Muscle relaxation has been widely shown to facilitate tracheal intubation and decrease complications associated with both laryngoscopy and endo-tracheal tube placement in adults.1,2 Consequently, the use of muscle relaxant during induction of anaesthesia has become the standard of care when tracheal intubation is performed. However, this practice in children and infants is still a subject of discussion amongst paediatric anaesthesiologists.3 Detractors of the use of muscle relaxation for intubation express concerns about the risk of anaphylaxis associated with these compounds and the efficacy of alternative methods such as hypnotic agents4 with opioids, used in combination or independently,5 for improving intubation conditions. Supporters of relaxants commonly highlight the potential benefit of using these agents for better intubation conditions, as recommended in adults.
A recent quantitative analysis of intubation conditions using opioids5 found remifentanil combined with propofol to be the better association, providing 90% of acceptable intubation conditions in children. However, the use of suxamethonium was associated with an average 99% of acceptable intubation conditions. Although this study supported the use of muscle relaxation for intubation in children, there was no direct comparison between opioids and muscle relaxants.
Without firm evidence to guide the use of muscle relaxation during intubation in children, we decided to conduct a meta-analysis of controlled trials comparing opioids with muscle relaxant for intubation during paediatric anaesthesia. The primary outcome of the study was the percentage of children with excellent intubation conditions. The secondary outcomes of the study were the percentage of children with acceptable intubation conditions and the incidence of respiratory complications.
Materials and methods
Bibliographic search and analysis
We conducted this meta-analysis according to the guidelines of the Cochrane Handbook for Systematic Reviews,6 the PRISMA statements7 and the GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology.8,9 The search criteria and statistical methodologies were similar to previous works on other topics.10,11 There was no pre-published protocol for this meta-analysis.
Searched databases included PubMed, Embase Cochrane central register of controlled trials, clinical trials register and open-access journals not indexed in major databases (Directory of Open Access Journals, Open Journal of Anesthesiology, Anesthesiology Research and Practice, Journal of Anesthesia & Clinical Research, Journal of Anesthesiology & Clinical Science, Journal of Anesthesiology and Critical Care Medicine). The following keywords associated with ‘children or infant’ were ‘muscle relaxation’ or ‘atracurium’ or ‘mivacurium’ or ‘suxamethonium’ or ‘pancuronium’ or ‘rapacuronium’ or ‘cisatracrium’ or ‘rocuronium’.
Articles obtained from queries were independently analysed by four senior anaesthesiologists (results checked twice), and those with the following criteria were included in analysis: randomised controlled study, double blinded (especially the blinding of outcome assessment: the evaluation of intubation conditions unequivocally performed by an operator blinded to the drug administered), a standardised anaesthesia protocol for all patients and the presence of a control group using an opioid (studies with a placebo arm were excluded). Meeting abstracts were not included in this meta-analysis. These criteria allow us to select articles with the lowest methodological bias. The most recent search was April 2016.
Readers (two per article) assessed article quality and the presence of potential bias using the following criteria: randomisation, detailed description of methodology demonstrating whether intervention allocations could have been foreseen before or during enrolment, double-blind study, incomplete data report statements (excluded patients and data), selective reporting (presence of studied outcomes report verified) and additional bias including the clear definition of a main outcome and the method used for calculating the sample size to correctly power the study. Blinding of outcome assessment bias was considered as affecting overall results, and studies with this bias (either undetermined or high-risk) were excluded from the analysis.
Extracted data consisted of country where the study took place, the registration of the study in a trial database, patient ages, ASA physical status, the procedure performed, the hypnotic agents used during induction of anaesthesia, type and doses of opioid used for anaesthesia induction, type and dose of muscle relaxant during anaesthesia induction and respiratory complications during the intubation process. Intubation conditions were rated as excellent or acceptable, according to the points scale defined by the Copenhagen consensus conference:12 laryngoscopy, jaw relaxation, vocal cords movement, coughing, and limb movement. The primary outcome of this study was the proportion of children with excellent intubation conditions in the muscle relaxant and opioids groups. Secondary outcomes were the proportion of children with acceptable intubation conditions and the rate of postoperative respiratory complications in the muscle relaxant and opioids groups.
Classical meta-analysis was conducted using the Review Manager 5 software (RevMan 5.3, The Cochrane Collaboration; Oxford, United Kingdom) and the Trial Sequential Analysis (TSA) Software (Copenhagen Trial Unit's TSA Software; Copenhagen, Sweden) to evaluate the effect of random error and calculate the information size (the power of the meta-analysis). For this meta-analysis, we calculated the risk ratio (RR) using the Mantel–Haenszel method (RevMan) and the cumulative z-score curve (TSA). Finally, a summary of results was analysed using the GRADE methodology (using the online GRADEpro software1) to allow clear recommendations concerning the impact of muscle relaxant in comparison with opioids, on the quality of intubation during induction of anaesthesia in children and infants.8,9,13,14
Heterogeneity was assessed using I2 statistics, which describes the percentage of the variability in effect estimates (RR) due to heterogeneity rather than sampling error. According to the Cochrane review guidelines2, an I2 more than 40% and a P less than 0.1 were considered as the threshold for heterogeneity and indicated the need for a random effects model. When statistical heterogeneity was present, each study was removed from the overall analysis of the considered outcome to examine its effect on the overall heterogeneity.
Subgroup analyses were also performed (when at least two studies included the considered outcome) according to the type of hypnotic agent used (intravenous versus volatile), the dose of opioid used (according to Aouad's review5 with high doses defined as remifentanil ≥3 μg kg−1) and the type of muscle relaxant used (nondepolarising versus suxamethonium). Comparison between subgroups, called the interaction test, consisted of determining the heterogeneity of subgroups. The random effects model was also used in case of heterogeneity between studies concerning one of the following points: age of patients, anaesthesia protocols used during induction (intravenous versus volatile agents) and the nature of muscle relaxant used (suxamethonium versus nondepolarising).
Statistical methods are available to assess the effects of unpublished studies on the results of meta-analysis (publication bias). This type of bias is assessed by plotting the RR, or its logarithm, against a measurement of the precision of the RR (such as the standard error of the RR). This plot is named the Funnel plot, and its asymmetry indicates studies left unpublished because of negative results.15 Such asymmetry may also reveal data heterogeneity or poor methodology in included studies.15,16 Some studies, due to design, may also produce strongly positive results that lead to funnel plot asymmetry interpreted as ‘publication bias’. Finally, studies’ methodological bias could also give strongly positive results leading to funnel plot asymmetry.15,17 According to the Cochrane collaborative guideline3, publication bias can be assessed when analysing an aggregation of at least 10 studies.
Interaction tests between paired subgroups were expressed as χ2, df, I2 and P value for I2. Difference between subgroups was considered significant when P was less than 0.05. In studies with more than one opiate group, only the comparison between the muscle relaxant and the higher opiate dose was considered (to detect an effect of muscle relaxant on intubation conditions). However, sensitivity analyses were performed to study the impact of deleting all other groups on overall results. In the case of different results in overall analyses, subgroup analyses were undertaken to explore factors affecting results.
A second set of analyses were performed for both the primary and secondary outcomes using the trial sequential method.18,19 This method has been found to be more pertinent when analysing cumulative heterogeneous results, and it decreases type I error. TSA provides three important complementary data points compared with traditional meta-analysis: it combines results and provides a cumulated sample size of included trials using an approximate semi-Bayes procedure with an adjusted threshold for statistical significance and adjusted alpha risk to decrease type I error. The result is considered as significant when the boundary of significance is crossed.
The second is the effect of previous meta-analysis on overall results. Given that the alpha risk must be adjusted to the number of comparisons, TSA can perform corrections according to studies previously included in previous meta-analyses, whereas classical meta-analysis does not, leading to 20 to 30% of false-positive results.18
Lastly, TSA permits the description of further trial requirements, through a procedure known as trial sequential monitoring boundaries. When the cumulative z-scores for the studied outcome crosses the trial sequential monitoring boundary, the level of evidence for the intervention is considered reached, and no further trials are needed. TSA also determines the futility area, indicating that no significant result would be found with additional trials. Finally, the O’Brien–Fleming approach using the TSA (the number of patients to be included in the meta-analysis to reach the desired level of power) predicts the sample size to be included in future trials to achieve the desirable effect (according to current results) with sufficient statistical power. In our study, information size was computed for the primary study outcome assuming an alpha risk of 5%, a beta risk of 20% and a relative risk difference between muscle relaxant and opioid computed on overall available results of the classical meta-analysis method.
Finally, summary of overall analyses were included in the GRADE analysis. To assess the overall quality of evidence for each outcome (pooled data expressed as RR), we downgraded the evidence from ’high quality’ by one level for high-risk bias in included studies, indirectness of evidence, serious inconsistency (either because heterogeneity of the use of random statistical model), imprecision of effect estimates, potential publication bias (if number of included studies allow assessment) and the potential lack of power of analyses.13,14,20,21 Data from RevMan software were transferred in the GRADEPRO online programme4.
Results of intervention effects were expressed as RR (95% confidence interval), I2, P value for I2 statistics, number of trials for the outcome, number of patients and level of GRADE recommendation.
Using the above selection criteria, 1915 articles were identified, of which only seven were assessed as relevant (none published in an open-access journal).22–28 All data extracted were displayed as numerical values, and no data consisted of graphical conversion or median and range transformation. As a consequence, authors were not contacted for their results. The details of the selection process are summarised in Fig. 1, and the descriptions of included studies are displayed in Table 1. Bias was found in all studies except that of Devys et al.25 (Table 2). The most common form of bias identified was an undetermined risk of bias concerning the random sequence generation and the allocation concealment. Three studies26–28 did not display a primary outcome or did not state clearly the method used to compute the sample size for the main outcome (Table 2). All studies included in this meta-analysis used the recommended Copenhagen scale for the quality of intubation conditions except one27 that used three items (vocal cords, jaw relaxation and coughing) instead of the recommended five items (laryngoscopy, jaw relaxation, vocal cords, coughing and limb movement).
A total of 195 patients received a muscle relaxant, and 244 received an opioid. The secondary outcome – respiratory complication during intubation – was provided in only one study (Devys et al.25), and so this outcome was removed from the analysis. Finally, the study performed by Blair et al.23 included three opiate arms using 1, 2 or 3 μg kg−1 of remifentanil. As stated in the materials and methods section, the comparison between the higher opiate dose and the muscle relaxant was included in the analysis of overall results (and a sensitivity analysis was performed to assess effects of the other comparisons on the overall results).
When considering “excellent” intubation conditions, the administration of muscle relaxation (MR) during induction of anaesthesia did not provide better intubation conditions in comparison with opioids (RR = 1.17 [0.96, 1.43], I2 = 36%, P of I2 = 0.18, number of studies = 5, number of patients = 226: Fig. 2a). When considering “acceptable” intubation conditions, the administration of MR did provide better intubation conditions in comparison to opioids (RR = 1.25 [1.06, 1.47], I2 = 70%, P of I2 < 0.01, number of studies = 6, number of patients = 362: Fig. 2b).
Trial sequential analysis
TSA for the primary outcome (Fig. 3a) found that muscle relaxant had no significant effect on intubation conditions, and the z-curve did not cross the trial sequential monitoring boundary, which indicates that this effect remains uncertain. Moreover, given the informative size of 307 patients, this meta-analysis appears underpowered to pronounce on this outcome with only 226 patients included in the analysis. Finally, given that our meta-analysis is the only one performed on this specific topic, no correction for the alpha risk was needed.
The same analysis demonstrated a statistically significant advantage of muscle relaxant over opioid for acceptable intubation conditions as displayed in Fig. 3b (the z-line crossed the two-sided error target). In addition, the z-curve crossed the trial sequential monitoring boundary and did not cross the futility region. Moreover, the informative size based on the O’Brien–Fleming approach (the number of patients required to be included in the meta-analysis to reach the desired level of significance and power) found that 243 patients needed be included in this meta-analysis and this was exceed (362 recruited). Finally, given that our meta-analysis is the only one performed on this specific topic, no correction for the alpha risk was needed for the TSA.
Summary of the analysis is displayed in Table 3. The following adjustments were performed according to GRADE recommendations: one level downgrade for the bias of included studies (considered as serious), one level downgrade for inconsistency of results (considered as serious because of the heterogeneity of results and the random effects model used) and one level downgrade for the lack of power of the analysis concerning the primary outcome (insufficient studies).8,9,13,14,20,21,29–31 The quality of evidence for the second outcome was not downgraded because of lack of power of the analysis given that significant results were obtained, even when sensitivity analyses were performed. Overall, quality of evidence was found to be low for the primary outcome (RR = 1.17 [0.96, 1.43], I2 = 36%, P of I2 = 0.18, number of studies = 5, number of patients = 226, GRADE = low evidence: Table 3) and moderate for the secondary outcome (RR = 1.25 [1.06, 1.47], I2 = 70%, P of I2 < 0.01, number of studies = 6, number of patients = 362, GRADE = moderate evidence: Table 3).
Sensitivity analyses for excluded study arms
The study performed by Blair et al.23 included three opiate arms (1, 2 or 3 μg kg−1 of remifentanil). In that study,23 all were compared individually with the group receiving mivacurium. Initially, only the 3 μg kg−1 remifentanil group was included in our meta-analysis of the primary outcome (excellent intubation conditions, RR = 1.26 [0.88, 1.81], I2 = 67%, P of I2 = 0.03, number of studies = 5, number of patients = 225, GRADE = Low evidence) but including the 1 and 2 μg kg−1 remifentanil groups made no real difference to the result (RR = 1.28 [0.96, 1.69], I2 = 63%, P of I2 = 0.03, number of studies = 5, number of patients = 227, GRADE = Low evidence) As regards the secondary outcome (adequate intubation conditions), again there was little difference between using the data from only the 3 μg kg−1 remifentanil group or including the 1 and 2 μg kg−1 remifentanil groups: RR = 1.29 [1.07, 1.55], I2 = 73%, P of I2 < 0.01, number of studies = 6, number of patients = 361, GRADE = moderate evidence and RR = 1.35 [1.07, 1.70], I2 = 81%, P of I2 < 0.01, number of studies = 6, number of patients = 363, GRADE = moderate evidence, respectively.
Study bias analyses
Excluding studies with either a high-risk bias concerning the sample size calculation or those with an absence of a defined primary outcome (Table 2)26–28 did not alter the results for the primary outcome of this study, (RR = 1.3 [0.84, 2.02], I2 = 65%, P of I2 = 0.06, number of studies = 3, number of patients = 127, GRADE = low evidence), or the secondary outcome (RR = 1.41 [1.05, 1.88], I2 = 73%, P of I2 = 0.02, number of studies = 3, number of patients = 183, GRADE = low evidence: the GRADE evidence was downgraded because of the limited number of studies for this subgroup analysis). Finally, excluding the study of Ng et al.,27 in which not all items of intubation quality were used, did not alter the results for excellent intubation conditions (RR = 1.25 [0.98, 1.59], I2 = 42%, P of I2 = 0.16, number of studies = 4, number of patients = 187, GRADE = low evidence), or acceptable intubation conditions (RR = 1.30 [1.09, 1.55], I2 = 60%, P of I2 = 0.04, number of studies = 5, number of patients = 323, GRADE = moderate evidence).
To explore the heterogeneity of the secondary outcome, we removed studies individually from the analyses. Removal of the Blair et al.22 study (published in 2000) alone from the analysis led to a homogenous result without influencing the directness of results (RR = 1.13 [1.04, 1.24], I2 = 0%, P of I2 = 0.42, number of studies = 5, number of patients = 282, GRADE = moderate evidence).
Regarding the primary outcome (excellent intubation conditions), neither the anaesthetic agent (Fig. 4a) nor the muscle relaxant used affected the results (Fig. 4b). Concerning the secondary outcome (acceptable intubation conditions), the use of an intravenous hypnotic agent (Fig. 5a) or a nondepolarising muscle relaxant in the active group (Fig. 5b) was associated with improvement in intubation conditions when using muscle relaxation compared with opioids. In contrast, the use of volatile agents (Fig. 5a) or suxamethonium in the active group (Fig. 5b) was not associated with better intubation conditions when comparing muscle relaxant with opioids. However, given that no significant difference was found within these two subgroups, no conclusion can be drawn from these analyses. Subgroup analyses of the effect of opiate dose were not possible due to the lack of studies with a high dose group; just one study was found with a remifentanil bolus at least 3 μg kg−1.5
The main finding of this meta-analysis can be summarised as follows: acceptable intubation conditions are more frequently observed when a hypnotic drug is combined with a muscle relaxant than when it is combined with an opioid. However, our meta-analysis lacked sufficient statistical power to provide an answer on the effect of muscle relaxant on excellent intubation conditions. In addition, results were heterogeneous because of the difference in anaesthesia protocols used in included studies. The quality of evidence according to GRADE recommendations ranged from low for the primary outcome of this meta-analysis to moderate for the secondary one.
We failed to find a difference between muscle relaxant and opioid for the primary outcome. This was confirmed by TSA: that clearly identified a lack of power in the meta-analysis for detecting a difference between these two drugs. Interestingly, this result remained unchanged when performing subgroup analyses according to confounding factors (intravenous versus volatile anaesthetic agents or suxamethonium versus nondepolarising relaxant) or when considering sensitivity analyses according to study bias. Results for the primary outcome indicate that further investigation is required to clarify whether muscle relaxant or opioid is better for intubation of children. An estimated 307 patients were needed to allow conclusions on this outcome, which means that 81 additional patients (41 in each arm) remain to be included in further trials.
In contrast, results of the current meta-analysis indicate that acceptable intubation conditions were significantly more frequent when using a muscle relaxant instead of an opioid agent. This was confirmed by the TSA analysis where we found a significant benefit of muscle relaxants over opioids following an analysis that had sufficient power. In addition, a homogenous result (with no impact on the result) was obtained when removing the study published by Blair et al. that used a low dose of alfentanil (10 μg kg−1) as a comparator to suxamethonium, with more frequent acceptable intubation conditions in the muscle relaxant group. Subgroup analyses of this outcome also found non-significant differences between subgroups and did not permit any conclusion regarding the confounding factors (anaesthetics or muscle relaxants). The significant result for this outcome (acceptable intubation conditions) supports previous studies that found suxamethonium providing acceptable intubation conditions in 99% of cases (using doses ranging from 1 to 2 mg kg−1 at 45 to 60 s after its injection). Nondepolarising relaxants also led to improvement of intubation conditions in more than 95% of cases, using the following agents: rocuronium, 30 s after injection of 1.2 mg kg−132–34, 60 s after injection of 0.6 mg kg−1, and 120 s after 0.3 mg kg−1; mivacurium, 90 s after injection of 0.2 mg kg−1 or 45 s after 0.3 mg kg−135,36; cisatracrium, 90 s after injection of 0.2 mg kg−137; and atracurium, 180 s after injection of 0.4 to 0.5 mg kg−1.38,39 Studies that compared standard doses of nondepolarising relaxant with suxamethonium did not find any significant difference in the provision of better intubation conditions using either agent.40–42 The rapid onset of arterial oxygen desaturation, especially in infants, strongly encourages the use of muscle relaxants to secure the airway, thus reducing the duration of apnoea and limiting the resulting arterial oxygen desaturation.43,44 This is particularly underlined by the study of Devys et al.25 where it was found that rocuronium reduced the incidence of adverse respiratory events and desaturation in comparison with alfentanil (20 μg kg−1).
The presence of excellent intubation conditions is a key point, given that adult studies have found the incidence of complications associated with intubation, such as airway injury, decreased when intubation conditions were excellent.1 For children, additional data are needed to reach any conclusion regarding airway injury and intubation conditions. Unlike adults, most adverse effects reported during paediatric intubation are related to the duration of apnoea and the resulting arterial desaturation rather than to airway injury.44 The second great limitation of this meta-analysis is the great heterogeneity observed, despite performing subgroup analyses. This was probably related to the variability in the range of ages of the children and in the anaesthesia protocols (variations in the hypnotic, the muscle relaxant and the opioid administered: Table 1) of the included studies, thus justifying the use of a random model in all statistical analyses.12 Studies included in this meta-analysis were almost all biased. However, studies with bias affecting results (blinding of outcome assessment) were discarded from this analysis. Bias detected in included studies consisted of an undetermined bias and high-risk bias resulting from the absence (or clear description) of a sample size calculation. Finally, as stated earlier, this meta-analysis was underpowered for a comparison between muscle relaxant and opioids on the primary outcome (excellent conditions for intubation). The limited number of studies exploring effects of muscle relaxants in paediatrics also impairs subgroup analyses. The impact of high-dose opioids could not be studied because only one study used a high dose of opioid (subgroup analyses were performed if at least two studies were available).
The GRADE analyses of the use of a muscle relaxant instead of an opioid found a low quality of evidence for excellent intubation conditions and a moderate quality of evidence for acceptable intubation conditions (Table 3). Quality of evidence was subject to the bias of included studies, the heterogeneity of results (and anaesthesia protocols used in included studies) and the absence of adequate power to explore the primary outcome. Consequently, in comparison with opioid, muscle relaxant should probably be recommended for obtaining acceptable intubation conditions in children. However, no recommendation can be made for excellent intubation conditions. Overall, these results suggest that during anaesthesia for children and infants, muscle relaxants may be recommended for intubation instead of opioids to improve intubation conditions. Further paediatric studies, including at least 81 children, are necessary to assess the effect of muscle relation in the production of excellent intubation conditions and a reduction in the incidence of respiratory complications related to direct laryngoscopy and tracheal intubation.
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: none.
Conflicts of interest: none.
1. Mencke T, Echternach M, Kleinschmidt S, et al. Laryngeal morbidity and quality of tracheal intubation: a randomized controlled trial. Anesthesiology
2. Baillard C, Adnet F, Borron SW, et al. Tracheal intubation in routine practice with and without muscular relaxation: an observational study. Eur J Anaesthesiol
3. Meakin GH. Role of muscle relaxants in pediatric anesthesia. Curr Opin Anaesthesiol
4. Kim SH, Hong JY, Suk EH, et al. Optimum bolus dose of propofol for tracheal intubation during sevoflurane induction without neuromuscular blockade in children. Anaesth Intensive Care
5. Aouad MT, Yazbeck-Karam VG, Mallat CE, et al. The effect of adjuvant drugs on the quality of tracheal intubation without muscle relaxants in children: a systematic review of randomized trials. Paediatr Anaesth
6. Cochrane C. Cochrane handbook 2016. http://www.cochrane-handbook.org
7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. J Clin Epidemiol
8. Atkins D, Best D, Briss PA, et al. Grading quality of evidence and strength of recommendations. BMJ
9. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ
10. Bellon M, Le Bot A, Michelet D, et al. Efficacy of intraoperative dexmedetomidine compared with placebo for postoperative pain management: a meta-analysis of published studies. Pain Ther
11. Michelet D, Hilly J, Skhiri A, et al. Opioid-sparing effect of ketamine in children: a meta-analysis and trial sequential analysis of published studies. Paediatr Drugs
12. Viby-Mogensen J, Engbaek J, Eriksson LI, et al. Good clinical research practice (GCRP) in pharmacodynamic studies of neuromuscular blocking agents. Acta Anaesthesiol Scand
13. Alonso-Coello P, Oxman AD, Moberg J, et al. GRADE evidence to decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 2: Clinical practice guidelines. BMJ
14. Alonso-Coello P, Schünemann HJ, Moberg J, et al. GRADE evidence to decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 1: Introduction. BMJ
15. Sterne JA, Egger M, Smith GD. Systematic reviews in healthcare: investigating and dealing with publication and other biases in meta-analysis. BMJ
16. Sutton AJ, Higgins JP. Recent developments in meta-analysis. Stat Med
17. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ
18. Afshari A, Wetterslev J. When may systematic reviews and meta-analyses be considered reliable? Eur J Anaesthesiol
19. Higgins JP, Whitehead A, Simmonds M. Sequential methods for random-effects meta-analysis. Stat Med
20. Andrews J, Guyatt G, Oxman AD, et al. GRADE guidelines: 14. Going from evidence to recommendations: the significance and presentation of recommendations. J Clin Epidemiol
21. Andrews JC, Schünemann HJ, Oxman AD, et al. GRADE guidelines: 15. Going from evidence to recommendation-determinants of a recommendation's direction and strength. J Clin Epidemiol
22. Blair JM, Hill DA, Bali IM, et al. Tracheal intubating conditions after induction with sevoflurane 8% in children. A comparison with two intravenous techniques. Anaesthesia
23. Blair JM, Hill DA, Wilson CM, et al. Assessment of tracheal intubation in children after induction with propofol and different doses of remifentanil. Anaesthesia
24. Crawford MW, Hayes J, Tan JM. Dose-response of remifentanil for tracheal intubation in infants. Anesth Analg
25. Devys JM, Mourissoux G, Donnette FX, et al. Intubating conditions and adverse events during sevoflurane induction in infants. Br J Anaesth
26. Morgan JM, Barker I, Peacock JE, et al. A comparison of intubating conditions in children following induction of anaesthesia with propofol and suxamethonium or propofol and remifentanil. Anaesthesia
27. Ng KP, Wang CY. Alfentanil for intubation under halothane anaesthesia in children. Paediatr Anaesth
28. Steyn MP, Quinn AM, Gillespie JA, et al. Tracheal intubation without neuromuscular block in children. Br J Anaesth
29. Russell SJ, Vowels MR, Vale T. Haemorrhagic cystitis in paediatric bone marrow transplant patients: an association with infective agents, GVHD and prior cyclophosphamide. Bone Marrow Transplant
30. Salazar OM, Castro-Vita H, VanHoutte P, et al. Improved survival in cases of intracranial ependymoma after radiation therapy. Late report and recommendations. J Neurosurg
31. Selcuk N, Elevli M, Inanc D, et al. Atypical teratoid/rhabdoid tumor mimicking tuberculous meningitis. Indian Pediatr
32. Eikermann M, Hunkemoller I, Peine L, et al. Optimal rocuronium dose for intubation during inhalation induction with sevoflurane in children. Br J Anaesth
33. Eikermann M, Renzing-Kohler K, Peters J. Probability of acceptable intubation conditions with low dose rocuronium during light sevoflurane anaesthesia in children. Acta Anaesthesiol Scand
34. Fuchs-Buder T, Tassonyi E. Intubating conditions and time course of rocuronium-induced neuromuscular block in children. Br J Anaesth
35. Green DW, Fisher M, Sockalingham I. Mivacurium compared with succinylcholine in children with liver disease. Br J Anaesth
36. Simhi E, Brandom BW, Lloyd ME, et al. Intubation in children after 0.3 mg/kg of mivacurium. J Clin Anesth
37. Kenaan CA, Estacio RL, Bikhazi GB. Pharmacodynamics and intubating conditions of cisatracurium in children during halothane and opioid anesthesia. J Clin Anesth
38. Lavery GG, Mirakhur RK. Atracurium besylate in paediatric anaesthesia. Anaesthesia
39. Ved SA, Chen J, Reed M, et al. Intubation with low-dose atracurium in children. Anesth Analg
40. Mangat PS, Evans DEN, Harmer M, et al. A comparison between mivacurium and suxamethonium in children. Anaesthesia
41. Naguib M, Samarkandi AH, Ammar A, et al. Comparison of suxamethonium and different combinations of rocuronium and mivacurium for rapid tracheal intubation in children. Br J Anaesth
42. Stoddart PA, Mather SJ. Onset of neuromuscular blockade and intubating conditions one minute after the administration of rocuronium in children. Paediatr Anaesth
43. Engelhardt T, Weiss M. A child with a difficult airway: what do I do next? Curr Opin Anaesthesiol
44. Kinouchi K, Tanigami H, Tashiro C, et al. Duration of apnea in anesthetized infants and children required for desaturation of hemoglobin to 95%. The influence of upper respiratory infection. Anesthesiology
1 https://gradepro.org/. Last assess October 2016.
2 http://www.cochrane-handbook.org/ (Section 9.5.2). Last access April 2016.
3 http://www.cochrane-handbook.org/ (Section 10.4.3.1). Last access April 2016.
4 Last assess October 2016.