British Journal of Anaesthesia 92 (4): 504±11 (2004)
DOI: 10.1093/bja/aeh093 Advance Access publication February 20, 2004
Agitation and changes of Bispectral IndexTM and
electroencephalographic-derived variables during sevo¯urane
induction in children: clonidine premedication reduces agitation
compared with midazolam²
I. Constant1*, Y. Leport1, P. Richard1, M.-L. Moutard2 and I. Murat1
1
Service d'AnestheÂsie-ReÂanimation, Hopital d'enfants Armand Trousseau, Assistance-Publique, HoÃpitaux de
Paris, Paris VI University, France. 2Service de Neurologie peÂdiatrique, Hopital Saint Vincent de Paul,
Assistance-Publique, HoÃpitaux de Paris, France*
*Corresponding author. E-mail: [email protected]
Background. This double-blind randomized study was undertaken to assess agitation,
Bispectral IndexÔ (BISÔ) and EEG changes during induction of anaesthesia with sevo¯urane in
children premedicated with midazolam or clonidine.
Methods. Children were allocated randomly to receive rectal midazolam 0.4 mg kg±1 (n=20)
or oral clonidine 4 mg kg±1 (n=20) as premedication. Rapid induction of anaesthesia was
achieved with inhalation of sevo¯urane 8% in nitrous oxide 50%±oxygen 50%. After tracheal
intubation, the children's lungs were mechanically ventilated and the inspired sevo¯urane concentration was adjusted to achieve an end-tidal fraction of 2.5%. The EEG and BISÔ were
recorded during induction until 10 min after tracheal intubation. The EEG was analysed using
spectral analysis at ®ve points: baseline, loss of eyelash re¯ex, 15 s before the nadir of the BISÔ
(BISnadir), when both pupils returned to the central position (immediately before intubation),
and 10 min after intubation.
Results. Agitation was observed in 12 midazolam-treated and ®ve clonidine-treated patients
(P=0.05). At baseline, EEG rhythms were slower in the clonidine group. Induction of anaesthesia was associated with similar EEG changes in the two groups, with an increase in total spectral
power and a shift towards low frequencies; these changes were maximal around the end of the
second minute of induction (BISnadir). When the pupils had returned to the central position,
fast EEG rhythms increased and BISÔ was higher than BISnadir (P<0.05). In both groups, agitation was associated with an increase in slow EEG rhythms at BISnadir.
Conclusions. Compared with midazolam, clonidine premedication reduced agitation during
sevo¯urane induction. During induction with sevo¯urane 8% (oxygen 50%±nitrous oxide 50%),
the nadir of the BISÔ occurred at the end of the second minute of inhalation. Agitation was
associated with a more pronounced slowing of the EEG rhythms at BISnadir compared with
inductions in which no agitation was observed. The BISÔ may not follow the depth of anaesthesia during sevo¯urane induction in children.
Br J Anaesth 2004; 92: 504±11
Keywords: anaesthetic techniques, induction; anaesthetics volatile, sevo¯urane; brain,
electroencephalography; monitoring, bispectral index; complications, agitation; children
Accepted for publication: November 14, 2003
Sevo¯urane is becoming very popular for inhalational
induction of anaesthesia in children. However, we demonstrated that agitation occurred in 60% of children during
sevo¯urane induction.1 Clonidine given orally at a dose of
4 mg kg±1 is an effective premedicant in children.2 It has
sedative properties and prevents emergence agitation after
sevo¯urane anaesthesia.3 We speculated that premedication
²
Presented in part at the annual meeting of the European Society of
Anaesthesiologists, Nice, France, April 6±9, 2002.
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2004
BIS, EEG and agitation with sevo¯urane
with clonidine may decrease the incidence of agitation
during sevo¯urane induction in children. In an attempt to
understand more precisely the phenomenon commonly
observed during induction, we recorded the electroencephalogram (EEG) continuously.
The second goal of this study was to investigate the
general agreement of Bispectral IndexÔ (BISÔ) measurements with the frequency of the EEG during sevo¯urane
induction in children. The BISÔ has been shown to be a
quanti®able measure of the hypnotic effect of anaesthetic
drugs on the central nervous system.4± 6 This EEG-based
monitor of anaesthetic effect integrates various EEG
descriptors into a single dimensionless, empirically calibrated number. Although a BISÔ/anaesthetic concentration
relationship similar to that seen in adult patients6 may exist
in paediatric patients during sevo¯urane sedation,7 there are
no data describing the BISÔ changes during sevo¯urane
induction in children.
Therefore, this double-blinded randomized study was
designed to assess clinical, EEG and BISÔ changes during
sevo¯urane induction in children premedicated with midazolam or clonidine.
of the lungs was manually assisted before tracheal
intubation at a rate close to 20 min±1. Tracheal intubation
was performed using a preformed oral cuffed tracheal tube
after placement of an i.v. line, and just after visualization of
the pupils in the central position. After placement of the
tracheal tube, the lungs were mechanically ventilated with a
tidal volume of 10 ml kg±1 at the rate of 20 min±1 (Aestiva;
Datex-Ohmeda, Helsinki, Finland). The inspired concentration of sevo¯urane was reduced to obtain an end-tidal
fraction (SD) of 2.5 (0.1)% and this expired concentration
was maintained for 10 min. Surgery started after completion
of the study, and anaesthesia was maintained with
sevo¯urane in all children.
The induction period until 10 min after placement of
tracheal tube was videotaped and reviewed by a blinded,
independent paediatric neurologist (MLM) in order to assess
the presence of agitation. Agitation was de®ned as involuntary movements of upper and/or lower limbs during
induction and before tracheal intubation that excluded
involuntary movements related to mask placement or
venous puncture. The time of the beginning and end of
the agitation phase was recorded.
Methods
Analysis of BISÔ data
Patients and study design
The study was approved by our institutional ethics committee (CCPPRB Saint-Antoine, Paris, France) and written
informed consent was obtained from parents. We studied 40
children ASA PS I aged 2±10 yr who were scheduled for
elective tonsillectomy. They were allocated randomly to
one of two treatment groups using a prospective randomized
design. Patients and investigators were blinded to the
premedicant and the code was broken after completion of
the study. Patients in the clonidine group received oral
clonidine 4 mg kg±1 with a small amount of water (total
volume of 2 ml) 1 h before induction and 2 ml of saline by
the rectal route 30 min before induction. Patients in the
midazolam group received 2 ml of water administered by
the oral route 1 h before induction and our standard
premedication with intrarectal midazolam 0.4 mg kg±1 with
a small amount of saline (total volume of 2 ml) 30 min
before induction.
Anaesthesia was induced with sevo¯urane 8% in a
mixture of oxygen and nitrous oxide (50±50), and this
inspired concentration was maintained up to tracheal
intubation. The circuit was primed with sevo¯urane 8%
and mask induction was performed using an open circuit at
high fresh gas ¯ow (8 litres min±1). Expired gases and
oxygen saturation were recorded continuously (Capnomac
Ultima; Datex, Instrument Corporation, Helsinki, Finland).
The time to loss of eyelash re¯ex and the time for the pupils
to return to the central position were recorded. In addition,
the degree of airway obstruction and the time when the oral
airway was inserted if required were recorded. Ventilation
The disposable BisSensor (Aspect Medical Systems,
Newton, MA, USA) was applied to the forehead of each
child before induction of anaesthesia and was connected to
an Aspect Medical Systems Model A-2000 BIS Monitor
2.10. The adult sensor was used for all patients. The skin
was prepared to ensure low impedance and a good quality of
signal. The smoothing window was set at 15 s and the values
of BISÔ were downloaded online onto a computer for later
analysis (sample time 5 s). BISÔ values at loss of eyelash
re¯ex and when the pupils returned to the central position
were noted. The other values of BISÔ were extracted
retrospectively from the ®le downloaded for each patient, as
follows: the BISÔ at baseline was calculated as the mean of
BISÔ values recorded during the 60 s period preceding
induction, the nadir of the BISÔ was the lowest value of
BISÔ detected from the start of induction to the time
when the pupils returned to the central position (the
corresponding time was also noted), the BISÔ at 2.5
(0.1)% end-tidal sevo¯urane was calculated as the mean of
BISÔ values recorded during the tenth minute after tracheal
intubation.
Analysis of EEG data
Independently from the BISÔ monitor, the EEG was
recorded continuously in awake children before anaesthesia,
during induction until 10 min after tracheal intubation. A
single channel was recorded from Ag±AgCl electrodes
placed on the forehead and left mastoid, with the right
mastoid as the common electrode. The electrode impedance
was checked automatically and maintained at less than 5 kW.
505
Constant et al.
Fig 1 Example of time course of BISÔ during sevo¯urane induction in a 5-yr-old child premedicated with midazolam. The agitation period is shown.
The EEG signal was acquired (256 Hz) on a microcomputer
(Compaq) using the Brain-QuickÔ program stem II,
(Micromed, Merignac, France) and stored on hard disk.
Bandpass ®lters were set at 0.5±30 Hz.8
The EEG was analysed in the frequency domain. Spectral
analysis of the EEG signal was performed using fast Fourier
transformation (FFT) (Acqknowledge v3.25; Biopac
Systems, Santa Barbara, CA, USA) on 8-s epochs.9 The
following variables were calculated:10 total spectral power
(TSP), de®ned as the area under the curve of the spectrum
(mV2), spectral edge frequency 95 (SEF95 = the frequency
below which 95% of the EEG power is located) and median
power frequency (MPF = the frequency below which 50%
of the EEG power is located). Spectral bands of 0±4 Hz
(delta), 4±8 Hz (theta), 8±13 Hz (alpha) and 13±30 Hz (beta)
were analysed and the power of the spectral bands was
calculated and expressed as a percentage of total spectral
power. The delta ratio (ratio of the power in the 8±30 Hz
band to power in the 0±4 Hz band) was calculated. These
EEG-derived variables were calculated at baseline, at loss of
eyelash re¯ex, when the pupils returned to the central
position, at 2.5 (0.1)% end-tidal sevo¯urane and 15 s before
the occurrence of the nadir of BISÔ (BISnadir).
Statistical analysis
A previous study showed that 60% of children premedicated
with midazolam demonstrated agitation during sevo¯urane
induction.1 In addition, compared with placebo, clonidine
(2 mg kg±1, i.v.) has been demonstrated to reduce postoperative sevo¯urane-induced agitation by more than 80%
in children premedicated with midazolam.3 Therefore, we
calculated the sample size according to an expected
reduction of 60% of agitation in children premedicated
with oral clonidine vs midazolam. We estimated that 20
subjects were required per group (power of 80% and
P<0.05).
Data for the two groups (clonidine and midazolam) were
analysed using non-parametric repeated measures analysis
of variance (Friedman's test). Incidence of agitation was
compared between the two groups using the c2 test. A nonparametric equivalent of the unpaired Student's t test
(Mann±Whitney U test) was performed for comparison of
EEG data between children showing agitation during
induction and the other subjects (Statview version 5.0,
Abacus Concept, Berkeley, CA). P<0.05 was considered
signi®cant. Clinical and EEG data are expressed as mean
(SD) and BIS data are expressed as median (range).
Results
Clinical data (Table 1)
Twenty children were enrolled in each group. The two
groups were comparable regarding age and weight. The
time to obtain loss of eyelash re¯ex was similar in the two
groups, but the pupils were centred earlier in the clonidine
group (P=0.027). In all children the trachea was intubated
successfully at the ®rst attempt (at central pupils).
Most of the children required insertion of an oral airway
within the ®rst 3 min of induction to relieve partial airway
obstruction (12 in the clonidine group vs 11 in the
midazolam group). Peripheral oxygen saturation was maintained above 96% throughout the study in all children. In the
two groups, respiratory frequency increased signi®cantly at
loss of eyelash re¯ex compared with control awake values
(20.6 (2.9) in the midazolam group, and 19.2 (2.8) in the
clonidine group at baseline vs 30.5 (8.2) in the midazolam
group and 28.1 (7.1) in the clonidine group at loss of eyelash
re¯ex). This increase was not signi®cantly different between
the two groups. No hypercarbia (de®ned as end-tidal carbon
dioxide >48 mm Hg) or hypocarbia (de®ned as end-tidal
carbon dioxide <30 mm Hg) was observed.
506
BIS, EEG and agitation with sevo¯urane
Fig 2 Examples of EEG tracings recorded during induction of anaesthesia with sevo¯urane in children premedicated with midazolam or clonidine at
baseline (left), 15 s before the occurrence of the nadir of BISÔ (BISnadir) (middle) and at central pupils (right), with the corresponding power spectra
and BISÔ values.
Agitation
Agitation occurred in 12 midazolam-treated and ®ve
clonidine-treated children (P=0.05). This transient agitation
started earlier (45 (7) vs 57 (15) s, P=0.04) and lasted longer
(56 (24) vs 28 (14) s, P=0.03) in the midazolam group
compared with the clonidine group.
BISÔ and EEG data (Table 2)
The BISÔ was similar in the two groups throughout the
study period. Induction of anaesthesia was associated with a
rapid decrease of BISÔ values, reaching the value of about
50 at loss of eyelash re¯ex. The maximal depression of
BISÔ (nadir of BISÔ) occurred at the end of the second
minute of induction (126.0 (35.7) s in the midazolam group
vs 130.7 (22.7) s in the clonidine group, not signi®cant).
From the nadir, the BISÔ increased rapidly to stabilize at
around 40 within the third minute of induction.
Consequently, the BISÔ was higher when the pupils
returned to the central position than at the end of the second
Table 1 Population data. Mean (SD) (range). *Signi®cant differences
between the clonidine and midazolam groups (P<0.05)
Midazolam
Number of patients
Age (months)
Weight (kg)
Time to loss of eyelash re¯ex (s)
Expired concentration at loss of
eyelash re¯ex (%)
Airway obstruction (Y/N)
Agitation (Y/N)
Time to agitation (s)
Time to manual ventilation (s)
Time to central pupils (s)
Expired concentration at central
pupils (%)
Expired concentration 10 min after
intubation (%)
20
62 (25) (37±118)
18 (4) (13±28)
49 (12)
5.7 (1.1)
Clonidine
20
55 (13) (34±78)
18 (4) (13±26)
53 (10)
5.1 (1.0)
11/9
12/8
45 (8)
206 (31)
283 (46)
6.0 (0.6)
12/8
5/15*
57 (15)*
198 (31)
250 (44)*
6.2 (0.5)
2.5 (0.2)
2.5 (0.2)
minute (BISnadir). An example of the time course of BISÔ is
shown in Figure 1.
Regarding the EEG data, the baseline EEG patterns were
different in the clonidine group compared with the
507
40.0 [22±55]
41.0 [28±56]
0.6 (0.4)
0.6 (0.4)
19.0 (2.7)
19.1 (2.7)
62.3 (13.9)
55.0 (15.6)
66.1 (12.5)
60.4 (13.0)
55.2 (10.7)²²
54.7 (10.3)²
52.2 (10.7)
52.3 (11.2)
27.9 (11.2)
21.9 (9.4)
42.2 (15.7)
35.3 (13.9)
40.2 (14.3)
37.9 (6.3)
36.5 (11.6)
34.1 (8.5)
13.2 (5.4)
14.1 (6.3)
14.3 (3.6)
13.8 (4.0)
20.3 (3.1)
19.9 (2.7)
4.0 (1.3)
4.1 (1.8)
39.5 [24±50]²²²
40.0 [20±63]²²²
0.6 (0.4)
0.6 (0.4)
21.0 (2.2)²²²
20.7 (6.3)²
15.3 (5.8)²²
15.0 (5.4)²
12.7 (3.1)²²
12.5 (3.6)
16.8 (3.1)
17.8 (2.7)
4.4 (3.6)
3.7 (1.3)
10 [5±46]
18 [6±36]
0.3 (0.4)
0.4 (1.3)
17.9 (3.1)
18.4 (14.3)
9.2 (4.9)
11.3 (4.5)
9.5 (4.9)
10.5 (5.4)
15.0 (5.4)
16.6 (5.8)
2.7 (1.3)
3.3 (1.3)
67.5 [9±74]
40.0 [6±88]
0.5 (0.4)
0.8 (0.9)
19.5 (2.3)
20.7 (3.1)
30.4 (8.9)
39.6 (7.2)***
Baseline
Midazolam
Clonidine
Loss of eyelash re¯ex
Midazolam
Clonidine
BISnadir
Midazolam
Clonidine
Central pupils
Midazolam
Clonidine
10 min after tracheal intubation (ET sevo¯urane ~2.5%)
Midazolam
Clonidine
10.0 (2.7)
8.2 (2.3)*
13.4 (6.3)
17.2 (8.9)
10.6 (6.7)
12.4 (6.7)
13.7 (3.6)
15.4 (3.6)
3.0 (2.3)
4.4 (3.6)
94.6 [92±97]
94.0 [88±98]
2.0 (0.9)
1.1 (0.4)***
25.6 (1.3)
25.1 (1.3)
9.8 (3.1)
6.2 (2.3)***
14.2 (2.7)
13.5 (2.7)
14.0 (3.1)
18.7 (3.1)***
41.3 (8.9)
28.1 (7.6)***
BISTM
d Ratio
Spectral edge
95 (Hz)
Median power
frequency (Hz)
b Band?
(%)
a Band
(%)
Band
(%)
d Band
(%)
Total spectral
power (mV2)
Table 2 EEG data (mean (SD) and BISÔ data (median [range]). Asterisks indicate signi®cant differences between clonidine and midazolam groups (*P<0.05; ***P<0.001); ²signi®cant differences between BISnadir
and central pupils (²P<0.05; ²²P<0.01; ²²²P<0.001). BISnadir is 15 s before the time of occurrence of the nadir of BISÔ
Constant et al.
midazolam group. The slow waves (delta and theta) were
more prominent and the fast rhythms (beta) were less
pronounced, as re¯ected by a lower delta ratio and median
power frequency after clonidine premedication compared
with midazolam. These differences disappeared when
sevo¯urane was introduced. Induction of anaesthesia was
associated with an increase in the total spectral power of the
EEG signal in all children and a shift towards the lowfrequency spectral bands (Table 2); these changes were
maximal at BISnadir. In the third minute, the relative
contribution of fast oscillations (alpha and beta) increased.
When the pupils returned to the central position, the SEF95
was markedly increased compared with the end of the
second minute of induction (BISnadir) (Table 2). This change
was associated with a signi®cant shift of the EEG oscillations towards high frequencies (Table 2, Fig. 2).
Burst suppressions were observed in four out of 20
patients in each group. Burst suppression occurred in all
cases after centralization of the pupils.
From our systematic clinical observations, we noted that
the nadir of the BISÔ occurred at the end of the agitation
period (Figs 1 and 3).
When the children demonstrating agitation during induction were taken together whatever the premedication group,
they showed a lower mean value of BISÔ and a different
EEG pro®le at BISnadir compared with the children in whom
agitation was not observed. The slow waves (delta) were
more prominent and the faster rhythms (alpha and beta)
were less pronounced, as re¯ected by lower delta ratio and
median power frequency (Table 3, Fig. 3).
Discussion
Clonidine premedication compared with midazolam reduced the incidence of agitation during sevo¯urane induction and reduced the time to obtain centralized pupils in
children. During rapid induction with sevo¯urane 8% in a
50±50 mixture of nitrous oxide and oxygen, the nadir of
BISÔ occurred at the end of the second minute of inhalation
and corresponded to a noticeable shift to lower frequency
bands of the EEG tracing. Whatever the premedication
group, children who demonstrated agitation during induction showed slower EEG traces during the second minute of
induction compared with those who did not demonstrate
agitation.
Clonidine premedication vs midazolam
premedication
In our institution, sevo¯urane induction of anaesthesia is
performed using high inspired concentrations of sevo¯urane
(8%) to speed up loss of consciousness and to reduce the
child's distress scores.11 12 In these conditions we found that
the time required to obtain the clinical end-point of
centralization of the pupils is reduced in children premedicated with clonidine compared with midazolam. These
508
BIS, EEG and agitation with sevo¯urane
Fig 3 Mean values of BISÔ as a function of time, measured during sevo¯urane induction in children who demonstrated agitation and those who
demonstrated no agitation (mean and 95% con®dence interval). These traces are synchronized from the beginning of sevo¯urane inhalation. Individual
agitation periods are represented by black marks.
Table 3 Comparison of EEG and BISÔ data calculated at BISnadir from children showing agitation during induction of anaesthesia [agitation (+)] vs those
showing no agitation [agitation (±)]. EEG data are mean (SD). BISÔ data are median [range]. BISnadir is 15 s before the time of occurrence of the nadir of
BISÔ (median [range])
Agitation (+) (n = 17)
Agitation (±) (n = 23)
P
Total spectral
power (mV2)
d Band
(%)
q Band
(%)
a Band
(%)
b Band
(%)
Median power
frequency (Hz)
Spectral edge
95 (Hz)
d Ratio
Nadir of BISTM
43.2 (14.8)
33.3 (9.8)
0.028
69.2 (8.9)
59.3 (13.4)
0.024
14.8 (4.5)
16.8 (5.8)
0.17
7.8 (2.7)
11.8 (4.9)
0.012
8.8 (3.1)
12.2 (4.0)
0.016
2.4 (0.4)
3.4 (1.3)
0.026
17.3 (3.6)
18.9 (2.7)
0.13
0.3 (0.0)
0.5 (0.4)
0.01
9.4 [6±20]
20.4 [6±46]
0.03
results are in accordance with the shortening of induction
time when anaesthesia is induced with 5% sevo¯urane using
the triple-breath method in adults premedicated with
clonidine compared with placebo.13 This speeding-up of
sevo¯urane induction may be explained by the decrease in
anaesthetic requirement induced by clonidine.
Agitation is not uncommon during induction of anaesthesia with sevo¯urane in children, whatever the induction
technique.14 15 Agitation usually occurs after loss of
consciousness and lasts about 45 s when sevo¯urane 7%
is administered in a mixture of oxygen and nitrous oxide in
children premedicated with midazolam.1 Clonidine premedication was associated with a reduced incidence and a
shorter duration of agitation than that observed with
midazolam. This is in agreement with previous adult and
paediatric studies. Indeed, in adult intensive care patients,
clonidine is effective in preventing and treating clinical
excitement.16 Moreover, compared with placebo, i.v.
clonidine prevents sevo¯urane-induced postoperative clinical excitement in children premedicated with midazolam.3
Both midazolam and clonidine are known to have speci®c
effects on EEG tracings. Intravenous clonidine administration increases delta band activity and decreases alpha band
activity,17 whereas the EEG effects of midazolam are
mainly characterized by an increase in beta activity.18 The
EEG patterns observed at baseline in the two groups of
children premedicated with clonidine or midazolam are
similar to those described in previous adult studies,
509
Constant et al.
suggesting that an effective plasma concentration was
achieved. These EEG differences were not associated with
similar differences in BISÔ values at baseline. This
discrepancy between the BISÔ and EEG data may suggest
that the algorithm of the BISÔ has low sensitivity in this
range of frequencies in an awake patient. However, the
differences between the two EEG pro®les disappeared as
soon as the induction began. The EEG effects of sevo¯urane
seem to override the effects of premedication on EEG
tracings.
paradoxical increase in BISÔ was also observed when the
end-tidal sevo¯urane concentration of 4% was compared
with 3%.7
EEG effects of anaesthetics may vary according to the
cortical topography of recording.23 Our ®ndings result from
frontal and cortical EEG data. This limited ®eld of
investigation may highlight a fast and transient pharmacodynamic effect of sevo¯urane on the frontal cerebral cortex.
However, this hypothesis requires further investigation.
Agitation and EEG-derived variables
EEG and BISÔ data
Sevo¯urane allows tracheal intubation to be performed
without muscle relaxants in children.19 20 In our study the
decision about when the child was ready for intubation was
based on clinical indicators of depth of anaesthesia.
Constriction and centralization of the pupils indicated that
the level of surgical anaesthesia had been reached.
In adults, biphasic EEG changes with increasing drug
concentration have been described for most hypnotic agents,
including sevo¯urane.21 These so-called biphasic effects are
described as an increase in fast rhythms followed by a
decrease in fast rhythms associated with a simultaneous
increase in delta activity. They have been observed during
incremental sevo¯urane inhalation and they correspond to
the sedative effects of hypnotic agents. In our study, using a
high inspired sevo¯urane concentration, the initial EEG
changes were small in the ®rst few seconds after the start of
induction. Induction was associated with an increase in slow
rhythms from loss of eyelash re¯ex to the end of the second
minute of induction. From this point, the frequency of
rhythms increased noticeably within the third minute, with a
shift towards faster rhythms, and then remained unchanged
until the time of tracheal intubation. The BISÔ followed the
EEG changes closely, with a short delay of ~15 s due to the
calculation epoch. The nadir of BISÔ observed at the
beginning of the third minute may re¯ect the slow EEG
frequency observed around the end of the second minute of
induction. The subsequent increase in BISÔ may re¯ect the
shift of EEG rhythms towards faster frequencies.
Consequently, the BIS was noticeably higher when the
plane of surgical anaesthesia was reached, as attested by
centralization of the pupils, compared with a few minutes
sooner. It is interesting that during sevo¯urane induction
with nitrous oxide, when the clinically assessed depth of
anaesthesia increases, BISÔ values increased paradoxically. This discrepancy between the clinically assessed depth
of anaesthesia and BISÔ values during induction of
anaesthesia may be explained by the calculation algorithm,
which was essentially based on i.v. anaesthesia in adults.
However, BISÔ and EEG changes are very similar. Our
®ndings are consistent with the paradoxical increase in
BISÔ values demonstrated in adults when iso¯urane
concentration was increased: this phenomenon has been
explained by a pre-burst EEG state.22 In children this
Agitation phenomena have been reported in almost all
studies designed to assess induction characteristics using
sevo¯urane in children. Agitation is described in most
studies as the occurrence of non-purposeful movements
requiring restraint during induction. In the present study,
agitation was observed in 17 children out of 40. Because
under sevo¯urane anaesthesia we found similar EEG
pro®les in the two premedication groups, we analysed the
children retrospectively according to the presence or
absence of agitation during induction. Our results reveal
that, whatever the premedication group, children who
demonstrated agitation during induction exhibited lower
EEG frequencies 15 s before the nadir of BISÔ compared
with children who remained quiet during induction. It is
interesting to underline the fact that agitation was observed
when BISÔ values were at their lowest and EEG frequencies were slowest. These ®ndings might suggest a relationship between the occurrence of agitation and the rapid
increase in delta activity on EEG observed in the second
minute of induction. Furthermore, the clinical and EEG
events were transient and almost simultaneous in the ®rst 2
min of induction. These observations might be explained by
the rapid onset of cerebral effects of sevo¯urane, which may
affect cortical and subcortical area differently.24 25
However, further research would be needed to investigate
the mechanisms of the sevo¯urane-induced agitation.
In conclusion, we have demonstrated that clonidine
premedication reduces the incidence of agitation compared
with midazolam premedication during sevo¯urane induction. Whatever the premedication group, during rapid
induction with sevo¯urane 8% in a mixture of nitrous
oxide and oxygen (50±50), the nadir of BISÔ occurred at
the end of the second minute of inhalation and corresponded
to a noticeable shift to lower frequency bands of the EEG
tracing. This slowing was more pronounced when agitation
was observed. In addition, our study demonstrated that the
BISÔ may not re¯ect the clinically assessed depth of
anaesthesia during sevo¯urane induction.
References
510
1 Constant I, Dubois MC, Piat V, Moutard ML, McCue M, Murat I.
Changes in EEG and autonomic cardiovascular activity during
BIS, EEG and agitation with sevo¯urane
2
3
4
5
6
7
8
9
10
11
12
13
14
induction of anesthesia with sevo¯urane compared to halothane
in children. Anesthesiology 1999; 91: 1604±15
Mikawa K, Maekawa N, Nishina K, Takao Y, Yaku H, Obara H.
Ef®cacy of oral clonidine premedication in children. Anesthesiology
1993; 79: 926±31
Kulka PJ, Bressem M, Tryba M. Clonidine prevents sevo¯uraneinduced agitation in children. Anesth Analg 2001; 93: 335±8
Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P.
Bispectral analysis measures sedation and memory effects of
propofol, midazolam, iso¯urane, and alfentanil in healthy
volunteers. Anesthesiology 1997; 86: 836±47
Sebel PS, Lang E, Rampil IJ, et al. A multicenter study of bispectral
electroencephalogram analysis for monitoring anesthetic effect.
Anesth Analg 1997; 84: 891±9
Katoh T, Suzuki A, Ikeda K. Electroencephalographic derivatives
as a tool for predicting the depth of sedation and anesthesia
induced by sevo¯urane. Anesthesiology 1998; 88: 642±50
Denman WT, Swanson EL, Rosow D, Ezbicki K, Connors PD,
Rosow CE. Pediatric evaluation of the bispectral index (BIS)
monitor and correlation of BIS with end-tidal sevo¯urane
concentration in infants and children. Anesth Analg 2000; 90:
872±7
Rampil IJ. Elements of EEG signal processing. Int J Monit Comput
1987; 4: 85±98
Levy WJ. Effect of epoch length on power spectrum analysis of
the EEG. Anesthesiology 1987; 66: 489±95
Rampil IJ. A primer for EEG signal processing in anesthesia.
Anesthesiology 1998; 89: 980±1002
Dubois MC, Piat V, Constant I, Lamblin O, Murat I. Comparison
of three techniques for induction of anaesthesia with sevo¯urane
in children. Paediatr Anaesth 1999; 9: 19±23
Baum V, Yemen T, Baum L. Immediate 8% sevo¯urane induction
in children: a comparison with incremental sevo¯urane and
incremental halothane. Anesth Analg 1997; 85: 313±6
Inomata S, Yaguchi Y, Toyooka H. The effects of clonidine
premedication on sevo¯urane requirements and anesthetic
induction time. Anesth Analg 1999; 89: 204±8
Sarner JB, Levine M, Davis PJ, Lerman J, Cook DR, Motoyama EK.
Clinical characteristics of sevo¯urane in children. A comparison
with halothane. Anesthesiology 1995; 82: 38±46
511
15 Piat V, Dubois MC, Johanet S, Murat I. Induction and recovery
characteristics and hemodynamic responses to sevo¯urane and
halothane in children. Anesth Analg 1994; 79: 840±4
16 Ip Yam PC, Forbes A, Kox WJ. Clonidine in the treatment of
alcohol withdrawal in the intensive care unit. Br J Anaesth 1992;
68: 106±8
17 Bischoff P, Scharein E, Schmidt GN, von Knobelsdorff G,
Bromm B, Esch JS. Topography of clonidine-induced electroencephalographic changes evaluated by principal component
analysis. Anesthesiology 2000; 92: 1545±52
18 Greenblatt DJ, Ehrenberg BL, Gunderman J, et al. Pharmacokinetic and electroencephalographic study of intravenous
diazepam, midazolam, and placebo. Clin Pharmacol Ther 1989;
45: 356±65
19 Inomata S, Watanabe S, Taguchi M, Okada M. End-tidal sevo¯urane concentration for tracheal intubation and minimum
alveolar concentration in pediatric patients. Anesthesiology 1994;
80: 93±6
20 Thwaites AJ, Edmends S, Tomlinson AA, Kendall JB, Smith I.
Double-blind comparison of sevo¯urane vs propofol and
succinylcholine for tracheal intubation in children. Br J Anaesth
1999; 83: 410±4
21 Kuizenga K, Wierda JM, Kalkman CJ. Biphasic EEG changes in
relation to loss of consciousness during induction with
thiopental, propofol, etomidate, midazolam or sevo¯urane. Br J
Anaesth 2001; 86: 354±60
22 Detsch O, Schneider G, Kochs E, Hapfelmeier G, Werner C.
Increasing iso¯urane concentration may cause paradoxical
increases in the EEG bispectral index in surgical patients. Br J
Anaesth 2000; 84: 33±7
23 Gugino LD, Chabot RJ, Prichep LS, John ER, Formanek V, Aglio
LS. Quantitative EEG changes associated with loss and return of
consciousness in healthy adult volunteers anaesthetized with
propofol or sevo¯urane. Br J Anaesth 2001; 87: 421±8
24 Osawa M, Shingu K, Murakawa M, et al. Effects of sevo¯urane on
central nervous system electrical activity in cats. Anesth Analg
1994; 79: 52±7
25 Gouvitsos F, Rey M, Bruder N, Mercier LH, Peragut JC. Scalp and
deep brain EEG changes during anesthesia induction with
propofol or sevo¯urane. Anesthesiology 2001; 95: A798