Does anaesthesia with nitrous oxide affect mortality or cardiovascular morbidity? A systematic review with meta-analysis and trial sequential analysis

British Journal of Anaesthesia Page 1 of 17 doi:10.1093/bja/aet416

5

Does anaesthesia with nitrous oxide affect mortality or cardiovascular morbidity? A systematic review with meta-analysis and trial sequential analysis

60

G. Imberger1*, A. Orr 2, K. Thorlund 3, J. Wetterslev 1, P. Myles2,4 and A. M. Møller 5,6 65 10

1

Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Blegdamsvej 9, Copenhagen Ø DK-2100, Denmark Department of Anaesthesia and Perioperative Medicine, Alfred Hospital, 55 Commercial Road, Melbourne 3004, Australia 3 Department of Clinical Epidemiology and Biostatistics, McMaster University, 1280 Main Street West, Hamilton L8S4L8, Canada 4 Academic Board of Anaesthesia and Perioperative Medicine, Monash University, 55 Commercial Road, Melbourne 3004, Australia 5 Cochrane Anaesthesia Review Group, Rigshospitalet, Blegdamsvej 9, Copenhagen Ø DK-2100, Denmark 6 Department of Anaesthesia, Herlev Hospital, Herlev DK-2730, Denmark 2

15

70

* Corresponding author. E-mail: [email protected]

Editor’s key-points 20

25

30

35

† The authors reviewed publications exploring cardiovascular outcomes after nitrous oxide anaesthesia. † The evidence-base was insufficient to draw robust conclusions regarding the effect of nitrous oxide anaesthesia on cardiovascular outcomes.

Background. The role of nitrous oxide in modern anaesthetic practice is contentious. One concern is that exposure to nitrous oxide may increase the risk of cardiovascular complications. ENIGMA II is a large randomized clinical trial currently underway which is investigating nitrous oxide and cardiovascular complications. Before the completion of this trial, we performed a systematic review and meta-analysis, using Cochrane methodology, on the outcomes that make up the composite primary outcome.

80

Methods. We used conventional meta-analysis and trial sequential analysis (TSA). We reviewed 8282 abstracts and selected 138 that fulfilled our criteria for study type, population, and intervention. We attempted to contact the authors of all the selected publications to check for unpublished outcome data. Results. Thirteen trials had outcome data eligible for our outcomes. We assessed three of these trials as having a low riskof bias. Using conventional meta-analysis, the relative riskof short-term mortality in the nitrous oxide group was 1.38 [95% confidence interval (CI) 0.22–8.71] and the relative risk of long-term mortality in the nitrous oxide group was 0.94 (95% CI 0.80–1.10). In both cases, TSA demonstrated that the data were far too sparse to make any conclusions. There were insufficient data to perform meta-analysis for stroke, myocardial infarct, pulmonary embolus, or cardiac arrest. Conclusion. This systematic review demonstrated that we currently do not have robust evidence for how nitrous oxide used as part of general anaesthesia affects mortality and cardiovascular complications.

40

75

85

90

95

Keywords: meta-analysis; nitrous oxide; review, systematic Accepted for publication: 17 July 2013

45

50

55

Nitrous oxide has been used as a general anaesthetic for more than 160 years. Collective anecdotal experience with this drug must be larger than with any other drug used in anaesthesia. Despite this experience, opinions about the role of nitrous oxide in modern-day practice continue to diverge.1 2 One concern is that exposure to nitrous oxide may increase the risk of cardiovascular complications. Nitrous oxide oxidizes the cobalt atom in vitamin B12, inactivating methionine synthase, causing a decrease in folate metabolism and an increase in homocysteine.3 Homocysteinaemia after exposure to nitrous oxide has been well demonstrated in vivo4 – 6 and long-term homocysteinaemia is known to be associated with an increased risk of ischaemic heart disease.7 It remains

unclear, however, whether this information about this surrogate outcome translates into real clinical risk. To investigate the possible causal association between mortality, cardiac morbidity, and nitrous oxide, the ENIGMA (Evaluation of Nitrous oxide In the Gas Mixture for Anaesthesia) trial group has designed a large, multi-centre randomized clinical trial: ENIGMA II is enrolling at-risk patients and is powered to investigate a composite primary outcome of mortality, non-fatal acute myocardial infarction, cardiac arrest, pulmonary embolism, and stroke.8 While this trial is underway, several observational studies have been published looking at similar outcomes. Using data from the Intraoperative Hypothermia in Aneurysm Surgery

& The Author [2013]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please email: [email protected]

100

105

110

BJA 115

120

125

130

135

140

145

150

Trial,9 a secondary analysis was performed, finding an association between nitrous oxide exposure and delayed ischaemic neurological deficits.10 Sanders and colleagues11 also performed a secondary subgroup analysis on 1615 participants from the General Anaesthesia compared with Local Anaesthesia for Carotid Surgery Trial, finding no evidence that nitrous oxide increases the risk of perioperative vascular adverse events.12 Turan and colleagues13 published a retrospective cohort study of 49,016 patients and found that patients who had received nitrous oxide had decreased odds for 30-day mortality. Leslie and colleagues14 performed a post hoc analysis on 5133 of the patients from the Perioperative Ischemic Evaluation trial, and found no association between nitrous oxide and cardiovascular adverse events. These studies herald a focusing interest on the association between nitrous oxide and cardiovascular adverse events and they certainly contribute to the debate. Being post hoc observational studies, however, potential confounders, especially those confounding by indication, increase the risk of bias and prevent any conclusions from being definitive. The association between nitrous oxide and cardiovascular complications remains unclear, and the results from ENIGMA II are keenly awaited. We thus performed a systematic review and meta-analysis of the five outcomes used as the composite primary outcome in ENIGMA II. We plan to update this meta-analysis when the results from ENIGMA II, and any other future trials, are available. Repeated updates in meta-analysis are analogous to those in interim analyses in a clinical trial. In clinical trials where interim analyses are performed, the concern about increased risk of type 1 error because of repetitive testing and sparse data is well known and sequential hypothesis testing designs and procedures are used to control this increased risk.15 16 In systematic reviews with meta-analysis, similar sequential hypothesis testing procedures, such as Trial Sequential Analysis (TSA), can also be applied.17 – 21 Given our plan to update our meta-analysis in the future, we performed TSA as part of our current analysis.

Methods Protocol and registration

155

160

165

The protocol for the review was published on the website of the Copenhagen Trial Unit (www.ctu.dk) in January 2011. The protocol was registered with the PROSPERO database in December 2011 (registration no CRD42011001831). There were two small deviations from this protocol. First, there were errors in the search strategies published in the protocol. These errors were corrected before the searches were run. The corrected searches are provided in Appendix 1. Secondly, we clarified the definition of patients at increased risk of cardiovascular complications slightly, by stating that we considered patients undergoing day surgery as being at low risk of cardiovascular complications.

Systematic literature search We conducted a sensitive systematic literature search of Medline, EMBASE, the Cochrane controlled trial register, CINAHL, and ISI Web of Science, updated until May 2012.

Page 2 of 17

Imberger et al.

There were no date or language restrictions. We also checked the references of included studies. Two authors (G.I. and A.O.) independently screened all of the abstracts produced by the search to identify eligible studies.

Study selection, data extraction, and quality assessment We included all randomized clinical trials, irrespective of language or publication status. We included trials with human adult patients receiving general anaesthesia, for any surgery. We defined a general anaesthetic as any procedure where inhalation agents, systemic agents, or both are given as part of general anaesthesia for the purposes of undertaking a medical procedure. We did not include studies where nitrous oxide was given for the purposes of sedation. We included trials where patients receiving nitrous oxide were compared with patients receiving no nitrous oxide. We included trials only when it was clear that the control group received no nitrous oxide throughout the perioperative period. We excluded trials where participants were randomized to different anaesthetic techniques (apart from the administration of nitrous oxide). We included studies that were able to provide data on any of the following outcomes: (1) Mortality (all cause)—including all trials where mortality data were available for the same follow-up period in both the exposure and control groups. We included two comparisons for the mortality outcome: † Mortality—short term, follow-up ranging from discharge from the perioperative care unit (PACU) until 30 days after operation (or discharge from hospital). The end point needed to be consistent in the two intervention groups. † Mortality—long term, including the longest follow-up starting from 30 days after operation. (2) Stroke. (3) Myocardial infarction (MI). (4) Pulmonary embolus (PE). (5) Cardiac arrest. For short-term mortality, stroke, MI, PE, and cardiac arrest, we included data from any trials where the patients were at increased risk of cardiovascular complications. See Appendix 2 for our definition of this increased risk. We accepted all clear and reasonable definitions of these outcomes in individual trials, as long as they were used consistently in both groups and reported explicitly. We selected possible inclusions on the basis of the trial type, participants and interventions described in the abstract and retrieved full copies of these publications. For trials that fulfilled these parameters, but did not report relevant outcomes, we wrote to the authors to ask whether any unpublished data were available. When no email address was available on the publication, we searched for previous contact email addresses for authors and we contacted departments, co-authors of other publications, or both, in order to find contact details. In

170

175

180

185

190

195

200

205

210

215

220

BJA

Nitrous oxide and cardiovascular outcomes

225

230

235

240

245

250

the case of mortality, when an author specified that all patients were followed up after leaving PACU, either until a specific time point or until discharge, and that they could confirm that there was no mortality in either group for this period, then we selected these trials for inclusion. In the case of stroke, MI, PE, and cardiac arrest, we only included trials for inclusion if the authors confirmed that these outcomes had been formally registered in each group. Two authors (G.I. and A.O.) independently extracted data from the included trials, including information about the methodology and the data for our outcomes. All duplicate publications were examined to ensure consistency. For the studies with relevant unpublished outcome data, extraction came from the email correspondence with authors. Two authors (G.I. and A.O.) reviewed all included trials with regard to their quality. We rated the risk of bias using the guide provided in The Cochrane Handbook of Systematic Reviews of Interventions.22 We made the assessments with regard to each outcome being included in our review, so these assessments are not necessarily reflective of the quality of the trials for their intended primary outcomes. We assessed each outcome for sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias. When all of the five assessment categories were assessed as adequate for a given outcome, we assigned a low risk of bias for the findings for that outcome. When we assessed one or more categories as inadequate, we assigned an increased risk of bias.

Statistical analysis 255

260

265

270

275

280

All of our outcomes were categorical and were expressed as risk ratios (RRs) with 95% confidence intervals (CIs). We performed meta-analysis when more than one trial with low risk of bias was eligible for inclusion for a given outcome. All testing was two-sided. To control the increased risk of random error because of sparse data and repetitive testing, we also performed TSA. Conventional meta-analysis was performed using the Review Manager software (RevMan 5.0). TSA was performed using the TSA software (www.ctu.dk/tsa).

Trial sequential analysis Repeated updates (sequential multiplicity) and sparse data increase the risk of random error.23 TSA is a method for metaanalysis that aims to correct for this increased risk.17 – 20 In a way similar to monitoring boundaries for interim analyses in single trials, TSA combines an estimation of required information size (RIS) for meta-analysis with monitoring boundaries used as thresholds for statistical significance. The less data that have accumulated, the more conservative the TSA boundaries, making it less likely to declare statistical significance before the RIS has been reached. Similar to a sample size calculation for a single trial, estimating RIS involves a calculation that includes type 1 error, type 2 error, the control event proportion, and the effect size. The calculation for RIS also requires an estimate of heterogeneity; if more heterogeneity is present, RIS increases.21 For our TSAs,

we estimated the RIS using 0.05 for type 1 error, 0.10 for type 2 error, the control event proportions calculated from the no nitrous oxide groups in all the included trials and the effect size estimated from the included trials with a low risk of bias. We repeated the analysis using relative risk reductions of 10 and 20% as effect size estimates. We used the D 2 (Diversity)24 present in the included trials as the estimate for heterogeneity. The TSA can be interpreted by viewing the boundaries and whether the cumulative meta-analysis has crossed them. Alternatively, the results can be translated into TSA-adjusted CIs.21 In this review, we presented the TSA-adjusted CIs.

285

290

Sensitivity analyses When statistical heterogeneity was present, we presented the results of random-effects meta-analysis [DerSimonian and Laird (DL)].25 When no statistical heterogeneity was present, we presented the results of fixed-effect meta-analysis. Groups with zero events were adjusted with a constant continuity adjustment of 0.5 in each arm (as per the default adjustment in the Revman software).26 A sensitivity analysis was performed using both analysis models (random effects and fixed effect), different continuity adjustments (constant and empirical),27 and using a Peto fixed-effect model.28

295

300

305

Results Figure 1 shows the study flow diagram. From the 8282 abstracts reviewed, we considered 138 publications eligible on the basis of the trial type, the participants and the intervention. Of note, none of these publications had any of our outcomes as primary outcomes. Table 1 shows the primary outcomes described in these publications. In 21 (14 being duplicates), authors reported our outcomes as secondary findings. Of the 117 publications that did not mention any of our outcomes, authors provided unpublished but eligible information about mortality from nine publications (three being duplicates). In total, 13 trials were selected as eligible to be included in our meta-analyses. The characteristics of the included studies are summarized in Table 2. Risk-of-bias assessments are presented in Table 3.

310

315

320

Synthesis of results Mortality—short term

325

Figure 2 shows a forest plot for the conventional meta-analysis for this outcome. Including only the trials with low risk of bias,37 83 160 the relative risk in the nitrous group was 1.38 (95% CI 0.22– 8.71). For TSA, using a control event proportion of 0.5%, a relative risk increase of 40%, and a constant continuity adjustment of 0.5 events per group, the accrued information size (2648) was only 1.63% of the estimated RIS (162802). Consequently, there was too little information for the TSA software to construct boundaries and calculate TSA-adjusted 95% CI. Sensitivity analyses were performed using both analysis models (random effects and fixed effect), different continuity

Page 3 of 17

330

335

BJA

Imberger et al.

395

8282 titles and abstracts reviewed

340

400 345

146 selected as potentially eligible and full papers reviewed

8 excluded based on study type, participants or intervention 405

350

138 selected as eligible based on study type, participants and intervention

117 publications not reporting relevant outcomes 410

355

21 publications reporting relevant outcomes

360

365

7 eligible trials (+14 duplicates)

26—unable to find a contact email address 37—email sent, no reply 45—email sent, reply stating that no relevant outcomes were measured 9—email sent, reply giving relevant outcomes that were measured but not published

415

420

6 eligible trials (+3 duplicates)

425 370

13 randomized clinical trials included

430 375

380

385

390

Fig 1 Study flow diagram.

adjustments (constant and empirical), and using a Peto fixed-effect model. The point estimates (and corresponding 95% CIs), the calculated heterogeneity (I 2) and the consequential estimation of RIS all varied substantially when using different models. This variation is summarized in Table 4.

Mortality—long-term For both trials included in this meta-analysis,83 160 these longterm data were extracted from follow-up publications.35 87 Using conventional meta-analysis, the relative risk in the nitrous group was 0.94 (95% CI 0.80 –1.10), as shown in Figure 3. For TSA, the control event proportion used was 25% and the relative risk reduction 6%. The relative risk in the nitrous group was 0.94 (TSA adjusted 95% CI 0.49–1.79), the TSA did not

Page 4 of 17

cross the futility threshold for rejecting a 6% relative risk reduction, and the accrued information size (1873) was only 6% of the estimated RIS (31 699). Using a relative risk reduction of 10%, the point estimate and CIs were the same, the futility threshold was not crossed and the accrued information was 15% of the estimated RIS (12 179). Using a relative risk reduction of 20%, the point estimate was 0.94 (TSA adjusted 95% CI 0.76–1.15) and the TSA did cross the futility threshold. This result indicates that exposure to nitrous oxide does not alter long-term mortality by 20%. See Figure 4 for this analysis.

Stroke Figure 5 shows a forest plot for stroke. There was only one trial with low risk of bias.83 Meta-analysis was not performed.

435

440

445

BJA

Nitrous oxide and cardiovascular outcomes

450

Table 1 Primary outcomes studied in the publications eligible based on study type, population, and intervention Primary outcome(s)

1989 – 200329 – 31

Anaesthetic depth

3

32 – 34

1999 – 2006

Cancer recurrence

1

200935

Cognitive function

2

2006 – 201136 37

Cost

2

1996 – 201138 39

15

1987 – 201140 – 54

Drug kinetics 460

Folate/homocysteine metabolism General tolerance

465

470

5 25

4 – 6 55 56

1990 – 2010

1990 – 201257 – 81

IVF

1

198782

Length of stay

1

200783

Laryngeal mask related endpoints

3

2000 – 200484 – 86 87

Long-term morbidity and mortality

1

2011

Myocardial ischaemia

3

1990 – 200088 – 90

Neuropsychology

1

199491

Operating conditions

6

1987 – 200792 – 97

Pain

3

201198 – 100

Cardiology

7

1990 – 2005101 – 107

Neurology

9

1993 – 2010108 – 116

Respiratory

12

1977 – 2010117 – 128

9

1985 – 2002129 – 137

23

138 – 160

Physiological parameters

Other 475

Years of publication (range)

3

Air-filled spaces 455

n

PONV

1986 – 2008

Viability of bone marrow cells

1

1995161

Wound infection

2

2005162 163

480

Myocardial infarction Four trials registered data for this outcome.61 76 83 89 There was only one trial with low risk of bias.83 Meta-analysis was not performed (Fig. 6). 485

Pulmonary embolus Only one trial reported data on PE.76 There were zero events in each group. Meta-analysis was not performed. 490

Cardiac arrest We found no data for this outcome.

Discussion 495

500

The purpose of this systematic review was to summarize the current evidence from randomized clinical trials associating nitrous oxide with serious cardiovascular complications. Our review presents results using both conventional meta-analysis and TSA. We were able to conduct meta-analyses for shortterm mortality and long-term mortality. For the other four outcomes, there was no more than one trial with low risk of bias, and we therefore did not conduct meta-analyses. In the case of short-term mortality, the conventional meta-analysis (with low risk of bias trials) gives a point estimate indicating increased risk in the nitrous oxide group (RR

1.38) but with very wide CIs (95% CI 0.22– 8.71). There is huge imprecision in the estimate provided by this evidence. This imprecision, combined with the low control event proportion and the relatively small number of patients included, means that we cannot say that the two groups are different for this outcome. Nor can we say that they are similar. For TSA, the accrued information was such a small percentage of the estimated RIS (,2%) that the boundaries were unable to be constructed. That is, the quantity of the information was too small to even be analysed with TSA. We conclude, therefore, that there is currently insufficient evidence to assess whether nitrous oxide increases the risk of short-term mortality in high-risk groups having surgery. For the meta-analysis for short-term mortality, events were sparse and there were zero-event groups in two of the three included trials. In this situation (sparse data and zero-event groups), the choice of continuity adjustment can have a large impact on the estimate of effect (RR) and on the calculated heterogeneity (I 2). I 2 affects the decision about which metaanalytic model to use, which in turn can have a large impact on the estimate of effect when the data are sparse. Each combination of effect measure, continuity adjustment, and meta-analytic model represents a different way to model the situation. When data are sparse, it is difficult to judge which model is most appropriate. We, therefore, did sensitivity analyses using each different model. The Cochrane handbook suggests using the Peto odds ratio (with fixed-effect metaanalysis) for zero-event data,26 and we included this in the sensitivity analysis. An extensive variation in point estimates, calculated heterogeneity (and consequently estimated RIS) was demonstrated. This variation reflects the enormous amount of uncertainty, based on these data, of the effect of nitrous oxide on short-term mortality. Long-term mortality provides an example of how TSA can assist with interpreting the results of meta-analysis. Using conventional techniques, the point estimate was close to 1 (RR 0.94) with rather narrow CIs (95% CI 0.80–1.10). One might be tempted to conclude, on the basis of these results, that there is no difference in long-term mortality with or without nitrous oxide for this population. For the TSA for this outcome, we used a control event proportion of 25% and a relative risk reduction of 6% for the effect size (derived from included trials with a low risk of bias). To test this model, the RIS was large (24 627). However, the number of randomized patients actually accrued in this meta-analysis was ,10% (1873). So concluding no effect here would be similar to stopping a trial after only 10% of the sample size had been recruited, finding narrow CIs that crossed 1.0 and concluded that an intervention had no effect (despite being grossly underpowered). Indeed, the futility boundaries constructed with TSA were not crossed, and we conclude that there is insufficient evidence to assess whether nitrous oxide increases or decreases long-term mortality by an effect estimate of 6%. When we tested this model using effect estimates of relative risk reductions of 10%, there was also insufficient evidence. For a relative risk reduction of 20%, however, the TSA just crossed the futility boundary. The population of patients included in this

Page 5 of 17

505

510

515

520

525

530

535

540

545

550

555

560

565

570

575

580

585

590

595

600

605

610

615

BJA

Page 6 of 17

Table 2 Characteristics of the included trials. *Published in Fleischmann and colleagues.162 †Published in Leslie and colleagues148 Trial

Other publications reporting the same study

Intervention

Control

Outcomes for inclusion and number of participants able to be included

Primary outcome(s)

Antonini and colleagues58

40 patients having laparoscopic cholecystectomy

N2O and O2

Air and O2

Mortality—short (n¼40) Unpublished

Multiple

FIO2 0.4

FIO2 0.4

Chowdhury and colleagues36

90 patients having transphenoidal removal of pituitary tumours

O2 and N2O 1:2

40% O2 in air

Mortality—short (n¼85) Unpublished

Postoperative cognitive defects

100—elective total hip arthroscopy 70—carotid endarterectomy 100—transphenoidal hypophysectomy

60% N2O and 40% O2

100% O2

Mortality—short (n¼270) MI (n¼70) Published

Multiple postoperative complications

28 patients having laparoscopic donor nephrectomies

70% N2O in O2

Air and O2, mixture

Mortality—short (n¼28) Unpublished

Bowel distension

418 ASA I – III patients having elective colon resection, expected to last longer than 2 h

65% N2O and 35% O2

35% O2 in N2

Mortality—short (n¼408) Mortality—long (n¼203)* Published

Surgical wound infection Bowel distension

Hohner and colleagues89

49 patients with coronary artery disease having abdominal aorta surgery

60% N2O and 40% O2 (and 5 –10 mg fentanyl kg21)

40% O2, in air (and 15 – 20 mg fentanyl kg21)

Mortality—short (n¼47) MI (n¼47) Published

Haemodynamics Ventricular function Myocardial ischaemia

Iacopino and colleagues112

40 ASA I and II patients having neurosurgical procedures

30% O2 and 67% N2O

30% O2 and 66% N2

Mortality—short (n¼40) Unpublished

Cerebral autoregulation

40 patients having intra-abdominal operations on the colon and rectum

30% O2 in N2O (and higher doses of propofol and fentanyl)

Oxygen in air with an FIO2 of 0.3

Mortality short (n¼40) Unpublished

Bowel function Atelectasis Recovery

Knu¨ttgen and colleagues30

42 patients having intracranial procedures in the sitting position

O2 and N2O 1:1

O2 and N2 1:1

Mortality—short (n¼42) Unpublished

Incidence of venous air embolism

Krogh and colleagues97

139 patients having elective major colonic surgery

N2O in O2 with an F IO2 of 0.3

Air in oxygen with an FIO2 of 0.3

Mortality—short (n¼139) Unpublished

Operating conditions Postoperative course

Leung and colleagues37

228 patients aged .65 yr, having any surgical procedure and expected to stay in hospital for .48 h

N2O and O2 (% at anaesthetist’s discretion) no difference in F IO2 between 2 groups—P 0.48)

O2 with/without air (% at anaesthetist’s discretion)

Mortality—short (n¼210) Published

Postoperative delirium or cognitive decline

Eger and colleagues61

Koblin and colleagues55 Kozmary and colleagues88 Lampe and colleagues132 Lampe and colleagues104 Lampe and colleagues125 Waldman and colleagues137

El-Galley and colleagues95 Fleischmann and colleagues162

Jensen and colleagues131

Akca and colleagues92 Fleischmann and colleagues35

Jensen and colleagues122 Jensen and colleagues66 Kalman and colleagues67

Imberger et al.

Participants Randomized (nitrous vs no nitrous)

620

625

630

635

640

645

650

655

660

665

670

BJA

Nitrous oxide and cardiovascular outcomes

Length of stay (hospital and ICU) Mortality—short (n¼87) Stroke (n¼87) MI (n¼87) PE (n¼87) Published

685

Length of stay (hospital)

680

Mortality—short (n¼2012) Stroke (n¼2012) MI (n¼2012) Mortality—long (n¼1670)† Published

675

720

725

Air and O2 1.5:0.5 O2 and N2O 0.8:1.2 116 patients having elective craniotomies for supratentorial tumours Singh and colleagues76

715

70% N2O and 30% O2

710

2050 patients having a surgical procedure expected to last .2 h and expected to stay in hospital .72 h

705

Chan and colleagues98 Myles and colleagues5 Myles and colleagues56 Leslie and colleagues148 Leslie and colleagues87

700

Myles and colleagues83

695

80% O2 and 20% N2

690

meta-analysis was patients having non-cardiac surgery expected to last longer than 2 h and patients with colon cancer having colon resection expected to last longer than 2 h.35 87 For this population, we can conclude, with a type II risk of 10%, that exposure to nitrous oxide does not alter longterm mortality by 20% (number needed to harm¼5). Clearly, for mortality we are clinically interested in an effect size much smaller than this for an intervention as commonly used as nitrous oxide. We therefore conclude that, for longterm mortality, for a clinically relevant effect size, we have insufficient evidence. The major strength of this review was its thorough search. We designed our approach to maximize sensitivity. With the amount of published trial work including nitrous oxide as an intervention, this task was large, but we feel confident that our systematic and transparent approach gave us a good chance to collect a complete summary of the evidence derived from current randomized controlled trials. This systematic review did not aim to investigate all the outcomes relevant to the use of nitrous oxide. Rather, it aimed to systematically collect and analyse what we know about nitrous oxide and mortality and major cardiovascular complications, before the completion of ENIGMA II. Modelling the significance of the findings can be challenging. TSA allows the definition of a hypothesis, with regard to control event proportions and effect size estimate, in the same way as is done for a randomized controlled trial. Moreover, heterogeneity is included. As you change the results of these parameters, that is, as you change the assumptions in the model, you may change your conclusions. So the answers in the meta-analyses depend crucially on the questions being asked, and the assumptions being made. In this systematic review, we estimated control event proportions by pooling all the no-nitrous oxide groups included in the relevant meta-analysis, the effect size was estimated from the included trials with a low risk of bias and for an a priori anticipated relative risk reduction of 10 and 20%, and we used the D 2 present in the included studies to adjust for heterogeneity. Consequently, our conclusions answer the questions that are defined by these parameters and depend on these assumptions. As with any attempt to synchronize evidence and to model the significance of the findings, there are limitations. Of particular note, in this review, the potential for clinical heterogeneity was large because the inclusion criteria were broad. Estimates of heterogeneity in meta-analysis can be unreliable when numbers are small.164 Consequently, we may have under-estimated the heterogeneity, and this fact further reduces the confidence we can have in the existing evidence. Nitrous oxide has been investigated extensively with regard to many different primary outcomes. From our search, we did not find any randomized clinical trials that handled our outcomes as primary outcomes. In most cases, we were extracting data that were reported incidentally. The disadvantage of this is that the study design would not have been intended for our outcomes. Moreover, because our outcomes were incidental for many of these trials, we retrieved some of our data from email correspondence with authors. There were 117

Page 7 of 17

730

735

740

745

750

755

760

765

770

775

780

785

790

795

800

805

810

815

820

825

830

835

840

Trial

Randomization

Concealment

Blinding*

Attrition

Other†

Computer generated

Sealed envelopes

Anaesthetists not blinded, surgeons blinded, outcome assessors blinded

10 (2%)—short-term mortality Well explained Data sheet lost for 4 (unknown groups) 2 lost to follow-up from the nitrous group (surgical complications) 4 lost to follow-up from the no nitrous group (1 withdrawn by a physician and 3 found not to meet the inclusion criteria) 11 (5%) 10 from the initial study (see Fleischmann and colleagues 162) 1 extra lost to long-term follow-up (from the nitrous group)

–

BJA

Page 8 of 17

Table 3 Assessments of risk of bias for included trials. *Blinding: participants and personnel, outcome assessment. †Other: evidence of selective outcome reporting, other potential threats to validity. ‡Information sourced from email correspondence with investigators

Low risk of bias Fleischmann and colleagues162 Mortality—short

Mortality—long

Post-hoc analysis

Leung and colleagues 37 Mortality—short

Computerized random number list

Randomization sequence concealed

Outcome assessment blinded

18 (8%) for the primary outcome Reasons not described Denominator/percentage not quoted for the mortality data Assume it was the same as for the primary outcome

–

Myles and colleagues 83 Mortality—short Stroke MI Mortality—long

Computer-generated code

Automated telephone service

Anaesthetists not blinded Outcome assessment blinded Surgeon and patient not informed of treatment group

38 (2%) 23 (2%) from the group who got nitrous 15 (1%) from the group who did not get nitrous All explained—deferred surgery or development of exclusion criteria 380 (19%) 38 (2%) during original trial (see Myles and colleagues 83) 340 (17%) not contacted or uncontactable (49.1% from nitrous group, 50.9% from the no nitrous group) 2 (0.1%) declined to participate Equal numbers of nitrous/no nitrous in the attrition population

–

Post-hoc analysis

Increased risk of bias Not described

Not described

Surgeons blinded to intervention

Nil

–

Chowdhury and colleagues 36 Mortality—short

Computer generated‡

Not described‡

Those performing the analyses were blinded‡

5 (5%) All from the group who received no nitrous Well explained and unlikely to be linked with our outcomes‡

–

Eger and colleagues 61 Mortality—short MI

Not fully described

Not described

Outcome assessment blinded

Nil 40 (57% of those randomized in relevant group) Reason for attrition not described (postoperative MI was not a primary outcome/concern in the paper)

–

Imberger et al.

Antonini and colleagues 58 Mortality—short

845

850

855

860

865

870

875

880

885

890

895

900

905

910

915

920

925

930

935

940

945

950

Card withdrawal technique

Not described

Surgeon blinded

Nil

–

Hohner and colleagues 89 Mortality—short

Not fully described

Not described

Subjective outcome assessment blinded

2 (4%) Well described

–

Iacopino and colleagues 112 Mortality—short

Not described

Not described

‘Study conducted in a blinded fashion’

Nil

–

Jensen and colleagues131 Mortality—short

Set of numbered envelopes

‘Randomized blindly’

Surgical team members blinded

Nil

Multiple publications, each describing different outcomes

Knu¨ttgen and colleagues30 Mortality—short

Not fully described

Not described

Not described

Nil

–

Krogh and colleagues97 Mortality—short

Not described

Not described

Surgeons blinded

Nil

–

Singh and colleagues76 Mortality—short Stroke MI PE

Computer-generated

Opaque envelopes‡

Surgeons and outcome assessment blinded

29 (25%) 15 (27%) from the group who got nitrous 14 (23%) from the group who did not get nitrous All explained—remained intubated postoperative and therefore ineligible for measurement of the primary outcome of the study Quantity/nature of attrition a source for increased risk of bias for our outcomes

–

Nitrous oxide and cardiovascular outcomes

El-Galley and colleagues 95 Mortality—short

BJA

Page 9 of 17 955

960

965

970

975

980

985

990

995

1000

1005

BJA

Imberger et al.

1065 1010

1015

1020

1025

1030

Nitrous oxide No nitrous oxide Risk ratio Total Weight M-H, random, 95%Cl Study or subgroup Events Total Events Low risk of bias 0.14 [0.01, 2.70] 202 9.8% Fleischmann 2005 0 206 3 Leung 2006 114 3.00 [0.12, 72.88] 8.4% 0 114 1 Myles 2007 997 50.2% 2.95 [0.80, 10.85] 9 1015 3 Subtotal (95% Cl) 1335 1313 68.4% 1.38 [0.22, 8.71] Total events 10 6 Heterogeneity: t 2=1.25; c 2=3.64, df=2 (P=0.16); l 2=45% Test for overall effect: Z=0.34 (P=0.73) Increased risk of bias Antonini 1994 0 20 0 20 Chowdhury 2011 0 45 0 40 Eger 1990 1 133 0 137 El Gallery 2007 0 12 0 16 Hohner 1994 1 1 22 26 20 20 Iacopino 2003 0 0 Jensen 1992 0 20 20 0 Knuttgen 1989 1 1 21 21 Krogh 1994 67 72 0 0 Singh 2011 41 46 0 0 Subtotal (95% Cl) 401 418 3 2 Total events Heterogeneity: t 2=0.00; c 2=0.31, df=2 (P=0.86); l 2=0% Test for overall effect: Z=0.43 (P=0.67)

8.4% 11.0%

11.7%

31.6%

Risk ratio M-H, random, 95%Cl

1070

1075

Not estimable Not estimable 3.09 [0.13, 75.17] Not estimable 1.18 [0.08, 17.82] Not estimable Not estimable 1.00 [0.07, 14.95] Not estimable Not estimable 1.43 [0.28, 7.41]

1080

1085

1736 1731 100.0% Total (95% Cl) 13 8 Total events Heterogeneity: t 2=0.00; c 2=3.97, df=5 (P=0.55); l 2=0% Test for overall effect: Z=1.18 (P=0.24) Test for subgroup differences: c 2=0.00, df=1 (P=0.97); l 2=0%

1.75 [0.69, 4.40]

0.01

0.1 Favours nitrous oxide

1

10 100 Favours no nitrous oxide

1090

1035

Fig 2 Forest plot for short-term mortality.

1040

1045

Table 4 Point estimates, calculated heterogeneity and estimated RIS for short-term mortality (low risk of bias) as the meta-analysis model is varied. CIs shown are calculated using conventional meta-analysis. Analysis model Effect measure

Continuity adjustment

Point estimate (95% CI)

I 2 (%)

RIS

Random effects Relative risk

Constant– 0.001 Constant– 1.0 Empirical –0.001 Empirical –1.0

2.94 (0.80 – 10.84) 1.4 (0.23 – 8.45) 2.94 (0.24 – 9.58) 1.51 (0.24 – 9.58)

0 43 0 44

4357 162 802 4357 115 011

Fixed effect Relative risk

Constant– 0.001 Constant– 1.0 Empirical –0.001 Empirical –1.0

1.64 (0.60 – 4.50) 1.55 (060 – 3.98) 1.64 (0.60 – 4.50) 1.73 (0.66 – 4.52)

0 43 0 44

26 911 94 506 26 911 57 514

1.62 (1.61 – 4.34)

67

34 070

1050

Peto fixed effect Peto odds ratio

1095

1100

1105

1110 1055

1060

publications that were eligible based on trial type, population and intervention. Of these, we were unable to contact 63. It is possible that this approach to data collection introduces bias; perhaps those whom we could not contact represent a different population of studies. Our systematic review demonstrated the high number of outcomes—both surrogate and clinical—that have been considered in randomized controlled trials investigating nitrous oxide. This list does not include outcomes examined in

Page 10 of 17

observational studies, such as environmental issues and staff exposure, which would extend its size further. The length of this list reflects how many different outcomes have been considered relevant to nitrous oxide. Assessing the merits of using this drug involves evaluating the strength of evidence for how nitrous oxide affects each relevant outcome. Advantageous outcomes can then be weighed up against the disadvantageous and a judgement can be made about the merits of using the drug.

1115

1120

BJA

Nitrous oxide and cardiovascular outcomes

Study or subgroup 1125

Fleischmann 2005 Myles 2007

Nitrous oxide No nitrous oxide Risk ratio Events Total Events Total Weight M-H, fixed, 95%Cl 36 185

95 847

40 195

Total (95% Cl) 942 Total events 221 235 Heterogeneity: c 2=0.26, df=1 (P=0.61); l 2=0% Test for overall effect; Z=0.78 (P=0.44)

108 824

15.9% 84.1%

1.02 [0.72, 1.46] 0.92 [0.77, 1.10]

932 100.0%

0.94 [0.80, 1.10]

Risk ratio M-H, fixed, 95%Cl 1180

0.01

1130

0.1 Favours nitrous oxide

1

10 100 Favours no nitrous oxide

1185

Fig 3 Forest plot for long-term mortality. 1190 1135

Information size is a two-sided graph Cumulative Z-score

Information size=2932

8 1140

1195

7

1145

Favours nitrous

6 5 1200

4 3 2 1

1205

Z-curve

1150 1873

Number of patients (linear scaled)

–1

1155

Favours no nitrous

–2 –3

1210

–4 –5 –6 1215

–7 1160

–8

Fig 4 TSA for long-term mortality, testing for a relative risk reduction of 20%. 1220 1165

1170

Nitrous oxide No nitrous oxide Risk ratio Study or subgroup Events Total Events Total M-H, fixed, 95%Cl Myles 2007 1 1015 1 997 0.98 [0.06, 15.68] Singh 2011 3 41 2 46 1.68 [0.30, 9.58] 0.01

1175

1225

Risk ratio M-H, fixed, 95%Cl

0.1 Favours nitrous oxide

1

10 100 Favours no nitrous oxide

Fig 5 Forest plot for stroke.

Page 11 of 17

1230

BJA 1235

Imberger et al.

Nitrous oxide No nitrous oxide Risk ratio Study or subgroup Events Total Events Total M-H, fixed, 95%Cl Eger 1990 1 14 3 16 0.38 [0.04, 3.26] 1 22 5 26 0.24 [0.03, 1.87] Hohner 1994 13 1015 7 997 1.82 [0.73, 4.55] Myles 2007 1 41 0 46 3.36 [0.14, 80.20] Singh 2011 0.01

1240

Risk ratio M-H, fixed, 95%Cl

1290

1295

0.1 Favours nitrous oxide

1

10 100 Favours no nitrous oxide

Fig 6 Forest plot for MI.

1300

1245

1250

1255

1260

1265

1270

1275

1280

1285

While many outcomes can alter in their relative importance, depending on circumstances and individual judgement, mortality and cardiovascular complications are always important. In most situations, these outcomes would be considered more important than all others. If nitrous oxide does not affect mortality or cardiovascular complications, then the merits of its use lie with the balance of other outcomes. If nitrous oxide does affect these important outcomes, then this information should profoundly affect whether this drug is used. Our systematic review has demonstrated that we currently do not have robust evidence for how nitrous oxide affects mortality and cardiovascular complications. We can say neither that it does not have an effect nor that it does. This lack of evidence leaves a striking hole. We look forward with interest, therefore, to the results of ENIGMA II and to the continued debate on the role of nitrous oxide in modern anaesthetic practice.

Authors’ contributions Conceiving the review: G.I., K.T., A.O., J.W., P.M., and A.M.M. Coordinating the review: G.I. Undertaking the search: G.I. Screening the abstracts from the search: G.I. and A.O. Retrieving the full papers for review: G.I. Reviewing full papers for inclusions: G.I. and AO. Contacting the authors of review to confirm outcome measurement: G.I. Appraising the quality of included papers: G.I. and A.O. Abstracting data from included papers: G.I. and A.O. Data management for the review: G.I. and J.W. Statistical analysis: G.I. and J.W. Interpretation of data: G.I., J.W., P.M., and A.M.M. Writing the review: G.I. (principal), K.T., A.O., J.W., P.M., and A.M.M.

Acknowledgements We thank Eva Grogro and Federico Riezzo for their assistance with translation. There were many authors of nitrous oxide trials who replied to our emails of enquiry. The list is too long to publish here, but we thank each of them greatly for their time and their interest in the project.

Declaration of interest None declared.

Page 12 of 17

Funding G.I. received a PhD study grant from Rigshospitalet, Copenhagen, Denmark, while conducting this project. No other author received funding for this project.

1305

References 1

Baum VC, Willschke H, Marciniak B. Is nitrous oxide necessary in the future? Paediatr Anaesth 2012; 22: 981– 7 2 Myles PS, Leslie K, Silbert B, Paech MJ, Peyton P. A review of the risks and benefits of nitrous oxide in current anaesthetic practice. Anaesth Intensive Care 2004; 32: 165–72 3 Weimann J. Toxicity of nitrous oxide. Best Pract Res Clin Anaesthesiol 2003; 17: 47–61 4 Badner NH, Drader K, Freeman D, Spence JD. The use of intraoperative nitrous oxide leads to postoperative increases in plasma homocysteine. Anesth Analg 1998; 87: 711–3 5 Myles PS, Chan MT, Kaye DM, et al. Effect of nitrous oxide anesthesia on plasma homocysteine and endothelial function. Anesthesiology 2008; 109: 657– 63 6 Rao LK, Francis AM, Wilcox U, Miller JP, Nagele P. Pre-operative vitamin B infusion and prevention of nitrous oxide-induced homocysteine increase. Anaesthesia 2010; 65: 710–5 7 The Homocysteine Studies Collaboration. Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA 2002; 288: 2015–22 8 Myles PS, Leslie K, Peyton P, et al. Nitrous oxide and perioperative cardiac morbidity (ENIGMA-II) trial: rationale and design. Am Heart J 2009; 157: 488–94 9 Todd MM, Hindman BJ, Clarke WR, Torner JC. Mild intraoperative hypothermia during surgery for intracranial aneurysm. N Engl J Med 2005; 135: 352–45 10 Pasternak JJ, McGregor DG, Lanier WL, et al. Effect of nitrous oxide use on long-term neurologic and neuropsychological outcome in patients who received temporary proximal artery occlusion during cerebral aneurysm clipping surgery. Anesthesiology 2009; 110: 563– 73 11 GALA Trial Collaborative Group. General anaesthesia versus local anaesthesia for carotid surgery (GALA): a multicentre, randomised controlled trial. Lancet 2008; 372: 2132–42 12 Sanders RD, Graham C, Lewis SC, Bodenham A, Gough MJ, Warlow C. Nitrous oxide exposure does not seem to be associated with increased mortality, stroke, and myocardial infarction: a nonrandomized subgroup analysis of the general anaesthesia compared with local anaesthesia for carotid surgery (GALA) trial. Br J Anaesth 2012; 109: 361– 7

1310

1315

1320

1325

1330

1335

1340

Nitrous oxide and cardiovascular outcomes

1345

1350

1355

1360

1365

1370

1375

1380

1385

1390

1395

1400

13 Turan A, Mascha EJ, You J, et al. The association between nitrous oxide and postoperative mortality and morbidity after noncardiac surgery. Anesth Analg 2012; 116: 1026– 33 14 Leslie K, Myles P, Devereaux PJ, et al. Nitrous oxide and serious morbidity and mortality in the POISE trial. Anesth Analg 2013; 116: 1034–40 15 McPherson K. Statistics: The problem of examining accumulating data more than once. N Engl J Med 1974; 290: 501–2 16 Freidlin B, Korn EL, George SL. Data monitoring committees and interim monitoring guidelines. Control Clin Trials 1999; 20: 395–407 17 Brok J, Thorlund K, Gluud C, Wetterslev J. Trial sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses. J Clin Epidemiol 2008; 61: 763–9 18 Brok J, Thorlund K, Wetterslev J, Gluud C. Apparently conclusive meta-analyses may be inconclusive—trial sequential analysis adjustment of random error risk due to repetitive testing of accumulating data in apparently conclusive neonatal meta-analyses. Int J Epidemiol 2009; 38: 287– 98 19 Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol 2008; 61: 64– 75 20 Thorlund K, Devereaux PJ, Wetterslev J, et al. Can trial sequential monitoring boundaries reduce spurious inferences from meta-analyses? Int J Epidemiol 2009; 38: 276–86 21 Thorlund K, Engstrøm J, Wetterslev J, Brok J, Imberger G, Gluud C. User Manual for Trial Sequential Analysis (TSA). Copenhagen, Denmark: Copenhagen Trial Unit, Centre for Clinical Intervention Research, 2011; 1 –115. Available from www.ctu.dk/tsa (accessed 30 January 2013) 22 Higgins JPT, Altman DG, Sterne JAC.(editors) Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S, eds, Cochrane Handbook for Systematic Reviews of Interventions. Chichester: John Wiley & Sons, 2008; 187– 235 23 Thorlund K, Imberger G, Walsh M, et al. The number of patients and events required to limit the risk of overestimation of intervention effects in meta-analysis – a simulation study. PLoS One 2011; 6: e25491 24 Wetterslev J, Thorlund K, Brok J, Gluud C. Estimating required information size by quantifying diversity in random-effects model meta-analyses. BMC Med Res Methodol 2009; 9: 86 25 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88 26 Deeks JJ, Higgins JPT, Altman DG. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Chichester: John Wiley & Sons, 2008; 243– 93 27 Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004; 23: 1351– 75 28 Yusuf S, Peto R, Lewis J, Collins R, Sleight P. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovasc Dis 1985; 27: 335–71 29 Hernandez-Palazon J, Martinez-Lage JF, De la Rosa-Carrillo VN, Tortosa JA, Lopez F, Poza M. Anesthetic technique and development of pneumocephalus after posterior fossa surgery in the sitting position. Neurocirugia 2003; 14: 216–21 30 Knuttgen D, Stolzle U, Koning W, Muller MR, Doehn M. Air embolism in the sitting position. Oxygen/nitrogen versus oxygen/laughing gas. Anaesthesist 1989; 38: 490–7 31 Losasso TJ, Muzzi DA, Dietz NM, Cucchiara RF. Fifty percent nitrous oxide does not increase the risk of venous air embolism in neurosurgical patients operated upon in the sitting position. Anesthesiology 1992; 77: 21 –30

BJA 32 Al Zahrani T, Ibraheim O, Turkistani A, Mazen K. Bispectral index profile during general anaesthesia using nitrous oxide for lower segment caesarean delivery. Internet J Anesthesiol 2006; 10: 7p 33 Coenen P, Lauwers M, Vanlersberhge C, Camu F. Contribution of nitrous oxide to anesthetic depth of sevoflurane and desflurane anesthesia. Br J Anaesth 1999; 82(Suppl. 1): 17 34 Nakayama M, Iwasaki S, Ichinose H, et al. Effect of nitrous oxide on the bispectral index during sevoflurane anesthesia. Masui 2002; 51: 973–6 35 Fleischmann E, Marschalek C, Schlemitz K, et al. Nitrous oxide may not increase the risk of cancer recurrence after colorectal surgery: a follow-up of a randomized controlled trial. BMC Anesthesiol 2009; 9: 1 36 Chowdhury T, Prabhakar H, Kumar BP, Dash HH. To assess the effects of two anesthetic techniques (with and without nitrous oxide) on cognitive function in patients undergoing transsphenoidal pituitary surgery. J Neurosurg Anesthesiol 2011; 23: 436 (Conference publication) 37 Leung JM, Sands LP, Vaurio LE, Wang Y. Nitrous oxide does not change the incidence of postoperative delirium or cognitive decline in elderly surgical patients. Br J Anaesth 2006; 96: 754– 60 38 Adamus M, Koutna J, Hrabalek L, Konupcikova K, Fischerova L, Francakova J. Comparison of the cost of balanced anaesthesia maintained with or without nitrous oxide at a university hospital. Anesteziol Intenzivni Med 2011; 22: 90– 4 39 Heath KJ, Sadler P, Winn JH, McFadzean WA. Nitrous oxide reduces the cost of intravenous anaesthesia. Eur J Anaesthesiol 1996; 13: 369 – 72 40 Ekbom K, Assareh H, Anderson RE, Jakobsson JG. The effects of fresh gas flow on the amount of sevoflurane vaporized during 1 minimum alveolar concentration anaesthesia for day surgery: a clinical study. Acta Anaesthesiol Scand 2007; 51: 290–3 41 Garrido P, Caldera MV, Deudero E, Villagran MJ, Martinez B, Criado A. Effects of nitrous oxide as second gas during desflurane anaesthesia. Br J Anaesth 1998; 80(Suppl. 1): 136 42 Gin T, Yau G, Chan K, Gregory MA, Oh TE. Disposition of propofol infusions for caesarean section. Can J Anaesth 1991; 38: 31 –6 43 Harbut P, Wysoczanska-Harbut J, Boroch M, Kubler A. Propofol consumption and recovery from target-controlled anaesthesia: influence of the nitrous oxide administration. Br J Anaesth 1999; 82: A478 44 Hendrickx JF, Coddens J, Callebaut F, et al. Effect of N2O on sevoflurane vaporizer settings during minimal- and low-flow anesthesia. Anesthesiology 2002; 97: 400– 4 45 Hendrickx JFA, Carette R, Lemmens HJM, De Wolf AM. Large volume N2O uptake alone does not explain the second gas effect of N2O on sevoflurane during constant inspired ventilation. Br J Anaesth 2006; 96: 391–5 46 Illman H, Antila H, Olkkola KT. Quantitation of the effect of nitrous oxide on rocuronium infusion requirements using closed-loop feedback control. Anesthesiology 2008; 108: 388–91 47 Illman HL, Antila HMJ, Olkkola KT. Effect of nitrous oxide on cisatracurium infusion demands: a randomized controlled trial. BMC Anesthesiol 2010; 10: 14 48 Jakobsson J, Heidvall M, Davidson S. The sevoflurane-sparing effect of nitrous oxide: A clinical study. Acta Anaesthesiol Scand 1999; 43: 411–4 49 Kopman AF, Chin WA, Moe J, Malik R. The effect of nitrous oxide on the dose-response relationship of rocuronium. Anesth Analg 2005; 100: 1343–7 50 Koukita Y, Ichinohe T, Hosaka G, et al. The effect of nitrous oxide on the vecuronium-induced neuromuscular blockade during propofol anesthesia. J Jpn Dent Soc of Anesthesiol 2004; 32: 193–7

Page 13 of 17

1405

1410

1415

1420

1425

1430

1435

1440

1445

1450

1455

BJA

1460

1465

1470

1475

1480

1485

1490

1495

1500

1505

1510

51 Lindekaer AL, Skielboe M, Guldager H, Jensen EW. The influence of nitrous oxide on propofol dosage and recovery after total intravenous anaesthesia for day-case surgery. Anaesthesia 1995; 50: 397– 9 52 Liu N, Boichut N, Herail T, et al. Propofol and remifentanil sparing effect of nitrous oxide using closed-loop anesthesia controller. Anesth Analg 2011; 113: 8 53 Peyton PJ, Chao I, Weinberg L, Robinson GJB, Thompson BR. Nitrous oxide diffusion and the second gas effect on emergence from anesthesia. Anesthesiology 2011; 114: 596– 602 54 Rothhammer A. Nitrous oxide in combination anesthesia. Quantitative aspects of the nitrous oxide effects. Anaesthesist 1987; 36: 333– 9 55 Koblin DD, Tomerson BW, Waldman FM, Lampe GH, Wauk LZ, Eger EI. Effect of nitrous oxide on folate and vitamin B12 metabolism in patients. Anesth Analg 1990; 71: 610–7 56 Myles PS, Chan MT, Leslie K, Peyton P, Paech M, Forbes A. Effect of nitrous oxide on plasma homocysteine and folate in patients undergoing major surgery. Br J Anaesth 2008; 100: 780 – 6 57 Abballe C, Camaioni D, Mascaro A, Boccardi M, Evangelista M. Anesthesia for laparoscopic cholecystectomy: the use of nitrous oxide in the anesthetic mixture. G Chir 1993; 14: 493–5 58 Antonini M, D’Errico RR, Della RG, Oliva P, Coccia C, Pastore E. Anesthesia in laparoscopic cholecystectomy: evaluation of the use of nitrous oxide. Acta Anaesthesiol Ital 1994; 45: 243–9 59 Arellano RJ, Pole ML, Rafuse SE, et al. Omission of nitrous oxide from a propofol-based anesthetic does not affect the recovery of women undergoing outpatient gynecologic surgery. Anesthesiology 2000; 93: 332–9 60 D’Angelo C, Vacca F, Perin S, Cervi R, Pincelli D. Total intravenous anesthesia with propofol vs. propofol/N2O anesthesia. Minerva Anestesiol 1990; 56: 169– 74 61 Eger EI, Lampe GH, Wauk LZ, Whitendale P, Cahalan MK, Donegan JH. Clinical pharmacology of nitrous oxide: an argument for its continued use. Anesth Analg 1990; 71: 575–85 62 Einarsson S, Bengtsson A, Stenqvist O, Bengtson JP. Emergence from isoflurane/N2O or isoflurane anaesthesia. Acta Anaesthesiol Scand 1997; 41: 1292–9 63 Fagiano G, Berra L, Catalano C, Salamino A. Balanced anesthesia without N2O. Its use in thoracic surgery. Minerva Anestesiol 1990; 56: 681– 4 64 Gilani SM, Sofi K. Is nitrous oxide necessary for general anaesthesia? J Ayub Med Coll Abbottabad 2008; 20: 149–52 65 Girardi G, Rossi R, Cellai MP, Pieraccioli E, Novelli GP. Anesthesia with isoflurane in air and with isoflurane and nitrous oxide. Minerva Anestesiol 1994; 60: 321–8 66 Jensen AG, Prevedoros H, Kullman E, Anderberg B, Lennmarken C. Peroperative nitrous-oxide does Not influence recovery after laparoscopic cholecystectomy. Acta Anaesthesiol Scand 1993; 37: 683–6 67 Kalman SH, Jensen AG, Ekberg K, Eintrei C. Early and late recovery after major abdominal surgery. Comparison between propofol anaesthesia with and without nitrous oxide and isoflurane anaesthesia. Acta Anaesthesiol Scand 1993; 37: 730–6 68 Lampe GH, Wauk LZ, Donegan JH, et al. Effect on outcome of prolonged exposure of patients to nitrous oxide. Anesth Analg 1990; 71: 586–90 69 Lebenbom-Mansour MH, Pandit SK, Kothary SP, Randel GI, Levy L. Desflurane versus propofol anesthesia: A comparative analysis in outpatients. Anesth Analg 1993; 76: 936– 41 70 Lind L, Johansson S, Wickstrom E. Nitrous oxide is not necessary in propofol anesthesia. Lakartidningen 1993; 90: 485– 6

Page 14 of 17

Imberger et al.

71 Mandal P, Parray FA. Air-oxygen sevoflurane anaesthesia for laparoscopic surgery. J Anaesthesiol Clin Pharmacol 2003; 19: 69 – 72 72 Nishijima S, Konno M, Sakurada H. The effects of propofol anesthesia with or without the use of nitrous oxide on the intraoperative involuntary movement, the postoperative awareness and vomiting. Masui 1999; 48: 1216–9 73 Ozdogan B, Coskunfirat N, Arici G, Ertok E. General anesthesia for thyroid surgery: comparison of remifentanil infusion with nitrous oxide. Br J Anaesth 2012; 108: ii382– 83 (Conference publication) 74 Rapp SE, Conahan TJ, Pavlin DJ, et al. Comparison of desflurane with propofol in outpatients undergoing peripheral orthopedic surgery. Anesth Analg 1992; 75: 572– 9 75 Rocca G, Montecchi C, Baisi F, Monaco S, Romboli D, Gasparetto A. N2O-free sevoflurane anesthesia. Clinical evaluation. Minerva Anestesiol 2000; 66: 611– 9 76 Singh GP, Prabhakar H, Bithal PK, Dash HH. A comparative evaluation of nitrous oxide-isoflurane vs isoflurane anesthesia in patients undergoing craniotomy for supratentorial tumors: a preliminary study. Neurol India 2011; 59: 18 –24 77 Tang J, Chen L, White PF, et al. Use of propofol for office-based anesthesia: effect of nitrous oxide on recovery profile. J Clin Anesth 1999; 11: 226– 30 78 Taylor E, Feinstein R, White PF, Soper N. Anesthesia for laparoscopic cholecystectomy: is nitrous oxide contraindicated? Anesthesiology 1992; 76: 541–3 79 Van HJ, Smith I, White PF. Use of desflurane for outpatient anesthesia: a comparison with propofol and nitrous oxide. Anesthesiology 1991; 75: 197–203 80 Van HJ, Tempelhoff R, White PF, Jellish WS. EEG-assisted titration of propofol infusion during neuroanesthesia: effect of nitrous oxide. J Neurosurg Anesthesiol 1992; 4: 11– 20 81 Wammen S, Guldager H, Rasmussen AM, Hojsted J. Total intravenous anesthesia with propofol-fentanyl. Is it necessary to supply nitrous oxide?. Ugeskr Laeger 1990; 152: 2415–7 82 Rosen MA, Roizen MF, Eger EI, et al. The effect of nitrous oxide on in vitro fertilization success rate. Anesthesiology 1987; 67: 42 – 4 83 Myles PS, Leslie K, Chan MTV, et al. Avoidance of nitrous oxide for patients undergoing major surgery: a randomized controlled trial. Anesthesiology 2007; 107: 221–31 84 Asai T, Shingu K. Time-related cuff pressures of the laryngeal tube with and without the use of nitrous oxide. Anesth Analg 2004; 98: 1803–6 85 Brimacombe J, Keller C. Does efficacy of seal and fibreoptic view change during anaesthesia with the laryngeal mask airway: a comparison of oxygen and oxygen-nitrous oxide gas mixtures. Acta Anaesthesiol Scand 2000; 44: 1258– 60 86 Tamura H, Matayoshi Y, Nakamura K, et al. Propofol-air-oxygen anesthesia reduces the incidence of sore throat after laryngeal mask anesthesia. Masui 1998; 47: 1217– 20 87 Leslie K, Myles PS, Chan MT, et al. Nitrous oxide and long-term morbidity and mortality in the ENIGMA trial. Anesth Analg 2011; 112: 387– 93 88 Kozmary SV, Lampe GH, Benefiel D, et al. No finding of increased myocardial ischemia during or after carotid endarterectomy under anesthesia with nitrous oxide. Anesth Analg 1990; 71: 591–6 89 Hohner P, Backman C, Diamond G, et al. Anaesthesia for abdominal aortic surgery in patients with coronary artery disease, part II: effects of nitrous oxide on systemic and coronary haemodynamics, regional ventricular function and incidence of myocardial ischaemia. Acta Anaesthesiol Scand 1994; 38: 793–804

1515

1520

1525

1530

1535

1540

1545

1550

1555

1560

1565

Nitrous oxide and cardiovascular outcomes

1570

1575

1580

1585

1590

1595

1600

1605

1610

1615

1620

90 Badner NH, Beattie WS, Freeman D, Spence JD. Nitrous oxide-induced increased homocysteine concentrations are associated with increased postoperative myocardial ischemia in patients undergoing carotid endarterectomy. Anesth Analg 2000; 91: 1073– 9 91 Launo C, Palermo S, Riello R, Cammardella MP, Ughe R. Clinical and neuropsychologic evaluation of different anesthesia techniques (propofol vs isoflurane) in general surgery. Minerva Anestesiol 1994; 60: 95– 108 92 Akca O, Lenhardt R, Fleischmann E, et al. Nitrous oxide increases the incidence of bowel distension in patients undergoing elective colon resection. Acta Anaesthesiol Scand 2004; 48: 894 – 8 93 Boulanger A, Hardy JF. Intestinal distention during elective abdominal surgery: should nitrous oxide be banished? Can J Anaesth 1987; 34: 346– 50 94 Brodsky JB, Lemmens HJM, Collins JS, Morton JM, Curet MJ, Brock-Utne JG. Nitrous oxide and laparoscopic bariatric surgery. Obes Surg 2005; 15: 494–6 95 El-Galley R, Hammontree L, Urban D, Pierce A, Sakawi Y. Anesthesia for laparoscopic donor nephrectomy: is nitrous oxide contraindicated? J Urol 2007; 178: 225– 7 96 Karlsten R, Kristensen JD. Nitrous oxide does not influence the surgeon’s rating of operating conditions in lower abdominal surgery. Eur J Anaesthesiol 1993; 10: 215– 7 97 Krogh B, Jensen PJ, Henneberg SW, Hole P, Kronborg O. Nitrous oxide does not influence operating conditions or postoperative course in colonic surgery. Br J Anaesth 1994; 72: 55 –7 98 Chan MT, Wan AC, Gin T, Leslie K, Myles PS. Chronic postsurgical pain after nitrous oxide anesthesia. Pain 2011; 152: 2514–20 99 Echevarria G, Elgueta F, Fierro C, et al. Nitrous oxide (N(2)O) reduces postoperative opioid-induced hyperalgesia after remifentanilpropofol anaesthesia in humans. Br J Anaesth 2011; 107: 959– 65 100 Fievez L, Decottenier V, Salengros J, Sosnowski M. Effect of nitrous oxide and nefopam on hyperalgesia and chronic pain after breast cancer surgery. Anesth Analg 2011; 112(Suppl. 1): S-321 (Conference publication) 101 Adams HA, Kling D, Boldt J, Dapper F, Hempelmann G. Effects of nitrous oxide on endocrine stress response and haemodynamic parameters during coronary artery surgery. Thorac Cardiovasc Surg 1990; 38: 73–8 102 Hamamcioglu G, Askar FZ, Certug A, Kultursay H, Kayaalti B. Effect of nitrous oxide on cardiac dysrhythmias during anaesthesia. Turk Anesteziyoloji ve Reanimasyon 1994; 22: 89– 92 103 Ishiguro Y, Goto T, Nakata Y, Terui K, Niimi Y, Morita S. Effect of xenon on autonomic cardiovascular control—comparison with isoflurane and nitrous oxide. J Clin Anesth 2000; 12: 196–201 104 Lampe GH, Donegan JH, Rupp SM, et al. Nitrous oxide and epinephrine-induced arrhythmias. Anesth Analg 1990; 71: 602– 5 105 McKinney MS, Fee JPH. Cardiovascular effects of 50% nitrous oxide in older adult patients anaesthetized with isoflurane or halothane. Br J Anaesth 1998; 80: 169– 73 106 Shiga T, Wajima Z, Inoue T, Ogawa R. Nitrous oxide produces minimal hemodynamic changes in patients receiving a propofolbased anesthetic: an esophageal Doppler ultrasound study. Can J Anaesth 2003; 50: 649–52 107 Shirgoska B, Trajkovska T, Soljakova M, et al. Level of nitric oxide in hypertensive patients scheduled on general anaesthesia. Prilozi 2005; 26: 13–24 108 Aceto P, Valente A, Gorgoglione M, Adducci E, De CG. Relationship between awareness and middle latency auditory evoked responses during surgical anaesthesia. Br J Anaesth 2003; 90: 630–5

BJA 109 Aceto P, Valente A, Adducci E, Gorgoglione M, De CG. Implicit memory, dream and auditory evoked responses during anaesthesia. Acta Med Rom 2002; 40: 21– 9 110 Handa T, Koike S, Kohkita Y, et al. Frequency and content of dreams during propofol anesthesia in patients undergoing mandibular sagittal split ramus osteotomy. J Jpn Dent Soc Anesthesiol 2010; 38: 29– 34 111 Hans P, Dewandre PY, Brichant JF, Bonhomme V. Effects of nitrous oxide on spectral entropy of the EEG during surgery under balanced anaesthesia with sufentanil and sevoflurane. Acta Anaesthesiol Belg 2005; 56: 37 –43 112 Iacopino DG, Conti A, Battaglia C, et al. Transcranial Doppler ultrasound study of the effects of nitrous oxide on cerebral autoregulation during neurosurgical anesthesia: A randomized controlled trial. J Neurosurg 2003; 99: 58– 64 113 Kunisawa T, Nagata O, Nomura M, Iwasaki H, Ozaki M. A comparison of the absolute amplitude of motor evoked potentials among groups of patients with various concentrations of nitrous oxide. J Anesth 2004; 18: 181–4 114 Langeron O, Vivien B, Paqueron X, et al. Effects of propofol, propofol-nitrous oxide and midazolam on cortical somatosensory evoked potentials during sufentanil anaesthesia for major spinal surgery. Br J Anaesth 1999; 82: 340–5 115 Strebel S, Kaufmann M, Baggi M, Zenklusen U. Cerebrovascular carbon dioxide reactivity during exposure to equipotent isoflurane and isoflurane in nitrous oxide anaesthesia. Br J Anaesth 1993; 71: 272– 6 116 Strebel S, Kaufmann M, Anselmi L, Schaefer HG. Nitrous oxide is a potent cerebrovasodilator in humans when added to isoflurane. A transcranial Doppler study. Acta Anaesthesiol Scand 1995; 39: 653– 8 117 Abdullah NB, Rahman RA, Maaya M, Zain JM. Oxygen-air mixture does not differ from oxygen-nitrous oxide mixture in influencing PF ratio in healthy patients undergoing laparoscopic cholecystectomy. Int Med J 2010; 17: 205–7 118 Drummond GB. Breathing pattern during isoflurane anaesthesia with or without nitrous oxide. Br J Anaesth 1986; 58: 586–92 119 Einarsson SG, Cerne A, Bengtsson A, Stenqvist O, Bengtson JP. Respiration during emergence from anaesthesia with desflurane/N2O vs. desflurane/air for gynaecological laparoscopy. Acta Anaesthesiol Scand 1998; 42: 1192–8 120 Einarsson SG, Bengtsson A, Stenqvist O, Bengtson JP. Decreased respiratory depression during emergence from anesthesia with sevoflurane/N2O than with sevoflurane alone. Can J Anaesth 1999; 46: 335– 41 121 Fujii Y, Tanaka H, Toyooka H. Intraoperative ventilation with air and oxygen during laparoscopic cholecystectomy decreases the degree of postoperative hypoxaemia. Anaesth Intensive Care 1996; 24: 42–4 122 Jensen AG, Kalman SH, Eintrei C, Fransson S-G, Morales O. Atelectasis and oxygenation in major surgery with either propofol with or without nitrous oxide or isoflurane anaesthesia. Anaesthesia 1993; 48: 1094– 6 123 Joyce CJ, Baker AB. Effects of inspired gas composition during anaesthesia for abdominal hysterectomy on postoperative lung volumes. Br J Anaesth 1995; 75: 417– 21 124 Kumakura S, Kikuchi T, Yamaguchi K, Kugimiya T, Inada E. Exposure to nitrous oxide may increase airway inflammation during sevoflurane anesthesia. Masui 2008; 57: 1200–6 125 Lampe GH, Wauk LZ, Whitendale P, et al. Postoperative hypoxemia after nonabdominal surgery: a frequent event not caused by nitrous oxide. Anesth Analg 1990; 71: 597–601

Page 15 of 17

1625

1630

1635

1640

1645

1650

1655

1660

1665

1670

1675

1680

BJA

1685

1690

1695

1700

1705

1710

1715

1720

1725

1730

1735

126 Logan DA, Spence AA, Smith G. Postoperative pulmonary function. A comparison of ventilation with nitrogen or nitrous oxide during anaesthesia. Anaesthesia 1977; 32: 3–7 127 Maroof M, Khan RM, Siddique M. Ventilation with nitrous oxide during open cholecystectomy increases the incidence of postoperative hypoxemia. Anesth Analg 1993; 76: 1091– 4 128 Roberts CJ, Parke TJ, Sykes MK. Effect of intraoperative inspired gas mixtures on postoperative nocturnal oxygen saturation. Br J Anaesth 1993; 71: 476–80 129 Giuffre M, Gross JB. The effects of nitrous oxide on postoperative bowel motility. Anesthesiology 1986; 65: 699– 700 130 Goto T, Matsukawa T, Sessler DI, et al. Thermoregulatory thresholds for vasoconstriction in patients anesthetized with various 1-minimum alveolar concentration combinations of xenon nitrous oxide, and isoflurane. Anesthesiology 1999; 91: 626 – 32 131 Jensen AG, Kalman SH, Nystrom P-O, Eintrei C. Anaesthetic technique does not influence postoperative bowel function: a comparison of propofol, nitrous oxide and isoflurane. Can J Anaesth 1992; 39: 938– 43 132 Lampe GH, Wauk LZ, Whitendale P, Way WL, Murray W, Eger EI. Nitrous oxide does not impair hepatic function in young or old surgical patients. Anesth Analg 1990; 71: 606– 9 133 Mann MS, Woodsford PV, Jones RM. Anaesthetic carrier gases. Their effect on middle-ear pressure peri-operatively. Anaesthesia 1985; 40: 8–11 134 Pedersen FM, Wilken-Jensen C, Knudsen F, Lindekaer AL, Svare EI. The influence of nitrous oxide on recovery of bowel function after abdominal hysterectomy. Acta Anaesthesiol Scand 1993; 37: 692– 6 135 Scheinin B, Lindgren L, Scheinin TM. Peroperative nitrous oxide delays bowel function after colonic surgery. Br J Anaesth 1990; 64: 154– 8 136 Suttner SW, Lang K, Boldt J, Kumle B, Maleck WH, Piper SN. The influence of hyperoxic ventilation during sodium nitroprussideinduced hypotension on skeletal muscle tissue oxygen tension. Anesthesiology 2002; 96: 1103– 8 137 Waldman FM, Koblin DD, Lampe GH, Wauk LZ, Eger EI. Hematologic effects of nitrous oxide in surgical patients. Anesth Analg 1990; 71: 618– 24 138 Akhtar TM, Kerr WJ, Kenny GNC. Effect of nitrous oxide on postoperative nausea and vomiting during propofol anaesthesia for short surgical operations. Eur J Anaesthesiol 1993; 10: 337 – 41 139 Apfel CC, Bacher A, Biedler A, et al. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. Anaesthesist 2005; 54: 201– 9 140 Apfel CC, Korttila K, Abdalla M, et al. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. N Engl J Med 2004; 350: 2441–51 141 Apfel CC, Kranke P, Katz MH, et al. Volatile anaesthetics may be the main cause of early but not delayed postoperative vomiting: a randomized controlled trial of factorial design. Br J Anaesth 2002; 88: 659– 68 142 Bloomfield E, Porembka D, Grimes-Rice M. Avoidance of nitrous oxide and increased isoflurane during alfentanil based anesthesia decreases the incidence of postoperative nausea. Anesth Prog 1997; 44: 27–31 143 Felts JA, Poler SM, Spitznagel EL. Nitrous oxide, nausea, and vomiting after outpatient gynecologic surgery. J Clin Anesth 1990; 2: 168– 71

Page 16 of 17

Imberger et al.

144 Hovorka J, Korttila K, Erkola O. Nitrous oxide does not increase nausea and vomiting following gynaecological laparoscopy. Can J Anaesth 1989; 36: 145– 8 145 Ichinohe T, Kaneko Y. Nitrous oxide does not aggravate postoperative emesis after orthognathic surgery in female and nonsmoking patients. J Oral Maxillofac Surg 2007; 65: 936– 9 146 Katayama M, Vieira JI, De Alessio FP, Dos Santos GV, Gomes CC. Importance of droperidol and metoprolol in anesthesia with propofol, alfentanil and nitrous oxide for septoplasties and turbinectomies. Rev Bras Anestesiol 1995; 45: 225– 34 147 Korttila K, Hovorka J, Erkola O. Nitrous oxide does not increase the incidence of nausea and vomiting after isoflurane anesthesia. Anesth Analg 1987; 66: 761–5 148 Leslie K, Myles PS, Chan MT, et al. Risk factors for severe postoperative nausea and vomiting in a randomized trial of nitrous oxide-based vs nitrous oxide-free anaesthesia. Br J Anaesth 2008; 101: 498– 505 149 Lonie DS, Harper NJ. Nitrous oxide anaesthesia and vomiting. The effect of nitrous oxide anaesthesia on the incidence of vomiting following gynaecological laparoscopy. Anaesthesia 1986; 41: 703 – 7 150 Mraovic B, Simurina T, Sonicki Z, Skitarelic N, Gan TJ. The dose-response of nitrous oxide in postoperative nausea in patients undergoing gynecologic laparoscopic surgery: a preliminary study. Anesth Analg 2008; 107: 818– 23 151 Muir JJ, Warner MA, Offord KP, Buck CF, Harper JV, Kunkel SE. Role of nitrous oxide and other factors in postoperative nausea and vomiting: a randomized and blinded prospective study. Anesthesiology 1987; 66: 513–8 152 Nader ND, Simpson G, Reedy RL. Middle ear pressure changes after nitrous oxide anesthesia and its effect on postoperative nausea and vomiting. Laryngoscope 2004; 114: 883– 6 153 Paredi G, Sofi G, Bonazzi M, et al. Role of nitrous oxide on the incidence of postoperative emesis: randomized, double blinded study. Acta Anaesthesiol Ital 1994; 45: 239– 42 154 Piper SN, Rohm KD, Boldt J, et al. Inspired oxygen fraction of 0.8 compared with 0.4 does not further reduce postoperative nausea and vomiting in dolasetron-treated patients undergoing laparoscopic cholecystectomy. Br J Anaesth 2006; 97: 647– 53 155 Ranta P, Nuutinen L, Laitinen J. The role of nitrous oxide in postoperative nausea and recovery in patients undergoing upper abdominal surgery. Acta Anaesthesiol Scand 1991; 35: 339–41 156 Sengupta P, Plantevin OM. Nitrous oxide and day-case laparoscopy: effects on nausea, vomiting and return to normal activity. Br J Anaesth 1988; 60: 570– 3 157 Sukhani R, Lurie J, Jabamoni R. Propofol for ambulatory gynecologic laparoscopy: does omission of nitrous oxide alter postoperative emetic sequelae and recovery? Anesth Analg 1994; 78: 831– 5 158 Taki K, Sugimura M, Morimoto Y, et al. The influences of nitrous oxide inhalation during general anesthesia on postoperative nausea and vomiting. J Jpn Dent Soc Anesthesiol 2003; 31: 268– 73 159 Ture H, Takil A, Eti Z, Gogus FY. The effect of nitrous oxide on postoperative nausea and vomiting. Marmara Med J 2007; 20: 85– 91 160 Vanacker BF. The impact of nitrous oxide on postoperative nausea and vomiting after desflurane anesthesia for breast surgery. Acta Anaesthesiol Belg 1999; 50: 77 –81 161 Lederhaas G, Brock-Utne JG, Negrin RS, Riley E, Brodsky JB. Is nitrous oxide safe for bone marrow harvest? Anesth Analg 1995; 80: 770– 2

1740

1745

1750

1755

1760

1765

1770

1775

1780

1785

1790

BJA

Nitrous oxide and cardiovascular outcomes

1795

1800

162 Fleischmann E, Lenhardt R, Kurz A, et al. Nitrous oxide and risk of surgical wound infection: a randomised trial. Lancet 2005; 366: 1101– 7 163 Mayzler O, Weksler N, Domchik S, Klein M, Mizrahi S, Gurman GM. Does supplemental perioperative oxygen administration reduce the incidence of wound infection in elective colorectal surgery? Minerva Anestesiol 2005; 71: 21– 5 164 Thorlund K, Imberger G, Johnston BC. Evolution of heterogeneity (I 2) estimates and their 95% confidence intervals in large meta-analyses. PLoS One 2012; 7: e39471

Handling editor: J. G. Hardman

Appendix 1—search strategies 1805

1810

1820

1825

#1 #2 #3 #4 #5

# 5 #4 AND #3 # 4 TS¼(random* or placebo* or prospective* or multicenter*) or (TS¼(clinical or controlled) SAME TS¼(trial*)) or TS¼ ((single or double or triple or treble) SAME (mask* or blind*)) # 3 #2 AND #1 # 2 TS¼(surg* or anaesth* or anesth*) # 1 TS¼(nitrous or N2O)

† † † †

1 exp Nitrous Oxide/ 2 (nitrous or N2O).ti,ab. 3 1 or 2 (20607) 4 (surg* or ana?Sth*).mp. [mp¼title, original title, abstract, name of substance word, subject heading word, unique identifier] † 5 3 and 4 † 6 ((randomized controlled trial or controlled clinical trial).pt. or randomized.ab. or placebo.ab. or clinical trials as topic.sh. or randomly.ab. or trial.ti.) not (animals not (humans and animals)).sh. † 7 5 and 6 EMBASE (Ovid SP)

1830

† S5 S3 and S4 † S4 TI (surg* or anaesth* or anesth*) or AB (surg* or anaesth* or anesth*) † S3 S1 or S2 † S2 TI (nitrous or N2O) or AB (nitrous or N2O) † S1 (MM ‘Nitrous Oxide’) ISI Web of Science

MeSH descriptor Nitrous Oxide explode all trees (nitrous or N2O):ti,ab (#1 OR #2) (surg* or anaesth* or anesth*):ti,ab (#3 AND #4)

† † † † †

1 exp nitrous oxide/ 2 (nitrous or N2O).ti,ab. 3 1 or 2 4 (surg* or ana?Esth*).mp. 5 3 and 4

1850

CINAHL (EBSCO host)

CENTRAL

Medline (Ovid SP) 1815

† 6 (placebo.sh. or controlled study.ab. or random*.ti,ab. or trial*.ti,ab.) not (animals not (humans and animals)).sh. † 7 5 and 6

1855

1860

1865

1870

Appendix 2—definition of increased risk of cardiovascular complications We defined a participant as being at increased risk if they fulfilled any of the following criteria: † Underwent surgery defined as moderate or high risk surgery in the ACC/AHA (American College of Cardiology /American Heart Association) guidelines for perioperative cardiovascular evaluations for non-cardiac surgery or underwent cardiac surgery. We considered all day surgery as low risk of cardiovascular complications. † Had clinical predictors (mild, moderate, or severe) of increased perioperative cardiovascular risk as defined in the ACC/AHA guidelines for perioperative cardiovascular evaluations for non-cardiac surgery.

1875

1880

1885

We did not include day surgery patients as being at increased risk of cardiovascular complications. 1890

1835

1895 1840

1900 1845

Page 17 of 17