Académique Documents
Professionnel Documents
Culture Documents
12713
Review Article
A comparison of total intravenous anaesthesia using propofol with
sevoflurane or desflurane in ambulatory surgery: systematic
review and meta-analysis
G. Kumar,1,2 C. Stendall,1 R. Mistry,2 K. Gurusamy3 and D. Walker4
1 Speciality Registrar, 4 Consultant, Department of Anaesthesia and Intensive Care, University College London
Hospitals NHS Foundation Trust, London, UK
2 Research Fellow, Department of Peri-operative Medicine, 3 Lecturer, Division of Surgery, University College London,
London, UK
Summary
With the popularity of ambulatory surgery ever increasing, we carried out a systematic review and meta-analysis to
determine whether the type of anaesthesia used had any bearing on patient outcomes. Total intravenous propofol
anaesthesia was compared with two of the newer inhalational agents, sevourane and desurane. In total, 18 trials
were identied; only trials where nitrous oxide was administered to, or omitted from, both groups were included. A
total of 1621 patients were randomly assigned to either propofol (685 patients) or inhalational anaesthesia
(936 patients). If surgical causes of unplanned admissions were excluded, there was no difference in unplanned
admission to hospital between propofol and inhalational anaesthesia (1.0% vs 2.9%, respectively; p = 0.13). The inci-
dence of postoperative nausea and vomiting was lower with propofol than with inhalational agents (13.8% vs 29.2%,
respectively; p < 0.001). However, no difference was noted in post-discharge nausea and vomiting (23.9% vs 20.8%,
respectively; p = 0.26). Length of hospital stay was shorter with propofol, but the difference was only 14 min on
average. The use of propofol was also more expensive, with a mean (95% CI) difference of 6.72 (5.138.31 (8.16
(6.23 10.09); $11.29 ($8.62$13.96))) per patient-anaesthetic episode (p < 0.001). Therefore, based on the published
evidence to date, maintenance of anaesthesia using propofol appeared to have no bearing on the incidence of unplanned
admission to hospital and was more expensive, but was associated with a decreased incidence of early postoperative
nausea and vomiting compared with sevourane or desurane in patients undergoing ambulatory surgery.
.................................................................................................................................................................
Correspondence to: G. Kumar
Email: gautamkumar@doctors.org.uk
Accepted: 7 April 2014
intravenous anaesthesia (TIVA), with modern drugs Table 1 Inclusion criteria according to the PICOS
like propofol administered by continuous infusion, is framework [12].
now commonplace in ambulatory surgical settings.
However, the proposed advantages and propagation of
Population Adult (16 years or over), elective
the relatively new and less-soluble inhaled anaesthetics, ambulatory surgery, undergoing general
sevourane and desurane, have added a new dimen- anaesthesia, non-sedation procedures
Interventions Maintenance anaesthesia with propofol
sion to patient care by rivalling propofol in ambula-
(total intravenous anaesthesia)
tory surgery. In theory, they allow faster recovery and Controls Maintenance with desflurane and/or
earlier discharge home than the more traditional (intervention) sevoflurane (inhalational anaesthesia)
Outcomes Primary: unplanned admission after
inhalational anaesthetic agents [4, 5]. planned day surgery
A number of anaesthetic issues may delay discharge Secondary: serious adverse events;
postoperative and post-discharge nausea
from hospital, including: cognitive recovery; cardiovas-
and vomiting; postoperative pain (48 h
cular recovery; pain; return to normal activity; and postoperatively); length of hospital stay;
postoperative nausea and vomiting (PONV) [6]. Postop- cost analysis; quality of life
Study design Randomised controlled studies in any
erative nausea and vomiting can be particularly prob- language
lematic in ambulatory surgery as it may lead to delay in
discharge or unscheduled admission to hospital [7]. PICOS, Population, Intervention, Control, Outcome, Study
design.
Additionally, it has been reported as the anaesthetic
complication that is of most concern to patients [8].
Multiple factors, including the anaesthetic agent deliv- irrespective of the method of induction of anaesthesia
ered, are associated with an increased incidence of (intravenous or inhalational), and whether an anti-
PONV and the optimal strategy for preventing PONV emetic was used intra-operatively (provided that it was
continues to be debated. Postoperative pain is also prob- used in both groups).
lematic, with over 30% of patients reporting moderate to We searched MEDLINE, EMBASE, the Cochrane
severe pain following ambulatory surgery [9, 10]. Central Register of Controlled Trials (CENTRAL) in the
We undertook a systematic review and meta- COCHRANE Library, Science Citation Index Expanded
analysis to assess whether the maintenance of anaes- and the meta-register of controlled trials were until
thesia using propofol TIVA is associated with fewer November 2013 through a formal search strategy. The
unplanned hospital admissions than maintenance with medical subject headings (MeSH) terms anaesthesia,
the inhalational agents sevourane or desurane. inhalational, anaesthesia, intravenous or TIVA, and
ambulatory or day-case surgery were used along with
Methods equivalent free-text search terms as part of the search
The systematic review was conducted following the Pre- strategy. Additional studies were identied by review of
ferred Reporting Items for Systematic Reviews and the reference sections of all eligible studies.
Meta-analyses (PRISMA) guidelines [11]. We designed The decision to include a study was based on an
a PICOS framework (Population, Intervention, Control, independent review of each of the identied manu-
Outcome, Study design) to identify controlled studies, scripts by two study investigators (GK and CS) Poten-
and used its elements as our primary selection criteria tially eligible studies (as determined by either reviewer)
[12]. Articles were regarded as potentially eligible if they underwent review of the full-text manuscript by both
met all of the criteria depicted in Table 1. Also shown reviewers working independently, who subsequently
are the studys primary and secondary endpoints. compared their judgements. Disagreements were har-
Only studies where the use of nitrous oxide was monised by consensus and if necessary by arbitration
matched (i.e. present or absent in both study groups) by the senior researcher. Data extraction tables and an
were included, as this was deemed a signicant con- extraction form, including a set of parameters for
founder having both analgesic properties and a higher relevant information on pre-operative testing, were
incidence of PONV [1315]. All studies were included created.
The risk of bias was assessed according to the propofol vs sevourane and propofol vs desurane.
guidelines of the Cochrane Collaboration [16, 17], Chi-squared tests for heterogeneity were performed to
since there is a risk of overestimation of benecial identify the differences between subgroups. A planned
treatment effects in randomised clinical trials with a subgroup analysis of ASA status was not possible due
high risk of bias [18, 19]. The assessment of risk of to the relatively low anaesthetic risk of included
bias was based on: sequence generation; allocation patients, and a planned subgroup analysis comparing
concealment; blinding (of patients, anaesthetists and different types of surgery was also not possible due to
outcome assessors); incomplete outcome data; selective there being too few trials in each category.
outcome reporting; and source of funding [20]. Sensitivity analysis was performed excluding all
Considering that follow-up was generally short trials where nitrous oxide was used. The impact of this
(< 24 h after discharge), and the incidence of compli- on the primary outcome and PONV is presented in
cations relatively low, trials that did not report the the results. Too few studies existed to perform it on
outcomes of all randomly assigned patients were other outcomes. All evidence were graded as low- or
considered to suffer from bias owing to incomplete out- moderate-quality as per the GRADE approach [20],
come data. Exploration of bias using a funnel plot was due to the limitations in trial design and implementa-
performed when assessing PONV and length of stay in tion and/or the imprecision of the results due to small
all trials, but was not possible for other outcomes sample sizes.
because of the small number of trials included [21].
The software package RevMan 5 [22], provided by Results
the Cochrane Collaboration, was used for analysis. For After removing duplicates a, total of 347 references
dichotomous data, the relative risk (RR) with the cor- were identied from medical journal databases and the
responding 95% condence intervals (CI) was calcu- World Health Organization International Clinical Tri-
lated for each study, and the results were pooled als Registry Platform. One additional trial was identi-
together using the MantelHaenszel method for com- ed by searching through references (Fig. 1). In total,
bining trials. For continuous data, the mean difference 18 publications describing 18 completed randomised
and its corresponding 95% CI were calculated using trials fullled the inclusion criteria. A total of 1621
the inverse variance method for meta-analysis. The patients were randomly assigned to either TIVA (685
individual effect sizes were weighted according to the patients) or inhalational anaesthesia (936 patients). All
reciprocal of their variance. If the standard deviation trials included adult patients (age 16 or above) who
was not available for continuous outcomes, it was underwent ambulatory surgery and compared TIVA vs
calculated according to the guidelines of the Cochrane sevourane (1052 patients; 10 trials) and TIVA vs des-
Collaboration [20]. urane (569 patients; eight trials) (Table 2). Five trials
The outcomes from both xed- and random-effects included patients who did not receive nitrous oxide
models were obtained in the meta-analysis; the model [27, 35, 3941]. No serious adverse incidents attribut-
used did not inuence the interpretation of effectiveness able to any of the anaesthetic agents were reported.
and the random-effects model is presented in the The majority of trials had a high risk of bias.
Results section as we included different types of ambu- Blinding was achieved in only one trial [38] as to
latory procedures. Heterogeneity (across-study inconsis- whether TIVA or inhalational anaesthesia was used,
tency) was quantied using the I2 statistic, which and in most trials, details of the randomisation proce-
estimates the percentage of variability across studies not dure or whether blinding of the outcome assessors
due to chance [23, 24]. Values of I2 < 40% might not be took place was unclear (Table 3). Dropouts after ran-
important, 3060% may represent moderate heteroge- domisation were reported in only three studies [29, 37,
neity, 5090% may represent high heterogeneity and 42]. Only 10 out of the 18 trials reported the primary
values > 75% considerable heterogeneity. outcome, thus raising the possibility of selective out-
Subgroup analysis was performed where possible come reporting. It was unclear whether any trial was
for each outcome to explore heterogeneity between free from funding bias.
Idenficaon
n = 547 n=1
Records screened
n = 348
Records excluded
based on abstract
screening
n = 311
Eligibility
Studies included in
qualitave synthesis
n = 18
Included
Studies Included in
quantave synthesis
n = 18
In total, 10 trials reported unplanned admission to with inhalational anaesthesia (13.8% vs 29.2%, respec-
hospital following ambulatory surgery. Four out of tively, RR 0.50 (95% CI 0.350.71, p < 0.001; Fig. 4).
them detailed no admissions [25, 30, 34, 35], and one Similarly, PONV was lower comparing propofol with
trial excluded inpatient admissions after randomisation sevourane (15.1% vs 26.7%, respectively, RR 0.56
[27]. The incidence of unplanned admission was lower (95% CI 0.340.90); p = 0.02), and propofol with des-
with propofol compared with inhalational agents (1.3% urane (10.9% vs 33.3%, respectively, RR 0.41 (95% CI
vs 4.4%, respectively, RR 0.35 (95% CI 0.140.91), 0.240.70); p = 0.001). There was moderate to high
p = 0.03, Fig. 2). However, if surgical causes of admis- heterogeneity in these results (I2 = 52%). Sensitivity
sion were excluded and only anaesthetic reasons for analysis excluding all trials involving nitrous oxide
admission included (pain, PONV, respiratory depres- showed no difference in PONV (17% vs 20%, respec-
sion), there was no difference between propofol and tively, RR 0.87 (95% CI 0.511.47); p = 0.60).
inhalational anaesthesia (1.0% vs 2.9%, respectively, Seven trials reported PONV up to 24 h after dis-
RR 0.44 (95% CI 0.151.27), p = 0.13; Fig. 3). No het- charge (Fig. 5). There was no difference between prop-
erogeneity was noted (I2 = 0%). Sensitivity analysis ofol and inhalational anaesthesia (23.9% vs 20.8%,
excluding all trials involving nitrous oxide showed no respectively, RR 1.17 (95% CI 0.891.55; p = 0.26).
difference in unplanned admission between the two Analogous results were seen when comparing propofol
anaesthetic techniques. with sevourane (27.7% vs 23.6%, respectively, RR
All trials reported pre-discharge PONV. The 1.22 (95% CI 0.891.67; p = 0.22) [27, 33, 34, 42], and
incidence of PONV with propofol was lower compared propofol with desurane (15.8% vs 15.0%; respectively,
Total Documented
number Was pre- or
Type of of Mean N2O Induction of intra-operative
Study Comparison surgery patients age; years used? anaesthesia anti-emetic
Ashworth and Propofol vs Body surface 60 44 Yes Propofol No
Smith [25] desflurane
Chen et al. [26] Propofol vs Gynaecology 80 29 Yes Propofol No
sevoflurane
Fish et al. [27] Propofol vs Urology 71 63 No Propofol or No
sevoflurane sevoflurane
Fredman et al. [28] Propofol vs Mixed 146 35 Yes Propofol or No
sevoflurane sevoflurane
van Hemelrijk et al. [29] Propofol vs Gynaecology 90 30 Yes Propofol or No
desflurane desflurane
Jakobsson et al. [30] Propofol vs Orthopaedic 80 35 Yes Propofol No
desflurane
Kurpiers et al. [31] Propofol vs Not stated 53 40 Yes Propofol No
desflurane
Lebenbom-Mansour Propofol vs Orthopaedic 60 29 Yes Propofol or No
et al. [32] desflurane desflurane
Nathan et al. [33] Propofol vs Gynaecology 52 ? Yes Propofol or No
sevoflurane sevoflurane
Raeder et al. [34] Propofol vs Orthopaedic 169 33 Yes Propofol No
sevoflurane
Raeder et al. [35] Propofol vs General 60 ? No Propofol Ondansetron
desflurane surgery or droperidol
Rapp et al. [36] Propofol vs Orthopaedic 91 34 Yes Propofol or No
desflurane desflurane
Smith et al. [37] Propofol vs Mixed 211 39 Yes Propofol or No
sevoflurane sevoflurane
Smith and Propofol vs Mixed 61 38 Yes Propofol or No
Thwaites [38] sevoflurane sevoflurane
Stevanovic et al. [39] Propofol vs General 60 48 No Propofol Metoclopramide
sevoflurane surgery
Tan et al. [40] Propofol vs Gynaecology 80 34 No Propofol or Dexamethasone
sevoflurane sevoflurane
Tang et al. [41] Propofol vs Mixed 75 55 Yes Propofol Metoclopramide,
desflurane ondansetron,
droperidol
White et al. [42] Propofol vs Gynaecology 122 38 No Propofol Dolasteron
sevoflurane
RR 1.03 (95% CI 0.561.87); p = 0.93) [25, 35, 41]. The chi-squared test for subgroup differences was not
Overall, heterogeneity was high (I2 = 76%). signicant (p = 0.92), and heterogeneity was moderate.
In total, nine trials reported postoperative pain; in Five trials reported a cost analysis (Fig. 7). Three
all these trials, there was no difference between propo- included the cost of maintenance anaesthesia only (i.e.
fol and inhalational anaesthesia. Eleven trials reported the cost of the propofol or desurane/sevourane
either the total hospital stay or t-for-discharge time used) [30, 38, 39], and two included the cost of all
(Fig. 6). These times were shorter in patients receiving ancillary drugs (induction agents, neuromuscular
propofol compared with those receiving inhalational blockers, analgesics and anti-emetics) [31, 37]. Total
anaesthesia; mean (95% CI) difference was 14 ( 21 costs were signicantly higher with propofol compared
to 8) min, p < 0.001. Hospital stay was also shorter with inhalational anaesthesia in each individual trial
comparing propofol with sevourane ( 14 ( 21 to and in the meta-analysis. The mean (95% CI) differ-
6) min, p < 0.001), but not when comparing propo- ence was 6.72 (5.138.31 (8.16 (6.23 10.09);
fol with desurane ( 15 ( 30 to 1) min, p = 0.06). $11.29 ($8.62 $13.96))) per patient anaesthetic
Table 3 Risk of bias (?, unclear risk of bias; , high risk of bias; +, low risk of bias).
Figure 3 Unplanned hospital admissions comparing propofol vs inhalational anaesthesia; anaesthetic causes.
A total overall unplanned admission rate of 2.1% was from 0.5% to 10% [43, 44], with an ideal standard of
identied in this review, which is comparable with sev- under 2% [45]. We have also shown that the incidence
eral clinical studies that have reported rates ranging of PONV was lower when propofol was used
compared with inhalational anaesthesia, but no differ- shorter with propofol but it is arguable whether the
ence was noted in the incidence of nausea and vomit- small difference found is likely to be clinically relevant.
ing after discharge. Length of hospital stay also was Additionally, no difference was noted in the incidence
SE(log[RRI) SE(MD)
0
0 (a) (b)
10
0.5
20
30
1.5
40
RR MD
2 50
0.01 0.1 1 10 100 50 25 0 25 50
Figure 8 Funnel plot for propofol vs inhalational anaesthesia, for (a) postoperative nausea and vomiting and (b)
length of hospital stay. The vertical axis shows the precision of the estimate of the treatment effect (the smaller the
condence interval, the more precise the study, and the further up the study is placed). The horizontal axis measures
the treatment effect (relative risk or mean difference.) The data points are from each study and the vertical line
shows where the pooled estimate from the meta-analysis lies.
of postoperative pain but maintenance anaesthesia expected, since major morbidity and mortality directly
with propofol was more expensive than with sevoura- attributed to anaesthesia now occur extremely rarely
ne or desurane. [46]. Likewise, no studies reported quality of life, and
No serious anaesthetic adverse events were this was expected due to the nature of surgery, with
reported in any of the selected studies, and this was as no difference in quality of life resulting from different
methods of anaesthesia anticipated several days follow- home compared with inhalational anaesthesia, sup-
ing discharge. Considering that mortality and quality ported by a previous study that concluded that nausea
of life contribute to the quality-adjusted life year, was the single most important factor which affected
which is the main measure used to assess benets in discharge following ambulatory surgery [7]. However,
cost-effectiveness analyses, the choice of anaesthetic although statistically signicantly different, the clinical
agent will depend on re-admission to hospital, costs, signicance of the difference in hospital readiness is
PONV and pain. very minimal, as the average difference was only about
We have shown that PONV was less common 1015 min between the two techniques.
when TIVA was used this correlates with extensive None of the trials reported any signicant differ-
evidence that propofol is the best anaesthetic agent for ence in postoperative pain or opioid use. This may be
preventing PONV [4749], and is of particular impor- because the analgesic properties of propofol have been
tance since it is the postoperative anaesthetic complica- well reported, and so this was not investigated [54
tion that is of most concern to patients [8]. It is also 56]. Alternatively, it may be that there was no
signicant because of the potential association with signicant difference in pain between TIVA and inha-
delayed discharge and unplanned admissions, and latational anaesthesia because of the nature of the pro-
related cost implications [6, 50]. This may be sup- cedures carried out.
ported by the fact that hospital admission was lower In interpreting differences in actual discharge and
with TIVA in all the selected studies, although this did t-for-discharge times, it is critical to understand that
not reach statistical signicance. The distinction other co-factors inuencing the interpretation of each
between pre- and post-discharge nausea and vomiting outcome may be relevant. Being ready for discharge is
has not been well investigated [42]. After discharge, usually assessed using standardised recovery protocols,
PONV has been reported to have an incidence of 30 whereas discharge time is the actual time when the
50%, and we have reported an incidence of 22% in this patient leaves hospital. The latter is affected by many
meta-analysis. Many patients experience PONV for the factors such as local guidelines, delay in transport, or
rst time only after leaving hospital [50, 51]. Interest- needing to pass urine, although type of anaesthetic
ingly, despite propofols being associated with signi- should not affect this. Unfortunately, many authors
cantly less in-hospital PONV, our analysis shows it have not made a distinction between these variables.
provides no benet in the rst 24 h after discharge. In our study, the cost of propofol as a mainte-
This may be explained by propofols pharmacokinetic nance anaesthetic was signicantly higher than that of
prole. A minimum plasma concentration of propofol desurane or sevourane. No trials calculated the total
(mean 343 ng.ml 1 [52]) is necessary to produce an cost including ancillary equipment or wastage. It has
anti-emetic effect; this is lower than that required to been recommended that drug wastage should be
produce an anaesthetic effect (approximately included in the total cost [57]. If this was the case,
500 ng.l 1), and since it has a short half-life, therapeu- propofol may cost even more, in part, because it is
tic anti-emetic plasma levels are unlikely to persist packaged as a preservative-free, single-patient-use
after several hours [53]. Propofol may provide prophy- agent [58]. Therefore, any opened propofol remaining
laxis against early PONV only, and suitable strategies at the end of surgery would have been wasted.
should be planned for patients after discharge. Although the saving per patient was small, the poten-
Although not having a direct cost implication to the tial for savings over the course of a year would be
hospital, preventing PONV once the patient has large. Whether this saving would be negated by the
returned home is still important as it may delay potential extra need for anti-emetics with inhalational
resumption of normal activities and readiness for anaesthesia, or the need for different equipment such
work, and negatively impact on patient experience. as infusion pumps [59], has not been analysed.
The reduction in the incidence of PONV may be There are a number of limitations to our meta-
the reason why our analysis demonstrated that propo- analysis. Although we conducted a thorough review of
fol was associated with earlier readiness for discharge work published in all languages, one trial was
excluded due to being untraceable [60]. Many authors meta-analysis of this nature and the results are compa-
did not differentiate between nausea and vomiting or rable with a previous systematic review that included
did not state whether it was counted twice as two seven of the trials we analysed and included compari-
pathologies. We included patients who were adminis- son with isourane, but did not differentiate use of
tered nitrous oxide; although it is less commonly used nitrous oxide [4].
in the UK [6163], it may be used in other countries. Although the evidence was of low quality overall,
However, our sensitivity analysis showed nitrous oxide based on the literature, TIVA appears to be preferable to
had no impact on hospital admission, but may have inhalational anaesthesia in patients undergoing ambula-
affected PONV. Patients undergoing inducution of tory surgery with regards to preventing PONV, but no
anaesthesia with inhalational agents were included, other major advantages were discovered. When hospital
and this could affect the primary endpoints since in a admissions due to surgical causes were excluded, there
few patients, induction was with desurane, an airway was no evidence of a difference in hospital admission
irritant and hence potentially emetogenic [64, 65]. between TIVA and inhalational anaesthesia. However,
This could have decreased the effect of propofol on given that there was an almost threefold difference in
PONV. However, we have presented the results from unplanned hospital admissions between TIVA and inha-
pragmatic trials that reect real-life situations and the lational anaesthesia (1.0% vs 2.9%, respectively) further
inuence of the induction method on the ability of robust randomised controlled trials are needed to iden-
propofol to reduce PONV should be further investi- tify whether a truly signicant difference exists.
gated. Perhaps the biggest limitation is the fact that
only three of the included trials were conducted Competing interests
within the past 10 years, with the majority in the No competing interests declared.
1990s. These older trials may have more inuence on
outcomes since they were conducted at a time when References
sevourane, desurane and TIVA were relatively new 1. Aylin P, Williams S, Jarman B, et al. Trends in day surgery
rates. British Medical Journal 2005; 331: 803.
and thus less rened, potentially producing more 2. Owings MF, Kozak LJ. Ambulatory and inpatient procedures in
adverse outcomes than may be expected in the mod- the United States, 1996. Vital and Health Statistics. Data from
the National Health Survey 1998; 13: 1119.
ern era. However, it was necessary to include all these 3. Kozak LJ, Owings MF, Hall MJ. National Hospital Discharge
trials due to their abundance and large trial numbers, Survey: 2002 annual summary with detailed diagnosis and
increasing the likelihood of producing signicant procedure data. Vital and Health Statistics. Data from the
National Health Survey 2005; 13: 1199.
results. 4. Gupta A, Stierer T, Zuckerman R, et al. Comparison of recovery
There was no evidence of heterogeneity in the pri- profile after ambulatory anesthesia with propofol, isoflurane,
sevoflurane and desflurane: a systematic review. Anesthesia
mary outcome of unplanned hospital admission, but for and Analgesia 2004; 98: 63241.
all other outcomes signicant heterogeneity was 5. Dexter F, Tinker JH. Comparisons between desflurane and
detected, with poor overlapping of condence intervals isoflurane or propofol on time to following commands and time
to discharge. A meta-analysis. Anesthesiology 1995; 83:
even within the subgroups, decreasing condence in the 7782.
effect estimate. A potential reason could be the type of 6. Shirakami G, Teratani Y, Namba T, et al. Delayed discharge
and acceptability of ambulatory surgery in adult outpatients
surgery, but there were not enough trials to investigate receiving general anesthesia. Journal of Anesthesia 2005; 19:
the inuence of each surgical specialty on the effective- 93101.
ness of TIVA. There was no evidence of publication bias 7. Green G, Jonsson L. Nausea: the most important factor deter-
mining length of stay after ambulatory anaesthesia. A com-
for the two outcomes where we could explore them. parative study of isoflurane and/or propofol techniques. Acta
The majority of trials in this analysis are relatively Anaesthesiologica Scandinavica 1993; 37: 7426.
8. Macario A, Weinger M, Carney S, et al. Which clinical anesthe-
small. A power calculation has identied that a large, sia outcomes are important to avoid? The perspective of
double-blinded randomised controlled trial of at least patients. Anesthesia and Analgesia 1999; 89: 6528.
820 participants in each of the TIVA and inhalational 9. McGrath B, Elgendy H, Chung F, et al. Thirty percent of
patients have moderate to severe pain 24 hr after ambulatory
groups would be needed to corroborate the results surgery: a survey of 5,703 patients. Canadian Journal of Anes-
found here with a condence of 95%. This is the rst thesia 2004; 51: 88691.
10. Rawal N, Hylander J, Nydahl PA, et al. Survey of postopera- undergoing ambulatory gynecological surgery. Acta Anaes-
tive analgesia following ambulatory surgery. Acta Anaesthesi- thesiologica Taiwanica 2006; 44: 1017.
ologica Scandinavica 1997; 41: 101722. 27. Fish WH, Hobbs AJ, Daniels MV. Comparison of sevoflurane
11. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items and total intravenous anaesthesia for daycase urological sur-
for systematic reviews and meta-analyses: the PRISMA state- gery. Anaesthesia 1999; 54: 10026.
ment. PLoS Medicine 2009; 6: e1000097. 28. Fredman B, Nathanson MH, Smith I, et al. Sevoflurane for out-
12. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement patient anesthesia: a comparison with propofol. Anesthesia
for reporting systematic reviews and meta-analyses of studies and Analgesia 1995; 81: 8238.
that evaluate healthcare interventions: explanation and elab- 29. van Hemelrijck J, Smith I, White PF. Use of desflurane for out-
oration. British Medical Journal 2009; 339: b2700. patient anesthesia. A comparison with propofol and nitrous
13. Visser K, Hassink EA, Bonsel GJ, et al. Randomized controlled oxide. Anesthesiology 1991; 75: 197203.
trial of total intravenous anesthesia with propofol versus 30. Jakobsson J, Rane K, Ryberg G. Anaesthesia during laparo-
inhalation anesthesia with isoflurane-nitrous oxide: postoper- scopic gynaecological surgery: a comparison between desflu-
ative nausea with vomiting and economic analysis. Anesthesi- rane and isoflurane. European Journal of Anaesthesiology
ology 2001; 95: 61626. 1997; 14: 14852.
14. Trame r M, Moore A, McQuay H. Meta-analytic comparison of 31. Kurpiers EM, Scharine J, Lovell SL. Cost-effective anesthesia:
prophylactic antiemetic efficacy for postoperative nausea and desflurane versus propofol in outpatient surgery. American
vomiting: propofol anaesthesia vs omitting nitrous oxide vs Association of Nurse Anaesthetists Journal 1996; 64: 6975.
total i.v. anaesthesia with propofol. British Journal of Anaes- 32. Lebenbom-Mansour MH, Pandit SK, Kothary SP, et al. Desflu-
thesia 1997; 78: 2569. rane versus propofol anesthesia: a comparative analysis in
15. Divatia JV, Vaidya JS, Badwe RA, et al. Omission of nitrous outpatients. Anesthesia and Analgesia 1993; 76: 93641.
oxide during anesthesia reduces the incidence of postopera- 33. Nathan N, Peyclit A, Lahrimi A, et al. Comparison of
tive nausea and vomiting. A meta-analysis. Anesthesiology sevoflurane and propofol for ambulatory anaesthesia in
1996; 85: 105562. gynaecological surgery. Canadian Journal of Anesthesia 1998;
16. Gurusamy KS, Gluud C, Nikolova D, et al. Assessment of risk 45: 114850.
of bias in randomized clinical trials in surgery. British Journal 34. Raeder J, Gupta A, Pedersen FM. Recovery characteristics of
of Surgery 2009; 96: 3429. sevoflurane- or propofol-based anaesthesia for day-care sur-
17. Higgins J, Altman G. Assessing risk of bias in included studies. gery. Acta Anaesthesiologica Scandinavica 1997; 41: 98894.
In: Higgins JP, Green S, eds. Cochrane Handbook for System- 35. Raeder JC, Mjaland O, Aasb V, et al. Desflurane versus propo-
atic Reviews of Interventions. Chichester: Wiley-Blackwell, fol maintenance for outpatient laparoscopic cholecystectomy.
2008: 187242. Acta Anaesthesiologica Scandinavica 1998; 42: 10610.
18. Wood L, Egger M, Gluud LL, et al. Empirical evidence of bias 36. Rapp SE, Conahan TJ, Pavlin DJ, et al. Comparison of desflurane
in treatment effect estimates in controlled trials with different with propofol in outpatients undergoing peripheral orthopedic
interventions and outcomes: meta-epidemiological study. Brit- surgery. Anesthesia and Analgesia 1992; 75: 5729.
ish Medical Journal 2008; 336: 6015. 37. Smith I, Terhoeve PA, Hennart D, et al. A multicentre compari-
19. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic son of the costs of anaesthesia with sevoflurane or propofol.
quality and discrepancies between large and small random- British Journal of Anaesthesia 1999; 83: 56470.
ized trials in meta-analyses. Annals of Internal Medicine 38. Smith I, Thwaites AJ. Target-controlled propofol vs. sevoflura-
2001; 135: 9829. ne: a double-blind, randomised comparison in day-case
20. Higgins J, Altman G. Assessing risk of bias in included studies. anaesthesia. Anaesthesia 1999; 54: 74552.
In: Higgins JP, Green S, eds. Cochrane Handbook for System- 39. Stevanovic PD, Petrova G, Miljkovic B, et al. Low fresh gas
atic Reviews of Interventions. Chichester: Wiley-Blackwell, flow balanced anesthesia versus target controlled intravenous
2008: 2304. infusion anesthesia in laparoscopic cholecystectomy: a cost-
21. Macaskill P, Walter SD, Irwig L. A comparison of methods to minimization analysis. Clinical Therapeutics 2008; 30: 1714
detect publication bias in meta-analysis. Statistics in Medicine 25.
2001; 20: 64154. 40. Tan T, Bhinder R, Carey M, et al. Day-surgery patients anes-
22. Deeks J, Higgins J. Statistical algorithms in Review Manager 5, thetized with propofol have less postoperative pain than
2010. www.cochrane.org/sites/default/files/uploads/hand- those anesthetized with sevoflurane. Anesthesia and Analge-
book/Statistical_Methods_in_RevMan5-1.pdf (accessed 21/ sia 2010; 111: 835.
12/2013). 41. Tang J, White PF, Wender RH, et al. Fast-track office-based
23. Cook DJ, Sackett DL, Spitzer WO. Methodologic guidelines for anesthesia: a comparison of propofol versus desflurane with
systematic reviews of randomized control trials in health care antiemetic prophylaxis in spontaneously breathing patients.
from the Potsdam Consultation on Meta-Analysis. Journal of Anesthesia and Analgesia 2001; 92: 959.
Clinical Epidemiology 1995; 48: 16771. 42. White H, Black RJ, Jones M, et al. Randomized comparison of
24. Hatala R, Keitz S, Wyer P, et al. Tips for learners of evidence- two anti-emetic strategies in high-risk patients undergoing
based medicine: 4. Assessing heterogeneity of primary studies day-case gynaecological surgery. British Journal of Anaesthe-
in systematic reviews and whether to combine their results. sia 2007; 98: 4706.
Canadian Medical Association Journal 2005; 172: 6615. 43. Rees S, Stocker M, Montgomery J. Paediatric outcomes in a
25. Ashworth J, Smith I. Comparison of desflurane with isoflurane District General Hospital day surgery unit. Journal of One-Day
or propofol in spontaneously breathing ambulatory patients. Surgery 2009; 19: 925.
Anesthesia and Analgesia 1998; 87: 3128. 44. Zulfiquer F, Pattanayak K. Evaluation of unplanned admission
26. Chen C, Yang Y, Chen W, et al. Comparison of the anesthesia following day surgery at a new surgical centre in London.
profiles between sevoflurane-nitrous oxide and propofol- 2009. www.iaas-med.com/files/Journal/15/15_4/Zulfiquer.
nitrous oxide conveyed by laryngeal mask airway in patients pdf (accessed 21/12/2013).
45. Colvin J, Peden C. Raising the standard: a compendium of experimentally induced tourniquet pain. American Association
audit recipes. The Royal College of Anaesthetists 2012. www. of Nurse Anesthetists Journal 2001; 69: 46670.
rcoa.ac.uk/system/files/CSQ-ARB-2012_1.pdf (accessed 21/ 56. Fassoulaki A. Is propofol an analgesic? European Journal of
12/2013). Anaesthesiology 2011; 28: 4812.
46. Lagasse RS. Anesthesia safety: model or myth? A review of 57. Weinger MB. Drug wastage contributes significantly to the
the published literature and analysis of current original data. cost of routine anesthesia care. Journal of Clinical Anesthesia
Anesthesiology 2002; 97: 160917. 2001; 13: 4917.
47. Martin TM, Nicolson SC, Bargas MS. Propofol anesthesia 58. Ravenelle F, Gori S, Le Garrec D, et al. Novel lipid and preser-
reduces emesis and airway obstruction in pediatric outpa- vative-free propofol formulation: properties and pharmacody-
tients. Anesthesia and Analgesia 1993; 76: 1448. namics. Pharmaceutical Research 2008; 25: 3139.
48. Watcha MF. Postoperative nausea and emesis. Anesthesiology 59. Eikaas H, Raeder J. Total intravenous anaesthesia techniques
Clinics of North America 2002; 20: 70922. for ambulatory surgery. Current Opinion in Anesthesiology
49. Watcha MF, White PF. Postoperative nausea and vomiting. Its 2009; 22: 7259.
etiology, treatment, and prevention. Anesthesiology 1992; 60. Motsch J, Wandel C, Neff S, et al. A comparative study of the
77: 16284. use of sevoflurane and propofol in ambulatory surgery. Der
50. Carroll NV, Miederhoff PA, Cox FM, et al. Costs incurred by Anaesthesist 1996; 45: S5762.
outpatient surgical centers in managing postoperative nau- 61. Turan A, Mascha EJ, You J, et al. The association between nitrous
sea and vomiting. Journal of Clinical Anesthesia 1994; 6: oxide and postoperative mortality and morbidity after noncardiac
3649. surgery. Anesthesia and Analgesia 2013; 116: 102633.
51. Wu CL, Berenholtz SM, Pronovost PJ, et al. Systematic review 62. Leslie K, Myles PS, Chan MT, et al. Nitrous oxide and long-
and analysis of postdischarge symptoms after outpatient sur- term morbidity and mortality in the ENIGMA trial. Anesthesia
gery. Anesthesiology 2002; 96: 9941003. and Analgesia 2011; 112: 38793.
52. Soppitt AJ, Glass PS, Howell S, et al. The use of propofol for 63. Baum VC, Willschke H, Marciniak B. Is nitrous oxide necessary
its antiemetic effect: a survey of clinical practice in the United in the future? Pediatric Anesthesia 2012; 22: 9817.
States. Journal of Clinical Anesthesia 2000; 12: 2659. 64. Eilers H, Cattaruzza F, Nassini R, et al. Pungent general anes-
53. DeBalli P. The use of propofol as an antiemetic. International thetics activate transient receptor potential-A1 to produce
Anesthesiology Clinics 2003; 41: 6777. hyperalgesia and neurogenic bronchoconstriction. Anesthesiol-
54. Bandschapp O, Filitz J, Ihmsen H, et al. Analgesic and antihy- ogy 2010; 112: 145263.
peralgesic properties of propofol in a human pain model. 65. White PF, Tang J, Wender RH, et al. Desflurane versus sevoflu-
Anesthesiology 2010; 113: 4218. rane for maintenance of outpatient anesthesia: the effect on
55. Hand R Jr, Riley GP, Nick ML, et al. The analgesic effects of early versus late recovery and perioperative coughing. Anes-
subhypnotic doses of propofol in human volunteers with thesia and Analgesia 2009; 109: 38793.