimed at a Ramsey sedation score of4–5. The Diprifusor was used to achieve target propofolplasma concentrations that ranged between 0.6 and 1.5 mg litre^–1.Propofol concentration was measured by high-performance liquidchromatography 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 6 h and8 h after starting TCI. The accuracy of TCI was evaluated bycalculating performance errors [PE=100x(measured concentration–predictedconcentration)/predicted concentration], absolute and relativeindividual median performance errors (MDAPE and MDPE) and divergence(the slope of individual regression lines between PEs and time). Results. PEs [median (range)] were –7 (–65, 79)in hypoalbuminaemic patients and –2 (–53, 188) innormoalbuminaemic patients; absolute PEs were 21 (1, 79) and22 (0, 188). No significant difference was observed betweenthe two groups. MDPE, MDAPE and divergence values were alsosimilar. In most patients the accuracy of TCI increased withtime because higher PE values were observed during the first30 min. Conclusions. Hypoalbuminaemia does not affect the accuracy ofDiprifusor during sedation with propofol in critically ill patients.     

EEG-controlled closed-loop dosing of propofol in rats

Background. Based on previous pharmacokinetic and pharmacodynamicstudies, we have developed an EEG-controlled closed-loop systemfor the i.v. hypnotic agent propofol in rats. Methods. Seven adult male Sprague–Dawley rats (weight423–584 g) were included in the study. EEG was recordedwith occipito–occipital needle electrodes and the EEGpower spectrum was estimated. The median frequency (MEF) wasextracted from the power spectrum and was modified MEF (mMEF)to account for the occurrence of spikes and burst suppressionpatterns in the EEG. Propofol infusion was controlled by a model-basedadaptive control algorithm to maintain a set point of mMEF=3.0(SD 0.5) Hz. The performance of the feedback system was characterizedby the median performance error MDPE=median{(mMEF–setpoint)/set point} and the median absolute performance error(MDAPE). The effective therapeutic infusion (ETI) to maintainthe set point was determined from the resulting infusion rates. Results. In all rats a feedback period of 90 min could be performed.Mean MDPE was 1.2 (SE 0.4)% and MDAPE was 13.9 (0.3)%. The ETIwas 0.73 (SD 0.20) mg kg^–1 min^–1. Mean arterialpressure before propofol infusion was 148 (14) mm Hg, with thelowest value during closed-loop infusion being 110 (20) mm Hg. Conclusions. The feedback system presented here may be a usefultool not only for automatic drug control to maintain a definedhypnotic effect but may also be a powerful device in pharmacologicalstudies such as the determination of dose requirements or theassessment of drug–drug interactions. Br J Anaesth 2004; 92: 564–9     

Comparison of Alaris AEP index and bispectral index during propofol-remifentanil anaesthesia

Background. The Alaris AEP monitor^TM (Alaris, UK, version 1.4)is the first commercially available auditory evoked potential(AEP) monitor designed to estimate the depth of anaesthesia.It generates an ‘Alaris AEP index’ (AAI), whichis a dimensionless number scaled from 100 (awake) to 0. Thisstudy was designed to compare AAI and BIS^TM (Aspect, USA, versionXP) values at different levels of anaesthesia. Methods. Adult female patients were premedicated with diazepam0.15 mg kg^–1 orally on the morning of surgery. Electrodesfor BIS and Alaris AEP monitoring and a headphone to give auditorystimuli were applied as recommended by the manufacturers. Anaesthesiawas induced with remifentanil (0.4 µg kg^–1 min^–1)and a propofol target-controlled infusion (Diprifusor^TM TCI,AstraZeneca, Germany) to obtain a predicted concentration ofinitially 3.5 µg ml^–1. After loss of consciousnessthe patients were given 0.5 mg kg^–1 of atracurium. Aftertracheal intubation, remifentanil was given at 0.2 µgkg^–1 min^–1 and the propofol infusion was adjustedto obtain BIS target values of 30, 40, 50, and 60. AAI and BISvalues were recorded and matched with the predicted propofoleffect-site concentrations. Prediction probability was calculatedfor consciousness vs unconsciousness. Values are mean (SD). Results. Fifty female patients, 53 (15), range 18–78 yr,ASA I or II were studied. Mean values before induction of anaesthesiawere 95 (4), range 99–82 for BIS and 85 (12), range 99–55for AAI. With loss of eyelash reflex both values were significantlyreduced to 64 (13), range 83–39 for BIS (P<0.05) and61 (22), range 99–15 for AAI (P<0.05). The predictionprobability P_K for consciousness vs unconsciousness (i.e. lossof eyelash reflex) was better for BIS (P_K=0.99) than for AAI(P_K=0.79). At a BIS of 30, 40, 50, and 60 the correspondingAAI values were 15 (6), 20 (8), 28 (11), and 40 (16), and thesewere significantly different. Conclusions. During propofol-remifentanil anaesthesia a decreaseof the depth of anaesthesia as indicated by BIS monitoring isaccompanied by corresponding effects shown by the AAI. However,wide variation in the awake values and considerable overlapof AAI values between consciousness and unconsciousness, suggestsfurther improvement of the AAI system is required. Br J Anaesth 2003; 91: 336–40     

Unconscious learning during surgery with propofol anaesthesia

Background. Learning during anaesthesia has been demonstrated,but little is known about the circumstances under which it mayoccur. This study investigated the hypothesis that learningduring anaesthesia occurs during, but not before, surgical stimulation. Methods. Words were played through headphones to 64 day-surgerypatients during propofol anaesthesia. Fourteen words were playedrepeatedly (15 times) for 1 min each either before (n=32) orduring (n=32) surgical stimulation. The depth of anaesthesiawas estimated using the bispectral index^TM (BIS^TM). Heart rate,ventilatory frequency, mean arterial pressure, end-tidal carbondioxide concentration, and infusion rate of propofol were recordedat 1 min intervals during word presentation. On recovery, memorywas assessed using an auditory word stem completion test andword recognition test. Results. The mean BIS^TM, arterial pressure, end-tidal carbondioxide and heart rate during word presentation did not differbetween the groups. The infusion rate of propofol and the ventilatoryfrequency were significantly greater in the during-surgicalstimulation group. There was no evidence for explicit recallor recognition, nor of awareness during anaesthesia (medianmean-BIS^TM=38 in the before-surgical stimulation group and 42in the during-surgical stimulation group). Only patients whowere played words during surgical stimulation showed significantimplicit memory on recovery (mean score=0.08, P<0.02) However,their scores were not significantly higher than those of thebefore-surgical stimulation group (mean score=0.01). Conclusions. Learning during anaesthesia seems more likely tooccur during rather than before surgical stimulation at comparableanaesthetic depth. We hypothesize that surgical stimulationfacilitates learning during anaesthesia, independently of itseffects on anaesthetic depth. Br J Anaesth 2004; 92: 171–7     

Predicted values of propofol EC50 and sevoflurane concentration for insertion of laryngeal mask Classic and ProSeal

Background. A new laryngeal mask airway, the ProSeal^TM (PLMA),is said to be more difficult to insert than the laryngeal maskairway Classic^TM (CLMA) using propofol anaesthesia. Therefore,we expected a greater dose of propofol and sevoflurane to berequired to insert the PLMA compared with the CLMA. We determinedthe effective concentration 50% (EC₅₀) of propofol and end-tidalsevoflurane to allow insertion of the PLMA and the CLMA. Methods. Seventy-six elective female patients (aged 20–60yr and ASA I–II) were randomly assigned to one of fourgroups. Either a PLMA or a CLMA was inserted using either propofoltarget controlled infusion or sevoflurane. Both propofol andsevoflurane targets were determined with a modified Dixon’sup-and-down method. After equilibration between the predeterminedblood and effect site concentrations, which had been held steadyfor more than 10 min, LMA insertion was attempted without neuromuscularblock. Results. The predicted EC_50CLMA and EC_50PLMA for propofol were3.14 (0.33) and 4.32 (0.67) µg ml^–1. E'_CLMAand E'_PLMA of sevoflurane (mean (SD)) were 2.36 (0.22) and 2.82(0.45)% (P<0.01 and 0.05, respectively). Conclusions. The estimated concentration of propofol and thesevoflurane concentration needed to allow insertion of the ProSeal^TMare respectively 38 and 20% greater than those needed for insertionof the Classic LMA. Br J Anaesth 2004; 92: 242–5     

Accuracy of the 'Paedfusor' in children undergoing cardiac surgery or catheterization

Background. A prototype paediatric propofol target-controlledinfusion (TCI) system, the ‘Paedfusor’ has beendeveloped. This system incorporates a paediatric pharmacokineticdata set and algorithm specific for children in a Graseby 3500anaesthesia syringe driver. In this study we have evaluatedthe accuracy of the Paedfusor TCI system in children who underwenteither cardiac surgery or cardiac catheterization procedures. Methods. Twenty-nine children aged 1–15 yr were investigated.General anaesthesia was provided using propofol administeredby the Paedfusor system. Accuracy of the system was evaluatedby obtaining up to 9 arterial samples for measurement of propofolconcentration both during anaesthesia and in the recovery period.Measured arterial propofol concentrations were then comparedwith values calculated by the Paedfusor. Results. The predictive indices of median performance error(MDPE), and median absolute performance error (MDAPE) of thePaedfusor system were found to be 4.1% and 9.7%, respectivelyand the median value for wobble was 8.3%. These values are muchbetter than those found with the adult ‘Diprifusor’system. Conclusion. The Paedfusor performance was found to be withinthe accepted limits for use as a TCI system. Br J Anaesth 2003; 91; 507–13     

The suppression of spinal F-waves by propofol does not predict immobility to painful stimuli in humans

Background. The immobilizing effects of volatile anaestheticsare primarily mediated at the spinal level. A suppression ofrecurrent spinal responses (F-waves), which reflect spinal excitability,has been shown for propofol. We have assessed the concentration-dependentF-wave suppression by propofol and related it to the logisticregression curve for suppression of movement to noxious stimuliand the effect on the bispectral index^TM (BIS^TM). The predictivepower of drug effects on F-waves and BIS for movement responsesto noxious stimuli was tested. Methods. In 24 patients anaesthesia was induced and maintainedwith propofol infused by a target controlled infusion pump atstepwise increasing and decreasing plasma concentrations between0.5 and 4.5 mg litre^–1. The F-waves of the abductor hallucismuscle were recorded at a frequency of 0.2 Hz. BIS values wererecorded continuously. Calculated propofol concentrations andF-wave amplitude and persistence were analyzed in terms of apharmacokinetic–pharmacodynamic (PK/PD) model with a simplesigmoid concentration–response function. Motor responsesto tetanic electrical stimulation (50 Hz, 60 mA, 5 s, volarforearm) were tested and the EC_50tetanus was calculated usinglogistic regression. Results. For slowly increasing propofol concentrations, computerfits of the PK/PD model for the suppression by propofol yieldeda median EC₅₀ of 1.26 (0.4–2.3) and 1.9 (1.0–2.8)mg litre^–1 for the F-wave amplitude and persistence, respectively.These values are far lower than the calculated EC₅₀ for noxiouselectrical stimulation of 3.75 mg litre^–1. This differenceresults in a poor prediction probability of movement to noxiousstimuli of 0.59 for the F-wave amplitude. Conclusions. F-waves are almost completely suppressed at subclinicalpropofol concentrations and they are therefore not suitablefor prediction of motor responses to noxious stimuli under propofolmono-anaesthesia. Presented in part at the annual meeting of the American Societyof Anesthesiologists 2004 in Las Vegas.     

Propofol sparing effect of remifentanil using closed-loop anaesthesia

Background. General anaesthesia is a balance between hypnosisand analgesia. We investigated whether an increase in remifentanilblood concentration would reduce the amount of propofol requiredto maintain a comparable level of anaesthesia in 60 patientsundergoing ambulatory surgery. Methods. Patients were allocated randomly to receive remifentanilto a target blood concentration of 2 ng ml^–1 (low), 4ng ml^–1 (medium), or 8 ng ml^–1 (high), administeredby target-controlled infusion (TCI). After equilibration, propofolTCI was commenced in closed-loop control, with auditory evokedpotentials (AEPex) as the input signal, aiming for an AEPexof 35. This was to ensure a comparable and unbiased level ofanaesthesia in all patients. Results. We found a dose-dependent decrease in propofol requirementswith increasing remifentanil concentrations. The mean (95% CI)propofol target blood concentration during adequate anaesthesiawas 4.96 (3.85–6.01) µg ml^–1 in the low, 3.46(2.96–3.96) µg ml^–1 in the medium, and 3.01(2.20–3.38) µg ml^–1 in the high group. Therewas no significant difference when recovery end points wereachieved between the groups. Cardiovascular changes were moderate,but most pronounced in the high concentration group, with adecrease in heart rate of 21% compared with baseline. The meancalculated effect site propofol concentration at loss of consciousnesswas 2.08 (1.85–2.32) µg ml^–1, and at recoveryof consciousness was 1.85 (1.68–2.00) µg ml^–1. Conclusions. This study confirms a synergistic interaction betweenremifentanil and propofol during surgery, whereas the contributionof remifentanil in the absence of stimulation seems limited.In addition, our results suggest that the propofol effect siteconcentration provides a guide to the value at which the patientrecovers consciousness. Br J Anaesth 2003; 90: 623–9     

Bispectral index, serum drug concentrations and emergence associated with individually adjusted target-controlled infusions of remifentanil and propofol for laparoscopic surgery

Background. Target-controlled infusions (TCI) are used to simplifyadministration and increase precision of i.v. drugs during generalanaesthesia. However, there is a limited relationship betweenpreset targets and measured concentrations of drugs and betweenmeasured concentrations and measures of brain function, suchas the bispectral index (BIS). Methods. We set out to evaluate the performance of TCI devicesfor propofol (Diprifusor^®) and remifentanil (Remifusor,prototype), during laparoscopic cholecystectomy in 21 patients.We also checked if there was any correlation between serum concentrationsof propofol and BIS during individually adjusted anaesthesia. Results. The Diprifusor and Remifusor had a median absoluteperformance error of 60% and 25% respectively. Propofol concentrationswere underpredicted by a median of 60%, and remifentanil concentrationswere slightly overpredicted by a median of 7%. When anaesthesiawas adjusted to keep BIS values between 45 and 60, no correlationexisted between measured concentrations of propofol and thecorresponding BIS values, although both BIS and serum propofolconcentration discriminated well between the awake and asleepstates. Emergence was rapid and uneventful in all patients.Female patients had a more rapid emergence than male patients(6.6 and 11.6 min respectively). Conclusions. TCI devices for remifentanil and propofol resultin large variation in measured serum concentrations. The lackof correlation between BIS and serum concentrations of propofoladds to the debate about whether BIS measures hypnosis as agraded state during surgery. This study confirms that womenwake up faster than men, but provides no explanation for thisrepeatedly shown difference. Br J Anaesth 2003; 91: 773–80     

Entropy of EEG during anaesthetic induction: a comparative study with propofol or nitrous oxide as sole agent

Background. The search continues for an anaesthetic monitorthat can define the level of anaesthesia in an individual patientirrespective of anaesthetic agent(s) used. Studies of availablemonitors based on bispectral analysis or evoked auditory potentialsshow the complexity of the problem. We assessed a new monitor,based on the entropy of the EEG, during induction of anaesthesiawith either propofol or nitrous oxide. Methods. In an open, randomized study (two groups; n=10) ofday surgical patients, we induced loss of response with incrementalboluses of propofol. The other group was given propofol 30 mgand then increasing concentrations of nitrous oxide until lossof response. We measured entropy with the M-Entropy Module S/5^TM(Datex-Ohmeda) using forehead electrodes and recorded responseentropy (RE; including frontal electromyogram) and state entropy(SE; only the cortical EEG). Values are median (range). Results. Baseline values were RE 98 (96–100), SE 89 (87–91)and RE 98 (96–99), SE 89 (87–91) for the propofoland nitrous oxide patients, respectively. During propofol induction,both entropy indices decreased with increasing sedation, withRE 40 (23–76) and SE 34 (17–70) at loss of response.Neither RE nor SE decreased during nitrous oxide inhalation,and at loss of response using nitrous oxide, RE and SE wereunchanged at 98 (96–100) and 88 (85–91) respectively. Conclusions. The entropy monitor of anaesthetic depth showsa successive decrease with propofol but loss of consciousnesswith nitrous oxide is not associated with change in entropyindices. Br J Anaesth 2004; 92: 167–70     

Predictive performance of 'Servin's formula' during BIS-guided propofol-remifentanil target-controlled infusion in morbidly obese patients

Background. The aim of this study was to assess the predictiveperformance of ‘Servin's formula’ for bispectralindex (BIS)-guided propofol-remifentanil target-controlled infusion(TCI) in morbidly obese patients. Methods. Twenty patients (ASA physical status II–III,age 32–64 yr) undergoing bilio-intestinal bypass surgery,were recruited. Anaesthesia was induced by using a TCI of propofolwith an initial target plasma concentration of 6 µg ml^–1,then adapted to maintain stable BIS values ranging between 40and 50. A TCI of remifentanil was added to achieve pain controland haemodynamic stability. For propofol, weight was correctedas suggested by Servin and colleagues. With ideal body weight(IBW) corrected according to formula suggested by Lemmens andcolleagues. For remifentanil, weight was corrected accordingto IBW. Arterial blood samples for the determination of bloodpropofol concentrations were collected at different surgicaltimes. The predictive performance of propofol TCI was evaluatedby examining performance accuracy. Results. Median prediction error and median absolute predictionerror were –32.6% (range –53.4%; –2.5%) and33.1% (10.8%; 53.4%), respectively. Wobble median value was5.9% (2.5%; 25.2%) while divergence median value was –1.5%h^–1 (–7.7; 33.8% h^–1). Conclusion. Significant bias between predicted and measuredplasma propofol concentrations was found while the low wobblevalues suggest that propofol TCI system is able to maintainstable drug concentrations over time. As already suggested before,a computer simulation confirmed that the TCI system performancecould be significantly improved when total body weight is used.     

Optimization of desflurane administration in morbidly obese patients: a comparison with sevoflurane using an 'inhalation bolus' technique

Background. The concept of an ‘inhalation bolus’can be used to optimize inhaled drug administration. We investigatedthe depth of anaesthesia, haemodynamic stability, and recoverytime in morbidly obese patients resulting from bispectral index^TM(BIS^TM)-guided sevoflurane or desflurane administration andBIS-triggered inhalation boluses of sevoflurane or desfluranecombined with titration of remifentanil. Methods. Fifty morbidly obese patients undergoing laparoscopicgastroplasty received either BIS-guided sevoflurane or desfluraneanaesthesia in combination with a remifentanil target-controlledinfusion. Intraoperative haemodynamic stability and BIS controlwere measured. Immediate recovery was recorded. Results. Intraoperatively, the BIS was between 40 and 60 fora greater percentage of time in the sevoflurane (78 (13)% ofcase time) than in the desflurane patients (64 (14)% of casetime), owing to too profound anaesthesia in the desflurane patientsat the start of the procedure. However, fewer episodes of hypotensionwere found in the desflurane group, without the occurrence ofmore hypertensive episodes. During immediate recovery, eye opening,extubation, airway maintenance, and orientation occurred soonerin the desflurane group. Conclusions. Immediate recovery was significantly faster inthe desflurane group. Overall hypnotic controllability measuredby BIS was less accurate with desflurane. Overall haemodynamiccontrollability was better when using desflurane. Fewer episodesof hypotension were found in the desflurane group. The use ofthe inhalation bolus was found to be appropriate in both groupswithout causing severe haemodynamic side effects. Minimal BISvalues were significantly lower after a desflurane bolus. Br J Anaesth 2003; 91: 638–50     

Comparison of propofol/remifentanil and sevoflurane/remifentanil for maintenance of anaesthesia for elective intracranial surgery

Background. Propofol and sevoflurane are suitable agents formaintenance of anaesthesia during neurosurgical procedures.We have prospectively compared these agents in combination withthe short-acting opioid, remifentanil. Methods. Fifty unpremedicated patients undergoing elective craniotomyreceived remifentanil 1 µg kg^–1 followed by an infusioncommencing at 0.5 µg kg^–1 min^–1 reducing to0.25 µg kg^–1 min^–1 after craniotomy. Anaesthesiawas induced with propofol, and maintained with either a target-controlledinfusion of propofol, minimum target 2 µg ml^–1 orsevoflurane, initial concentration 2%_ET. Episodes of mean arterialpressure (MAP) more than 100 mm Hg or less than 60 mm Hg formore than 1 min were defined as hypertensive or hypotensiveevents, respectively. A surgical assessment of operating conditionsand times to spontaneous respiration, extubation, obey commandsand eye opening were recorded. Drug acquisition costs were calculated. Results. Twenty-four and twenty-six patients were assigned topropofol (Group P) and sevoflurane anaesthesia (Group S), respectively.The number of hypertensive events was comparable, whilst morehypotensive events were observed in Group S than in Group P(P=0.053, chi-squared test). As rescue therapy, more labetolol[45 (33) vs 76 (58) mg, P=0.073] and ephedrine [4.80 (2.21)vs 9.78 (5.59) mg, P=0.020] were used in Group S. Between groupdifferences in recovery times were small and clinically unimportant.The combined hourly acquisition costs of hypnotic, analgesic,and vasoactive drugs appeared to be lower in patients maintainedwith sevoflurane than with propofol. Conclusion. Propofol/remifentanil and sevoflurane/remifentanilboth provided satisfactory anaesthesia for intracranial surgery.     

Baroreflex control of heart rate during and after propofol infusion in humans

Background. This study was designed to determine cardiovagalbaroreflex gain during propofol infusion and to characterizeits recovery profile using the pharmacological and spontaneoussequence methods in 13 healthy volunteers without cardiovascularor autonomic disorders. Methods. After an 8- to 10-h fast and no premedication, measurementsof RR intervals obtained from the electrocardiogram and non-invasivebeat-to-beat systolic blood pressure (SP) were made at consciousbaseline, at 60 and 120 min after induction of general anaesthesiausing propofol, and at 20, 60, 120 and 180 min after emergencefrom anaesthesia. During propofol anaesthesia, ventilation wasmechanically controlled to maintain normocapnia and calculatedpropofol concentration was adjusted by a TCI system at 5 µgml^–1. Baroreflex responses were triggered by bolus i.v.injections of phenylephrine and nitroprusside to alter SP by15–30 mm Hg. The linear portions of the baroreflex curvesrelating RR intervals and SP by least-square regression analysiswere determined to obtain pharmacological gains. In addition,spontaneous sequence baroreflex gains were calculated from spontaneouslyfluctuating SP and RR intervals. Results. Baseline pressor and depressor test gains before propofolanaesthesia were 29.1 (SD 14.9) and 12.5 (7.8) ms mm Hg^–1,respectively. They were significantly depressed by 65–73%during propofol infusions. Similarly, baseline up- and down-sequencebaroreflex gains were 33.8 (28.9) and 27.3 (19.8) ms mm Hg^–1,respectively, and were significantly depressed by 71–87%during propofol anaesthesia. Pressor test and up-sequence baroreflexgains returned to the baseline values 20 min after emergencefrom propofol anaesthesia, but depressor test and down-sequencebaroreflex gains did not recover until 60 min after emergence. Conclusions. We conclude that heart rate responses to both loweringand elevating blood pressure were depressed by propofol anaesthesia,and 60 min was required for their full recovery after discontinuationof propofol infusion.     

Early recovery after remifentanil-pronounced compared with propofol-pronounced total intravenous anaesthesia for short painful procedures

Background. We compared recovery from high-dose propofol/low-doseremifentanil (‘propofol-pronounced’) compared withhigh-dose remifentanil/low-dose propofol (‘remifentanil-pronounced’)anaesthesia. Methods. Adult patients having panendoscopy, microlaryngoscopy,or tonsillectomy were randomly assigned to receive either propofol-pronounced(propofol 100 µg kg^–1 min^–1; remifentanil0.15 µg kg^–1 min^–1) or remifentanil-pronounced(propofol 50 µg kg^–1 min^–1; remifentanil 0.45µg kg^–1 min^–1) anaesthesia. In both groups,the procedure was started with remifentanil 0.4 µg kg^–1,propofol 2 mg kg^–1, and mivacurium 0.2 mg kg^–1.Cardiovascular measurements and EEG bispectral index (BIS) wererecorded. To maintain comparable anaesthetic depth, additionalpropofol (0.5 mg kg^–1) was given if BIS values were greaterthan 55 and remifentanil (0.4 µg kg^–1) if heartrate or arterial pressure was greater than 110% of pre-anaestheticvalues. Results. Patient and surgical characteristics, cardiovascularmeasurements, and BIS values were similar in both groups. Therewere no differences in recovery times between the groups (timeto extubation: 12.7 (4.5) vs 12.0 (3.6) min, readiness for transferto the recovery ward: 14.4 (4.4) vs. 13.7 (3.6) min, mean (SD)). Conclusions. In patients having short painful surgery, lesspropofol does not give faster recovery as long as the same anaestheticlevel (as indicated by BIS and clinical signs) is maintainedby more remifentanil. However, recovery times were less variablefollowing remifentanil-pronounced anaesthesia suggesting a morepredictable recovery. Br J Anaesth 2003; 91: 580–2     

Agitation and changes of Bispectral Index and electroencephalographic-derived variables during sevoflurane induction in children: clonidine premedication reduces agitation compared with midazolam

Background. This double-blind randomized study was undertakento assess agitation, Bispectral Index^TM (BIS^TM) and EEG changesduring induction of anaesthesia with sevoflurane in childrenpremedicated with midazolam or clonidine. Methods. Children were allocated randomly to receive rectalmidazolam 0.4 mg kg^–1 (n=20) or oral clonidine 4µg kg^–1 (n=20) as premedication. Rapid inductionof anaesthesia was achieved with inhalation of sevoflurane 8%in nitrous oxide 50%–oxygen 50%. After tracheal intubation,the children’s lungs were mechanically ventilated andthe inspired sevoflurane concentration was adjusted to achievean end-tidal fraction of 2.5%. The EEG and BIS^TM were recordedduring induction until 10 min after tracheal intubation. TheEEG was analysed using spectral analysis at five points: baseline,loss of eyelash reflex, 15 s before the nadir of the BIS^TM (BIS_nadir),when both pupils returned to the central position (immediatelybefore intubation), and 10 min after intubation. Results. Agitation was observed in 12 midazolam-treated andfive clonidine-treated patients (P=0.05). At baseline, EEG rhythmswere slower in the clonidine group. Induction of anaesthesiawas associated with similar EEG changes in the two groups, withan increase in total spectral power and a shift towards lowfrequencies; these changes were maximal around the end of thesecond minute of induction (BIS_nadir). When the pupils had returnedto the central position, fast EEG rhythms increased and BIS^TMwas higher than BIS_nadir (P<0.05). In both groups, agitationwas associated with an increase in slow EEG rhythms at BIS_nadir. Conclusions. Compared with midazolam, clonidine premedicationreduced agitation during sevoflurane induction. During inductionwith sevoflurane 8% (oxygen 50%–nitrous oxide 50%), thenadir of the BIS^TM occurred at the end of the second minuteof inhalation. Agitation was associated with a more pronouncedslowing of the EEG rhythms at BIS_nadir compared with inductionsin which no agitation was observed. The BIS^TM may not followthe depth of anaesthesia during sevoflurane induction in children. Br J Anaesth 2004; 92: 504–11     

Target-controlled infusion of propofol and remifentanil in cardiac anaesthesia: influence of age on predicted effect-site concentrations

Background. Propofol-anaesthesia administrated via target-controlledinfusion (TCI) has been proposed for cardiac surgery. Age-relatedchanges in pharmacology explain why propofol dose requirementis reduced in elderly patients. However, the Marsh pharmacokineticmodel incorporated in the Diprifusor propofol device does nottake age into account as a covariable. In the absence of depthof anaesthesia monitoring, this limitation could cause adversecardiovascular effects resulting from propofol overdose in olderpatients. We assessed the influence of age on effect-site propofolconcentrations predicted by the Diprifusor and titrated to thebispectral index score (BIS) during cardiac anaesthesia. Methods. Forty-five patients received propofol by Diprifusorand remifentanil by software including Minto model. Propofoland remifentanil effect-site concentrations were adapted toBIS (40–60) and haemodynamic profile, respectively. Theinfluence of age on effect-site concentrations was assessedby dividing patients into two groups: young (<65 yr) andelderly (     

Desflurane compared with propofol for postoperative sedation in the intensive care unit

Background. We hypothesized that emergence from sedation inpostoperative patients in the intensive care unit would be fasterand more predictable after sedation with desflurane than withpropofol. Methods. Sixty patients after major operations were allocatedrandomly to receive either desflurane or propofol. The targetlevel of sedation was defined by a bispectral index^TM (BIS^TM)of 60. All patients were receiving mechanical ventilation ofthe lungs for 10.6 (SD 5.5) h depending on their clinical state.The study drugs were stopped abruptly in a calm atmosphere withthe fresh gas flow set to 6 litres min^–1, and the timeuntil the BIS increased above 75 was measured (t_BIS75, the mainobjective measure). After extubation of the trachea, when thepatients could state their birth date, they were asked to memorizefive words. Results. Emergence times were shorter (P<0.001) after desfluranethan after propofol (25th, 50th and 75th percentiles): t_BIS75,3.0, 4.5 and 5.8 vs 5.2, 7.7 and 10.3 min; time to first response,3.7, 5.0 and 5.7 vs 6.9, 8.6 and 10.7 min; time to eyes open,4.7, 5.7 and 8.0 vs 7.3, 10.5 and 20.8 min; time to squeezehand, 5.1, 6.5 and 10.2 vs 9.2, 11.1 and 21.1 min; time to trachealextubation, 5.8, 7.7 and 10.0 vs 9.7, 13.5 and 18.9 min; timeto saying their birth date, 7.7, 10.5 and 15.5 vs 13.0, 19.4and 31.8 min. Patients who received desflurane recalled significantlymore of the five words. We did not observe major side-effectsand there were no haemodynamic or laboratory changes exceptfor a more marked increase in systolic blood pressure afterstopping desflurane. Using a low fresh gas flow (air/oxygen1 litre min^–1), pure drug costs were lower for desfluranethan for propofol (95 vs 171 Euros day^–1). Conclusions. We found shorter and more predictable emergencetimes and quicker mental recovery after short-term postoperativesedation with desflurane compared with propofol. Desfluraneallows precise timing of extubation, shortening the time duringwhich the patient needs very close attention. Br J Anaesth 2003; 90: 273–80     

Relation between fentanyl dose and predicted EC50 of propofol for laryngeal mask insertion

Background. This study sought to determine the effective concentrationfor 50% of the attempts to secure laryngeal mask insertion (predictedEC_50LMA) of propofol using a target-controlled infusion (Diprifusor^TM)and investigated whether fentanyl influenced these requiredconcentrations, respiratory rate (RR) and bispectral index (BIS). Methods. Sixty-four elective unpremedicated patients were randomlyassigned to four groups (n = 16 for each group) and given saline(control) or fentanyl 0.5, 1 or 2 µg kg^–1.Propofol target concentration was determined by a modificationof Dixon’s up-and-down method. Laryngeal mask airway insertionwas attempted without neuromuscular blocking drugs after equilibrationhad been established for >10 min. Movement was defined aspresence of bucking or gross purposeful muscular movement within1 min after insertion. EC_50LMA values were obtained by calculatingthe mean of 16 patients in each group. Results. Predicted EC_50LMA of the control, fentanyl 0.5, 1 and2 µg kg^–1 groups were 3.25 (0.20), 2.06 (0.55),1.69 (0.38) and 1.50 (0.54) µg ml^–1 respectively;those of all fentanyl groups were significantly lower than thatof control. RR was decreased in relation to the fentanyl doseup to 1 µg kg^–1. BIS values after fentanyl1 and 2 µg kg^–1 were significantly greaterthan in the control and 0.5 µg kg^–1 groups. Conclusions. A fentanyl dose of 0.5 µg kg^–1is sufficient to decrease predicted EC_50LMA with minimum respiratorydepression and without a high BIS value. Br J Anaesth 2004; 92: 238–41