Академический Документы
Профессиональный Документы
Культура Документы
ORIGINAL ARTICLE
• Propofol infusion throughout maintenance of anesthesia can be effective in reducing emergence agitation (EA)
in children compared to sevoflurane anesthesia. However, there are mixed results of efficacy when propofol is
administered as a 1 mgkg 1 bolus at the conclusion of sevoflurane anesthesia.
What this article adds
• Anesthetists can use this simple highly effective intervention for reducing EA in children after sevoflurane anes-
thesia. Further research could assess this intervention in a surgical setting, as part of a multimodal approach to
EA prevention, or directly against other propofol interventions.
Keywords Summary
child; emergence; agitation; delirium;
propofol; sevoflurane Background: Emergence agitation (EA) is a common behavioral disturbance
after sevoflurane anesthesia in children. Propofol 1 mgkg 1 bolus at the end
Correspondence of sevoflurane anesthesia has had mixed results in reducing the incidence of
David Costi, Department of Paediatric EA, whereas propofol infusion throughout anesthesia maintenance seems
Anaesthesia, Women’s and Children’s
effective but is more complex to administer. If a simple, short transition to
Hospital, 72 King William Road, North
propofol anesthesia was found to be effective in reducing EA, this could
Adelaide 5006, SA, Australia
Email: david.costi@health.sa.gov.au enhance the recovery of children following sevoflurane anesthesia. We there-
fore aimed to determine whether transition to propofol over 3 min at the end
Section Editor: Brian Anderson of sevoflurane anesthesia reduces the incidence of EA in children.
Methods: In this prospective randomized controlled trial, 230 children aged
Accepted 15 December 2014 1–12 years, undergoing magnetic resonance imaging (MRI) scans under sevo-
flurane anesthesia were randomized to receive either propofol 3 mgkg 1 over
doi:10.1111/pan.12617
3 min (propofol group), or no propofol (control group), at the end of sevoflu-
rane anesthesia. EA was assessed by a blinded assessor using the Pediatric
Emergence Anesthesia Delirium (PAED) scale and the Watcha scale until
30 min after emergence. EA on the PAED scale was defined as a PAED score
>12. EA on the Watcha scale was defined as a score ≥3. Times to emergence,
postanesthesia care unit (PACU) discharge, and discharge home were also
recorded.
Results: Data were analyzed for 218 children. The incidence of EA was lower
in the propofol group on both PAED (29% vs 7%; relative risk = 0.25; 95%
confidence interval 0.12–0.52; P < 0.001) and Watcha (39% vs 15%; relative
risk = 0.37; 95% confidence interval 0.22–0.62; P < 0.001) scales. Duration
and severity of EA were also reduced in the propofol group. Preplanned sub-
group analyses for midazolam premedication, preexisting cognitive or behav-
ioral disturbance, and age group did not alter our findings. Emergence time
and time in PACU were both increased by a mean of 8 min in the propofol
group (P < 0.001) with no difference in time to discharge home.
Conclusions: Transition to propofol at the end of sevoflurane anesthesia
reduces the incidence of EA and improves the quality of emergence. There is
a small increase in recovery time, but no delay in discharge home.
mask airways were removed prior to emergence. All rousing (defined as the ‘onset of eye opening or purpose-
patients received face mask oxygen delivered by a ful movement’). Upon emergence, EA was assessed
T-piece circuit until emergence and were monitored with simultaneously using two scales. The Pediatric Emer-
continuous pulse oximetry until discharge from the gence Anesthesia Delirium (PAED) scale (12) (Table S1)
PACU. and Watcha scale (13) (Table S2) scores were recorded
at 5 min intervals until 30 min after emergence. EA on
the PAED scale was defined as a PAED score >12 at
Outcome assessment
any time in the 30 min after emergence (1). EA as
An outcome assessor who was blinded to group alloca- assessed by the Watcha scale was defined as a score ≥3
tion observed the child from the time of arrival in the at any time in the 30 min after emergence. The primary
PACU until 30 min after emergence. The time to emer- outcomes were the incidence of EA according to the
gence was defined as the time from cessation of sevoflu- PAED and Watcha scales. The secondary outcomes
rane or propofol administration until the first sign of were peak PAED scores, emergence time, time in
Excluded (n = 12)
♦ Declined to participate
Randomized (n = 230)
Excluded from analysis due to: Excluded from analysis due to:
♦ laryngospasm on emergence requiring treatment with ♦ laryngospasm on induction requiring treatment with
propofol and/or suxamethonium (n = 3) propofol and/or suxamethonium (n = 1)
♦ protocol violation: IV midazolam given (n = 1) ♦ laryngospasm on emergence requiring treatment with
♦ assessor unavailable due to overlapping recruits (n = 1) propofol and/or suxamethonium (n = 1)
♦ addition of an unscheduled painful procedure (n = 1) ♦ protocol violation: IV ketamine given (n = 1)
♦ assessor unblinded due to unflushed propofol visible
in IV bung (n = 1)
♦ missing data collection forms (n = 1)
PACU, and time to hospital discharge. We performed Table 1 Demographics of study participants
preplanned subgroup analyses of the incidence of EA Control Propofol
for the presence or absence of midazolam premedica- group group
tion, age group, and the presence or absence of a preex- (n = 109) (n = 109)
isting cognitive or behavioral disturbance (defined by
Sex, n
medical record diagnosis and/or parental verbal report,
Female 48 51
e.g., developmental delay, autism). Male 61 58
Age (years), 4.1 2.8 4.0 2.4
mean SD
Statistical analysis
Median (IQR) 3.5 (1.9, 5.4) 3.7 (2.3, 5.2)
A recent prospective audit of the incidence of EA in our Age groups, n
institution was found to be 32% (1). A sample size cal- 1–6 years 92 98
≥7 years 17 11
culation showed that 216 patients (108 in each group)
Weight (kg), 18.0 7.3 17.7 7.0
would be required to show a reduction in the incidence mean SD
of EA to 15% with a power of 80% and an a of 0.05. Median (IQR) 16.2 (12.7, 21.4) 16.3 (13.0, 20.6)
We recruited 230 patients to allow for protocol viola- Cognitive or behavioral 30 31
tions and exclusions. disorder, n
Continuous data were checked for normality using Premedication, n 12 10
the Shapiro–Wilk test. Demographics and group com- Induction type, n
Inhalational 106 105
parisons of continuous variables were conducted using
Intravenous 3 4
Mann–Whitney U-tests for nonnormally distributed Anesthetic durationa 73 25 65 20
data, with mean SD or median (interquartile range) (min), mean SD
reported. Incidence of EA and group comparisons of Median (IQR) 67 (56, 85) 62 (54, 74)
categorical data are reported as frequency and percent- MRI region scanned, n
age, using chi-squared analysis. A logistic regression Head 78 86
model with EA as our binary-dependent outcome vari- Head and spine 7 4
Spine 12 7
able was performed with emergence propofol (yes/no),
Other 12 12
midazolam premedication (yes/no), preexisting cognitive
or behavioral disturbance (yes/no), age (continuous), SD, standard deviation; IQR, interquartile range; MRI, magnetic reso-
and induction technique (i.v. or inhalational) as indepen- nance imaging.
a
dent predictor variables. Unadjusted (crude) odds ratios In contrast to all other baseline parameters, there was a statistical
difference in mean and median anesthetic duration of 8 and 5 min,
were obtained for each individual predictor and an over-
respectively.
all logistic regression was then run with all predictors
together to obtain adjusted odds ratios. P < 0.05 was
considered statistically significant. All statistical analyses
were performed using IBM SPSS Statistics for WINDOWS, Figure 2 illustrates the distribution of peak PAED
Version 20.0. (IBM Corp., Armonk, NY, USA). scores showing that children in the propofol group were
far more likely to have a zero PAED score throughout
their PACU stay (21% vs 9% P = 0.014). Figure 2 also
Results
shows a grouping of very high PAED scores in controls
A total of 230 children were randomized from January which is absent in the propofol group. Figure 3 shows
2010 to September 2013, of which 218 were analyzed the influence of time after emergence on the proportion
(Figure 1). The demographics of the two groups were of children with EA (as defined by the PAED score
comparable (Table 1). >12). Unlike the control group, no children in the prop-
Emergence agitation and recovery outcomes are pre- ofol group had PAED scale EA beyond 15 min after
sented in Table 2. The overall incidence of EA was emergence.
lower in the propofol group when measured by both Preplanned subgroup analyses of premedication and
PAED (29% vs 7%, P < 0.001) and Watcha (39% vs preexisting cognitive or behavioral disorder are consis-
15%, P < 0.001) scales. Peak PAED scores were also tent with the finding that propofol is effective in reduc-
significantly lower in the propofol group (P < 0.001). ing the incidence of EA in these patient subgroups
The emergence times and PACU times were a mean of (Table 2). With logistic regression, we found that no
8 min longer in the propofol group, but there was no emergence propofol was the only predictor associated
difference in time to discharge home (Table 2). with a higher incidence of EA (Table 3).
CI, confidence interval; EA, emergence agitation; PAED, Pediatric Anesthesia Emergence Delirium; IQR, interquartile range; SD, standard devia-
tion; PACU, postanesthesia care unit.
a
Preexisting cognitive or behavioral disorder (PCBD) and/or midazolam premedication and/or intravenous induction.
PAED, Pediatric Anesthesia Emergence Delirium; EA, emergence agitation; OR, odds ratio; CI, confidence interval.
intravenous anesthesia (TIVA) or throughout Our trial has a few limitations in the setting studied.
maintenance after sevoflurane induction (5,6). These To improve external validity, we elected to include chil-
approaches generally require the additional complexity dren with preexisting cognitive and behavioral disorders
of an infusion pump and application of an algorithm for (the majority of which had a mild cognitive impairment
infusion rate (e.g., target-controlled infusion or manu- due to developmental delay), and those requiring mi-
ally adjusted rate at certain time intervals). In our insti- dazolam premedication. Although this introduces other
tution, anesthetists often prefer the simplicity of variables, it was an a priori subgroup analysis and num-
manually administered propofol from a syringe incre- bers were evenly distributed across the two study
mentally in doses >1 mgkg 1 over a few minutes groups. Reassuringly, propofol administration seems
toward the end of anesthesia. At the time of designing effective in the presence or absence of these factors
this study, four RCTs had reported comparing the inci- (Tables 2 and 3). We did not study adverse events aside
dence of EA and/or agitation scores with two showing a from laryngospasm (which occurred in <3% of random-
benefit (7,8) and two no benefit (9,14) with propofol ized subjects) and desaturation. The hemodynamic and
1 mgkg 1. By the completion of this study, there were respiratory effects of propofol boluses ranging in doses
eight studies with five showing a benefit (7,8,15–17) and up to 3 mgkg 1 administered in the presence of sevoflu-
three no benefit (9,10,14) with propofol 1 mgkg 1. The rane anesthesia have already been well described in chil-
only study performed in the MRI setting was much dren, so this was not further investigated in this study
smaller than ours and showed a reduced incidence of (18). Although we did not study beyond 30 min after
EA when EA was defined as a PAED score ≥16 (8). This emergence, it may have been useful to measure the post-
score is a very high cut-off for defining EA and had we operative behavior beyond this time frame and also
used this definition, our results show an even greater some measure of parent/caregiver satisfaction. A less
effect in favor of propofol (see Figure 2). obvious limitation may be the choice of EA scale. Sikich
The etiology of EA is unknown and it may be multi- and Lerman (12) who devised the PAED scale did not
factorial. The emergence delayed by an average of 8 min describe a cutoff score that indicates the presence or
would suggest that the mechanism of action of the prop- absence of EA or ED. Cutoff scores ranging from ≥10 to
ofol transition may be facilitation of sevoflurane wash- ≥16 have subsequently been used without rigorous justi-
out prior to the point of emergence. We speculate that fication for the level chosen (1). Numerous scales other
sevoflurane-associated EA may occur within a window than PAED have also been used to assess EA (3). The
of brain concentrations of sevoflurane. For example, reason for using the Watcha scale in addition to PAED
children may be vulnerable to EA between a brain or is that it is simpler to use and provides a meaningful
endtidal concentration of 0.2% and 0.1%, and less so at clear endpoint, i.e., consolability, for the dichotomous
0.05% and below. It is possible that there is an intrinsic outcome of EA. We did not define a minimum time
property of propofol that accounts for its beneficial frame for which a child reached an EA threshold to be
effect in reducing EA. However, a diverse group of considered to have EA. A minimum time frame of
agents including dexmedetomidine, clonidine, fentanyl, 3 min (19) or 5 min (20) has been suggested previously.
and ketamine can also reduce sevoflurane EA (5,6), and In the current study, the majority of children assessed as
these agents all have in common with propofol the having EA were beyond the EA threshold in more than
capacity to delay emergence and facilitate sevoflurane one 5 min epoch.
washout. In contrast to propofol, they all have analgesic To avoid longer emergence times or PACU stays, the
properties which may contribute to their effectiveness transition to propofol with concurrent cessation of sevo-
after painful procedures; however, this is not a plausible flurane could be made prior to completion of the proce-
mechanism in the MRI setting. dure. Our findings suggest that a transition 8 min prior
References
1 Bajwa SA, Costi D, Cyna AM. A compari- 8 Abu-Shahwan I. Effect of propofol on emer- dure. J Shanghai Jiaotong Univ (Med Sci)
son of emergence delirium scales following gence behavior in children after sevoflurane 2010; 30: 73–75.
general anesthesia in children. Pediatr general anesthesia. Pediatr Anesth 2008; 18: 16 Kim YH, Yoon SZ, Lim HJ et al. Prophylac-
Anesth 2010; 20: 704–711. 55–59. tic use of midazolam or propofol at the end
2 Malarbi S, Stargatt R, Howard K et al. 9 Bakhamees HS, Mercan A, El-Halafawy of surgery may reduce the incidence of emer-
Characterizing the behavior of children YM. Combination effect of low dose fenta- gence agitation after sevoflurane anaesthesia.
emerging with delirium from general anes- nyl and propofol on emergence agitation in Anaesth Intensive Care 2011; 39: 904–908.
thesia. Pediatr Anesth 2011; 21: 942–950. children following sevoflurane anesthesia. 17 Kim MS, Moon BE, Kim H et al. Compari-
3 Vlajkovic GP, Sindjelic RP. Emergence Saudi Med J 2009; 30: 500–503. son of propofol and fentanyl administered at
delirium in children: many questions, few 10 Lee CJ, Lee SE, Oh MK et al. The effect of the end of anaesthesia for prevention of
answers. Anesth Analg 2007; 104: 84–91. propofol on emergence agitation in children emergence agitation after sevoflurane anaes-
4 Voepel-Lewis T, Malviya S, Tait AR. A pro- receiving sevoflurane for adenotonsillectomy. thesia in children. Br J Anaesth 2013; 110:
spective cohort study of emergence agitation Korean J Anesthesiol 2010; 59: 75–81. 274–280.
in the pediatric postanesthesia care unit. 11 Fronapfel PJ. Prevention of emergence delir- 18 Lerman J, Houle TT, Matthews BT et al.
Anesth Analg 2003; 96: 1625–1630. ium. Pediatr Anesth 2008; 18: 1113–1114. Propofol for tracheal intubation in children
5 Costi D, Cyna AM, Ahmed S et al. 12 Sikich N, Lerman J. Development and psy- anesthetized with sevoflurane: a dose-
Effects of sevoflurane versus other general chometric evaluation of the pediatric anes- response study. Pediatr Anesth 2009; 19: 218–
anaesthesia on emergence agitation in chil- thesia emergence delirium scale. 224.
dren. Cochrane Database Syst Rev 2014; Anesthesiology 2004; 100: 1138–1145. 19 Cravero J, Surgenor S, Whalen K. Emer-
9: CD007084. 13 Watcha MF, Ramirez-Ruiz M, White PF gence agitation in paediatric patients after
6 Dahmani S, Stany I, Brasher C et al. Phar- et al. Perioperative effects of oral ketorolac sevoflurane anaesthesia and no surgery: a
macological prevention of sevoflurane- and and acetaminophen in children undergoing comparison with halothane. Paediatr Ana-
desflurane-related emergence agitation in bilateral myringotomy. Can J Anaesth 1992; esth 2000; 10: 419–424.
children: a meta-analysis of published stud- 39: 649–654. 20 Cravero JP, Beach M, Thyr B et al. The
ies. Br J Anaesth 2010; 104: 216–223. 14 Chiba S, Shima T, Murakami N et al. Effect effect of small dose fentanyl on the emer-
7 Aouad MT, Yazbeck-Karam VG, Nasr VG of propofol on sevoflurane agitation in chil- gence characteristics of pediatric patients
et al. A single dose of propofol at the end of dren. Masui 2003; 52: 611–615. after sevoflurane anesthesia without surgery.
surgery for the prevention of emergence agi- 15 Sun Y, Hu J, Xu W-Y et al. Combination Anesth Analg 2003; 97: 364–367.
tation in children undergoing strabismus sur- effect of tramadol and low dose propofol on 21 Rosen HD, Cravero JP. Research on emer-
gery during sevoflurane anesthesia. emergence agitation in children receiving gence agitation in children. Can J Anaesth
Anesthesiology 2007; 107: 733–738. sevoflurane for adenotonsillectomy proce- 2013; 60: 822–823.