Airway management is required for anaesthetised patients undergoing surgical or diagnostic procedures and is essential for life-threatening conditions such as cardiopulmonary resuscitation and critical care. Several guidelines have been published to standardise airway management and tracheal intubation procedures for routine and emergency situations in patients with normal, known, or anticipated difficult airway.1–4 However, there are no specific guidelines for neonates and infants.

Children have a unique anatomy and physiology that can present clinicians with significant challenges.5 Younger children, term neonates, and pre-term neonates are at the highest risk for respiratory and traumatic complications from airway management. Most devices available for airway management are not specifically designed or tested for use in children.3,6–8

This practice guideline aims to provide an evidence-based approach to airway management in neonates and infants. It was developed by a core group of experts in paediatric airway management, with the intention to serve anaesthetists working in a variety of paediatric settings, from highly specialised to district centres. As expertise and resources differ across centres, these practice guidelines are not a standard of care, however, they should serve as a basis for developing local institutionally approved operating procedures and best practice guidelines.

Methods

In 2021, 23 internationally recognised experts in airway management formed a task force aimed at writing updated practice guidelines for airway management in neonates and infants. The proposal was submitted to the European Society of Anaesthesiology and Intensive Care (ESAIC) and to the Editorial Board of the British Journal of Anaesthesia (BJA) for logistic support and endorsement. A joint endeavour between ESAIC and BJA was approved.

Clinical queries were developed in the form of seven Population, Intervention, Comparison, Outcome (PICO) groups and further developed into five elements for the search strategy. The complete list of PICO groups was then revised and approved by the task force, and generated the following research questions:

1. Is physical assessment the best way to predict difficult airway management? Can physical assessment be improved by further measurements? Which normal values should be used? 2. What type of preparation and planning should be mandatory before starting airway management in neonates and infants? Is neuromuscular block recommended for tracheal intubation if maintaining spontaneous breathing is not necessary? 3. Is direct laryngoscopy the first-choice technique for tracheal intubation in neonates and infants? Should direct laryngoscopy be replaced by other techniques? What is the definition of a difficult intubation? Is a tracheal intubation algorithm needed? 4. What should be the gold standard for anticipated difficult airway management, and who should be involved? Where should difficult airway management be performed? 5. Which technique should be used for detecting the correct position of the tracheal tube in neonates and infants? 6. What is the best strategy for safe tracheal extubation in neonates and infants after difficult intubation, either anticipated or unanticipated? Should extubation or removal of an airway device be performed under deep anaesthesia or when the patient is awake? 7. What is the impact of human factors and the need for developing a specific paediatric airway curriculum? Criteria for considering studies for data analysis

Types of studies: Data analysis was based on all randomised, parallel, quasi-randomised studies (including crossover design) and observational studies that addressed the above queries. Systematic reviews and meta-analyses were considered on a case-by-case basis when meeting inclusion criteria. Data from quasi-randomised, observational and large retrospective studies were included as very few, if any, randomised controlled trials (RCTs) were anticipated.

Types of participants: The qualitative and quantitative analysis of the literature was confined to children up to 1 yr of age, with or without specified comorbidities. Studies including a mix of paediatric and adult populations were reviewed only if they included a relevant number of infants. In case there was a lack of data for neonates and infants undergoing airway management in the operating room, and if considered relevant, data were extrapolated from non-operating room settings (i.e. neonatal and paediatric intensive care unit and emergency department).

Types of interventions: We included the following interventions: i) physical assessment for detecting or predicting a potentially difficult airway; ii) child and staff preparation for airway management, including pharmacological treatment; iii) direct laryngoscopy for tracheal intubation; iv) specific competency and techniques for expected difficult intubation; v) chest auscultation confirmation of correct tracheal tube position; vi) strategies for tracheal extubation in children in whom tracheal intubation was difficult, including use of an airway management debriefing; vii) influence of human factors and competencies on successful airway management.

Types of comparators: Any technique or strategy for preparation, or both, initial and intraoperative management, and extubation different from the above-mentioned ‘interventions’ were considered as comparators, both in routine care and difficult airway management, either expected or unexpected.

Types of outcomes: First-pass tracheal intubation success, number of attempts until successful intubation was achieved, and any complication during and after airway management were considered as outcomes.

Search methods for identification of studies

An information specialist (AC) developed the literature search strategy in close collaboration with ND and the ESAIC group methodologists (AA, PK, and CSR). The literature search was conducted in PubMed, EMBASE, Web of Science, and Cochrane Central Register of Controlled Trials (CENTRAL). A similar search strategy was used for all databases. The electronic database search was run on November 17, 2021 by AC, and included articles published since 2011 to increase clinical relevance (Appendix 1). Panel members were also encouraged to add any missing papers of interest that they were aware of, and to conduct related searches themselves. The resulting titles were screened by two independent authors. A third author was assigned to resolve conflicts for inclusion or exclusion.

Search results

After removing duplicates, the authors screened titles with abstracts in a two-stage procedure. Relevant papers were retrieved for full text assessment and data extraction by task force subgroups who compiled and wrote the literature review for their respective PICO groups. The methodologist was responsible for choosing topics for possible meta-analyses based on the quality of the available data, reliability of the search (sensitivity), and the predefined inclusion and exclusion criteria. For this guideline, we found no data suitable for meta-analysis.

For simplicity of the search, screening, inclusion and exclusion of studies, all retrieved titles and articles were allocated into five folders: A) preparation of airway management, B) tracheal intubation, C) difficult airway, D) tracheal extubation, and E) human factors. The retrieved articles were:

A) From 972 publications on preparation, after removal of duplicates and date restrictions, the remaining 466 titles were screened, resulting in 60 abstracts. For the next step, 22 full articles were included. B) From 3991 publications on tracheal intubation, after removal of duplicates and date restrictions, the remaining 2486 titles were screened, resulting in 532 abstracts. For the next step, 94 full articles were included. C) From 570 publications on difficult airway, after removal of duplicates and date restrictions, the remaining 278 titles were screened, resulting in 191 abstracts. For the next step, 86 full articles were included. D) From 6507 titles on tracheal extubation, after removal of duplicates and date restrictions, the remaining 4637 titles were screened, resulting in 613 abstracts. For the next step, 180 full articles were included. E) From the 994 titles on human factors, 992 were screened, 64 abstracts were selected, and 37 full articles were used for a short scoping review.

A detailed description of the search strategy and PICO are shown in Appendix 1.

Data collection and analysis

Selection of studies: all publications meeting inclusion criteria were included. At least two authors in each PICO group assessed the relevant full text articles independently by using Covidence software. Disagreements were resolved by a member not involved in the screening.

Data extraction and management: Each task force group extracted data from relevant studies in a similar fashion with guidance from the methodologist using similar spreadsheets, including information on study design, population characteristics, interventions, and outcome measures. Task force group authors reached a consensus regarding extracted data through discussion, initially within the group, and secondly within the entire task force.

Assessment of risk of bias in included studies: Risk of bias was assessed for each PICO group in accordance with the Cochrane Handbook for Systematic Reviews of Interventions source for RCTs,9 and was assessed for the following domains:

Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of outcome assessors (performance and detection bias) Incomplete outcome data, intention-to-treat (attrition bias) Selective reporting

Trials were assessed as having a low risk of bias if all of the domains were considered adequate, medium risk if one domain was inadequate, and high risk if more than one domain was considered inadequate or unclear. Disagreement regarding assessment of the risk of bias was settled in a discussion with the methodologist (AA). For non-RCTs, checklists from SIGN (Scottish Intercollegiate Guidelines Network;https://www.sign.ac.uk/what-we-do/methodology/checklists/) were applied.

Assessment of the quality of the evidence: In accordance with the ESAIC guidelines policy, the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) methodology was used to assess methodological quality and to formulate recommendations.

Decisions to downgrade the level of evidence for a recommendation were based on the quality and type of the included literature, observed inconsistencies, indirectness or directness of the evidence, overall impression, and the presence of publication bias as proposed by GRADE. Decisions to upgrade the level of evidence for recommendations were based on study quality and magnitude of the effect ratio, dose-response gradient, and plausible confounding. The GRADE definitions are summarised in Table 1. A more detailed account of GRADE is available at https://www.uptodate.com/home/grading-guide.

Table 1 - GRADE definitions. GRADE, Grading of Recommendations, Assessment, Development and Evaluation. Grade of recommendation Clarity of risk/benefit Quality of supporting evidence 1A
Strong recommendation, high-quality evidence Benefits clearly outweigh risk and burdens, or vice versa Consistent evidence from well performed randomised, controlled trials or overwhelming evidence of some other form; further research is unlikely to change our confidence in the estimate of benefit and risk 1B
Strong recommendation, moderate quality evidence Benefits clearly outweigh risk and burdens, or vice versa Evidence from randomised, controlled trials with important limitations (inconsistent results, methodological flaws, indirect or imprecise), or very strong evidence of some other research design; further research (if performed) is likely to have an impact on our confidence in the estimate of benefit and risk and may change the estimate 1C
Strong recommendation, low-quality evidence Benefits appear to outweigh risk and burdens, or vice versa Evidence from observational studies, unsystematic clinical experience, or from randomised, controlled trials with serious flaws; any estimate of effect is uncertain 2A
Weak recommendation=suggestion, high-quality evidence Benefits closely balanced with risks and burdens Consistent evidence from well-performed randomised, controlled trials or overwhelming evidence of some other form; further research is unlikely to change our confidence in the estimate of benefit and risk 2B
Weak recommendation=suggestion, moderate-quality evidence Benefits closely balanced with risks and burdens, some uncertainty in the estimates of benefits, risks, and burdens Evidence from randomised, controlled trials with important limitations (inconsistent results, methodological flaws, indirect or imprecise), or very strong evidence of some other research design; further research (if performed) is likely to have an impact on our confidence in the estimate of benefit and risk and may change the estimate 2C
Weak recommendation=suggestion, low-quality evidence Uncertainty in the estimates of benefits, risks, and burdens; benefits may be closely balanced with risks and burdens Evidence from observational studies, unsystematic clinical experience, or from randomised, controlled trials with serious flaws; any estimate of effect is uncertain

Development of recommendations: After the above procedures, each group developed recommendations relevant to their PICO group and clinical questions. These were then discussed and rediscussed, as required, with the entire expert panel in light of the data synthesis (when available), the risk of bias, and the quality of the evidence.

One general limitation across all PICO groups was heterogeneity of the case mix, such that many studies included children >1 yr old. Consequently, the task force assessed external validity and generalisability of the findings of the included studies for each recommendation. When considered reasonable, evidence from older patients was used for drafting recommendations, suggestions, or clinical practice statements.

A two-step Delphi process using an online survey was used to produce expert recommendations and degree of agreement. A third Delphi round was performed by teleconference to discuss methodological quality of the supporting literature and when rephrasing recommendations was mandated. The same voting and consensus processes were applied to each recommendation, suggestion, or clinical practice statement.

Summary of recommendations, suggestions and clinical practice statements.

PICO 1. Preoperative airway assessment to predict difficulty Is physical examination the best way to predict difficult airway management? Can physical examination be implemented by further measures? Which normal values should be used?

Recommendation: We recommend use of medical history and physical examination to predict difficult airway management in neonates and infants (1C).

Evidence summary: No evidence is available supporting or refuting the use of physical assessment to predict difficult airway management in neonates and infants. However, clinician experience supported by several retrospective studies identified some physical risk factors for a difficult airway in neonates and infants: micro-, retro-, or prognathia, limited mouth opening, facial asymmetry, fixed cervical spine, labio-palatine cleft, and oral or neck mass.10–12 These physical features commonly exist in neonates and infants with syndromes associated with anticipated difficult airway management such as the Pierre Robin, Crouzon, and Treacher-Collins syndromes.11,13 However, there are no studies prospectively evaluating the role of physical features in predicting difficult airway management (mask ventilation or tracheal intubation) in neonates and infants with an apparently normal airway. While several adult based studies have suggested simple summations or weighted risk scores for predicting ‘unanticipated’ difficult tracheal intubation based on physical measures (e.g. symptoms of a pathological airway; inter-incisor gap; mandible luxation; thyromental distance; head and neck movement; and Mallampati score), none of these have been validated in children nor is it always possible to assess these parameters.14–16 The Mallampati score is not feasible in neo-nates and infants as it requires a cooperative patient who can follow verbal commands. Other measures (e.g. measuring thyromental distance) are not always possible and are subject to dramatic changes as neonates and infants grow rapidly in the first months of life. Therefore, no normal values are available for neonates and infants.

A single centre case series of eight infants aged 3 weeks to 1 yr has shown the potential benefit of using volumetric computer tomography (CT) imaging to delineate abnormalities of the lower trachea and lungs (Supplementary Table S1, https://links.lww.com/EJA/A891).17 However, the case series did not provide data for upper airway abnormalities that can be associated with difficult airway management. No other measures were found to help with prediction of difficult airway in neonates and infants.

PICO 2. Preparation for airway management and pharmacological treatment (outside resuscitation) What preparation and planning should be mandatory before airway management in neonates and infants? Is neuromuscular block mandatory if spontaneous breathing is not necessary (pharmacology)?

Recommendation: We recommend use of an adequate level of sedation or general anaesthesia in neonates and infants during airway management to ensure patient comfort and safety (1B).

Recommendation: We recommend use of neuromuscular block before tracheal intubation when maintaining spontaneous breathing is not necessary (1C). The risks and benefits of neuromuscular blocking agent administration should be balanced for the individual patient and team skills.

Evidence summary: Except in cases of resuscitation (e.g. in the delivery room), tracheal intubation in neonates and infants with minimal or no anaesthesia is a largely abandoned clinical practice. The reasons for promoting awake intubation in the past were varied, including fear of adverse effects such as pulmonary aspiration, poor tolerance of infants to hypo-xaemia, lack of knowledge of pharmacology in the youngest children, and the significant medicolegal implications of dealing with this population.

Seven RCTs,18–24 13 (five retrospective, eight prospective) observational studies,25–37 and one systematic review38 were identified that examined different sedation or anaesthesia regimens (Supplementary Tables S2 and S3, https://links.lww.com/EJA/A891). The age of patients included in the RCTs ranged from 26 weeks to 3 yr, and included a total of 486 patients. The age of patients in the observational studies was from 24 weeks to 7 yr, and studied 10 759 intubation attempts. One observational study was interrupted prema-turely,28 thus five retrospective studies27,31,35–37 and seven observational studies25,26,29,30,32–34 remained for analysis. Very heterogenous drug regimens were used in neonates and infants. Some studies assessed the effects of opioids during intubation, the differences between inhalation and i.v. anaes-thesia,24,33,37 the optimal dose for propofol,29,38 the use of neuromuscular blocking agents,21–23,31,34–36 and a variety of drugs used for premedication.25–27,30,32 All studies assessed orotracheal intubation except one RCT and two observational studies that recruited patients for nasotracheal intubation.26,36 Data on emergent intubation were only available in two observational studies.25,26 Finally, only one observational study compared orotracheal intubation in a group that maintained spontaneous ventilation and in a group with controlled venti-lation.31 To determine the best preparation for neonates and infants for undergoing orotracheal intubation, we extracted data on first attempt success, number of attempts, and adverse events reported in each group. Many studies did not report adverse events (e.g. hypotension, hypertension, hypercapnia, hypoxaemia, upper airway trauma, myoclonus, hypothermia, or laryngospasm), nor the occurrence of severe adverse events (e.g. death, tonic-clonic seizures, pneumothorax, sepsis, digestive tract perforation, pulmonary haemorrhage, cardiac arrest, supraventricular tachycardia, pulmonary hypertension, aspiration syndrome, or hyponatraemia). We did not find evidence of an increase in the total number of adverse events related to use of sedative or general anaesthetic drugs.

Several studies compared a group without sedation or anaesthesia with an anaesthetised group. Compared with no sedation or anaesthesia, anaesthesia increased the success rate of intubation on the first attempt, and reduced the number of attempts and incidence of complications. Some studies compared several anaesthesia regimens, in some cases up to 13 different regimens of sedative drugs30 or the combination of more than two drugs in the same group. Anticholinergic drugs were used in the majority of studies before induction of anaesthesia, with atropine the most frequently used agent. Use of an anticholinergic drug was not associated with a lower incidence of bradycardia. Based on the available evidence, neonates and infants receiving opioids will have lower noci-ception as measured with the premature infant pain profile than those who did not receive opioids.20 Use of a neuro-muscular blocking agent was found to improve the quality of intubation conditions and to decrease the median number of orotracheal intubation attempts. Two RCTs that used rocuro-nium reported a higher rate of successful first attempt tracheal intubation, even at doses as low as 0.2 mg kg−1 in one RCT.22,23 Atracurium was used in one underpowered RCT that failed to detect a clinically relevant difference.21 Suxamethonium was addressed in two observational studies.27,32 In summary, use of a neuromuscular blocking agent increases the success of tracheal intubation and reduces the incidence of complications such as laryngospasm.

PICO 3. Tracheal intubation Is direct laryngoscopy or videolaryngoscopy the first-choice technique for tracheal intubation in neonates and infants?

Recommendation: We recommend the use of a video-laryngoscope with an age-adapted standard blade (Macintosh or Miller) as first choice for tracheal intubation of neonates and infants (1B), including for tracheal intubation in the lateral position (1C).

Clinical practice statement: a videolaryngoscope should also be used for teaching purposes using a dual approach: direct laryngoscopy for the trainee and videolaryngoscopy for the tutor. The screen can serve as guide for feedback during the intubation manoeuvre performed by the trainee.

Clinical practice statement: Training is a mandatory and essential prerequisite for correct use of a videolaryngos-cope. The use of a videolaryngoscope is warranted in anaesthesia suites, intensive care units, and emergency departments.

Evidence summary: Five RCTs comparing videolaryngo-scopy (VL) with direct laryngoscopy (DL) were assessed (Supplementary Table S4, https://links.lww.com/EJA/A891).8,39–41 The first is an RCT that enrolled 564 infants and has indicated that videolaryngoscopy with a standard Miller blade was superior to direct laryngos-copy. Infants <6.5 kg in weight had greater first-attempt success rate for orotracheal intubation when videolaryngoscopy was used compared with direct laryngoscopy (92% vs. 81%).39 Another RCT among neonatology residents compared video-laryngoscopy with direct laryngoscopy for intubation of premature neonates.40 Overall intubation success rate was greater in the videolaryngoscopy group (57% vs. 33%). First-year residents and all residents intubating their first patient had higher intubation success rates with videolaryngoscopy than with direct laryngoscopy (58% vs. 23% and 50% vs. 17%, respectively). Among residents with <6 months’ tertiary neonatal experience, when the instructor was able to view the video-laryngoscope screen success rate was 66% (69 of 104) compared with 41% (42 of 102) when the screen was covered (odds ratio [OR] 2.81, 95% confidence interval [CI] 1.54–5.17).41 When pre-medication was used, the success rate in the intervention group (videolaryngoscopy) was 72% (56 of 78) compared with 44% (35 of 79) in the control group (direct laryngoscopy) (OR 3.2, 95% CI 1.6–6.6). A multicentre RCT compared video-laryngoscopy with direct laryngoscopy in neonates and infants with apnoeic oxygenation added in both groups. First attempt tracheal intubation success rate with no desaturation was greater with videolaryngoscopy (89% [108 of 121]; 95% CI 83.7–94.8%) compared with direct laryngoscopy (79% [97 of 123]; 95% CI 71.6–86.1%), with an adjusted absolute risk difference of 9.5% (CI O.8–18.1%).6

Two RCTs compared videolaryngoscopy with direct laryn-goscopy for tracheal intubation with the patient in the lateral position.42,43 The videolaryngoscopy group had a higher first-attempt success rate and provided a more favourable glottic view compared with conventional direct laryngoscopy with a Miller blade. These studies used the C-MAC (Karl Storz, Tut-tlingen, Germany) videolaryngoscope in comparison to direct laryngoscope for tracheal intubation. One single-centre RCT (Supplementary Table S5, https://links.lww.com/EJA/A891) enrolled 120 children, 1–24 months of age, undergoing elective surgery, and showed that there was no difference between direct laryngoscopy with a Miller or Macintosh blade for laryngoscopic views and intubation conditions.44

Should apnoeic oxygenation become standard of care during tracheal intubation?

Recommendation: We recommend the use of apnoeic oxygenation during tracheal intubation in neonates (1B).

Clinical practice statement: In infants, the of apnoeic oxygenation (low or high flow) during tracheal intubation should be based on the risk of hypoxaemia in the patient and the experience of the provider.

Evidence summary: Five RCTs have shown that supplemental oxygen increases the safe apnoea time in neonates and infants and reduces the incidence of hypoxaemia during tracheal intubation (Supplementary Table S6, https://links.lww.com/EJA/A891).45–49 Supplemental oxygen can be delivered directly via the airway instrumentation device AirTraq (Prodol Meditec S.A., Vizcaya, Spain), Truview (Truphatek International Ltd, Netanya, Israel), Oxiport (Truphatek International®, Israel), nasopharyngeal airway or via nasal cannula.

One RCT which studied use of high-flow nasal oxygenation (HFNO) before orotracheal intubation in 251 patients (median age 28 weeks)46 showed a success rate for orotracheal intubation on the first attempt of 50% (62 of 124 patients) when HFNO was used and 32% (40 of 127 patients) when standard care (morphine and midazolam, without HFNO) was used. An observational study found that the incidence of desaturation during attempts at tracheal intubation in paediatric patients was significantly lower when 5 L min−1 oxygen was provided via a nasal cannula.50 One RCT comparing the efficacy of vid-eolaryngoscopy and direct laryngoscopy in neonates and infants showed that the incidence of desaturation during intubation was low in both groups.6 Even though the study was designed to compare direct laryngoscopy vs. video-laryngoscopy, the authors stated that the low incidence of adverse events could be partially attributed to supplementary oxygen administration prolonging the safe apnoea time.6,50 The optimal oxygen flow remains to be determined; it should be reported in L kg−1 min−1 to assess the effectiveness and permit comparisons.6,50

There are sparse data for anticipated difficult airways. We found only a single case report in a preterm neonate with multiple airway abnormalities and repeated attempts at tracheal intubation.51

Cuffed or uncuffed tracheal tube as standard of care?

Recommendation: Cuffed and uncuffed tubes can both be safely used (cuffed tubes in children >3 kg) (1C).

Clinical practice statement: For the safe use of cuffed tubes, we recommend adherence to the manufacturer's instructions, including size and cuff inflation pressure (minimal cuff pressure to avoid air leak, not exceeding 20–30 cm H2O), to reduce the risk of postextubation stridor. Anatomical variation, clinical conditions, and degree of prematurity might warrant the use of an uncuffed tube.

Evidence summary: Cuffed tracheal tubes are increasingly used in neonates, infants, and preschool children, with RCTs showing fewer tube exchanges for selecting the best size without affecting postextubation complications (Supplementary Table S7, https://links.lww.com/EJA/A891).52–54 The effect of cuffed tubes on the incidence of postintubation stridor or complications in neo-nates remains unclear because of a lack of sufficient data.54–56

Retrospective studies and case series indicate the need to be vigilant regarding cuff inflation pressure limits.57–59 There are insufficient data to support the routine use of a cuffed tracheal tube in children <3 kg with external tube diameter being the limiting factor.58

What is unanticipated difficult intubation in neonates and infants?

Suggestion: We suggest defining unanticipated difficult intubation as: ‘two failed tracheal intubation attempts’ to facilitate (i) comparison between studies and (ii) assessment of the effectiveness of interventions (2C).

Evidence summary: A standardised definition of difficult intubation in neonates and infants is required, especially when difficult intubation is unanticipated by medical history and physical assessment. Four large prospective observational databases were considered, all of which define difficult tracheal intubation as a minimum of two failed attempts (Supplementary Table S8, https://links.lww.com/EJA/A891).3,4,60,61 This definition will allow standardised reporting of future studies.

Nasal or oral route for tracheal intubation in neonates and infants?

Clinical practice statement: the nasal route is often preferred for success rate in neonates, and the oral route for infants, but limited data are available to recommend the nasal or oral route. When the nasal route is chosen, the risk of bleeding and nose preparation with topical vasoconstrictors should be considered.

Evidence summary: Limited data are available to recommend the preferred intubation route (Supplementary Table S9, https://links.lww.com/EJA/A891). Nasal intubation is preferred in some institutions for ease in securing the tracheal tube, especially when prolonged intubation is expected.62

Can a supraglottic airway device be an alternative to a tracheal tube in difficult airway or emergency situations?

Recommendation: We recommend use of a supraglottic airway device for rescue oxygenation and ventilation when tracheal intubation has failed or if face mask ventilation is inadequate (1B).

Evidence summary: All studies are single-centre RCTs (Supplementary Table S10, https://links.lww.com/EJA/A891).63–66 A supraglottis airway device permits oxygenation and ventilation in neonates and infants either during elective surgery or resuscitation, and can be used as an alternative to either a tracheal tube or face mask ventilation. The incidence of major adverse respiratory events during elective surgery was significantly higher for use of a tracheal tube than for use of a supraglottic airway device, with a risk ratio of 5.3 (95% CI 1.6–17.4).63 Two RCTs in neonates aged ≥34 weeks reported that supraglottic airway devices were more effective than face mask for oxygenation during cardiopulmonary resuscitation, potentially avoiding the need for tracheal intubation.64,65 One multicentre RCT reported fewer attempts and faster placement of a supraglottic airway device compared with tracheal intubation (32 vs. 66 s).66

Is a neonatal and infant tracheal intubation algorithm (cognitive aid) necessary?

Recommendation: We recommend development of a multi-disciplinary consensus based tracheal intubation cognitive aid for neonates and infants to harmonise clinical practices and potentially reduce tracheal intubation related morbidity and mortality and to enable assessment of long term outcomes (1C).

Evidence summary: Two prospective observational multi-centre studies (NECTARINE and NEAR4NEOS) reported a high incidence of difficult tracheal intubation and adverse tracheal intubation-associated events in neonates and infants (Supplementary Table S11, https://links.lww.com/EJA/A891).3,67 The high variability of clinical approaches could have contributed to the high incidence of adverse events. The NECTARINE trial reported frequent use of neuromuscular blocking agents (72.6%) and limited use of videolaryngoscopy.3 Practices independently associated with reduced adverse events in the NEAR4NEOS registry included videolaryngoscopy (OR 0.46, 95% CI 0.28–0.73) and neuro-muscular block (OR 0.38, 95% CI 0.25–0.57).67 A prospective multicentre trial to investigate the benefits of neuromuscular blocking agents was abandoned early because of a lack of funding causing it to be underpowered.21

PICO 4. Difficult airway management What should be the gold standard for anticipated difficult airway management and who should be involved?

Clinical practice statement: For anticipated difficult airway management, at least one type of videolaryngoscope, flexible intubating scope, or rigid or semirigid scope should be available, including appropriate sizes for the patient's age, in addition to routinely used equipment such as supraglottic airway devices and face masks.

Evidence summary: There is no ‘one solution fits all’ for difficult airway management for neonates and infants. In any medical care area where infants might need respiratory support, established equipment for bag-mask ventilation, for opening of the upper airway (such as oral or nasal airways), supraglottic airway devices, suction tubes, laryngoscopes, and tracheal tubes in appropriate sizes should be readily available. Every instrument for instrumentation of the difficult airway has limitations in certain situations. As the source of airway difficulties varies individually, individuals’ airway restrictions inevitably demand one of the various types of airway devices.

Recommendation: We recommend limiting the number of tracheal intubation attempts by reassessing the clinical condition and by considering a change to a different technique, different provider, or both after each attempt (1C).

Clinical practice statement: after four attempts, clinicians should consider aborting intubation attempts and waking the patient if feasible.

Evidence summary: Multiple attempts at tracheal intubation with the same technique result in a higher probability of complications and can cause airway oedema or bleeding. This reduces the chance of success of subsequent attempts with the same or other techniques (Supplementary Table S12, https://links.lww.com/EJA/A891). A single centre retrospective analysis of 1341 intubations in healthy infants for routine operative procedures found an increased risk for hypoxaemia (OR 1.78, 95% CI 1.30–2.43) when multiple intubation attempts occurred.68 In a prospective observational single centre trial including 171 intubations on neonatal intensive care wards, occurrence of more than two intubations was identified as the only independent risk factor (OR 6.7, 95% CI 1.3–33.6) for adverse outcomes (brady-cardia, hypotension, or hypertension).69 Comparable results have been reported in a retrospective analysis of an international multicentre database for neonatal intubations because of respiratory failure.70 Two attempts had an OR of 1.6 and three or more attempts an OR of 1.8 for severe adverse events. Thus, after an initial failed attempt, assistance by additional personnel with special airway expertise and supportive devices must be consulted immediately.

Clinical practice statement: There is insufficient evidence to recommend which patients should be intubated with hyperangulated blades. However, in cases where standard blades fail and the airway is difficult (anterior larynx, suspected cervical spine injury, or limited movement), the next step should be an alternative advanced technique including use of hyperangulated blades with a stylet, flexible or rigid bronchoscopy alone or in combination with video-laryngoscopy, or flexible bronchoscopy via a supraglottic airway device.

Evidence summary: When conventional direct laryngos-copy fails to provide a sufficient view of the glottis for tracheal intubation, videolaryngoscopy is commonly used as the first advanced technique.

Hyperangulated (non-standard) videolaryngoscopes are often superior in visualising the glottis when conventionally shaped (standard) blades fail.71–73

There is a difference in intubation success rate in infants and neonates when different types of videolaryngoscopes are used. The success rate for intubation using hyperangulated videolaryngoscopes is lower than with Miller blade video-laryngoscopes. In conventional direct laryngoscopy, a good view of the glottis is usually associated with easy intubation. This might not be the case when performing indirect video-laryngoscopy, particularly when using a hyperangulated blade.74–77 Because videolaryngoscopy provides a full view of the glottis without aligning the oral and the tracheal axis, passage of a tracheal tube can be more difficult. This is particularly true if a hyperangulated blade is used, resulting in greater misalignment of the oral and tracheal axes and a longer and more angulated route for the tracheal tube to pass into the trachea. This makes use of a stylet mandatory when a hyperangulated videolaryngoscope is used. This can result in more time needed for tracheal intubation and lower overall success rates, especially in small children.71,72,78–84 Modification of the shape of a stylet by adding an angle of 15 degrees has been described when using a hyperangulated blade.85,86

Based on this evidence, other alternative techniques that have been studied in this population for tracheal intubation when direct laryngoscopy is difficult include flexible bron-choscopy, intubation through a supraglottic airway device, or combined use of videolaryngoscopy with flexible bronchoscopy. Data from the Pediatric Difficult Intubation Registry suggest that these advanced techniques have similar success rates, and it is not possible to recommend any single technique. The choice of intubation technique will depend on the equipment available, the experience of the clinician managing the patient, and the airway anatomy of the patient. It is important that practitioners not persist with failing techniques, and that an alternative be used early in the process, as severe complications associated with airway management occur with repeated attempts at laryngoscopy and intubation.

Recommendation: We recommend use of a stylet to reinforce and preshape tracheal tubes when a hyperangulated laryngoscope blade is used or when the larynx is anatomically anterior (1C).

Clinical practice statement: Routine use of a stylet to improve the success rate of tracheal intubation by novice practitioners and trainees cannot be recommended when performing laryngoscopy with standard blades. Bougies of appropriate size can be used to facilitate a difficult intubation or to guide tube insertion.

Evidence summary: Using a stylet during laryngoscopy with standard blades to reinforce or preshape the tracheal tube does not appear to increase the success rates of tracheal intubation in neonates and infants and might be traumatic to neonatal and infant airways (Supplementary Table S13, https://links.lww.com/EJA/A891). A retrospective analysis of >5000 neonatal intubations, one RCT, and two meta-analyses (with low quality of evidence) did not identify benefits for first time success rate or complications if a stylet was used or not.55,58,87,88 Nevertheless, a stylet or bougie might be essential if reinforcement or stronger angulation of the tracheal tube is needed or when the larynx is anatomically anterior and angulation of the tracheal tube is essential to facilitate intubation.85,89

Recommendation: We recommend flexible bronchoscopy by the nasal route in case of restricted mouth opening (1C).

Clinical practice statement: Flexible fibreoptic tracheal intubation can be performed through a supraglottic airway device, a specially designed face mask or via one nostril while a nasopharyngeal tube is in place in the other nostril for oxygenation. Intubation through a special face mask can be easier, especially for trainees or novices, and when performed via the nasal route. Another provider can assist with mask ventilation during the intubation. If not using a supraglottic airway device, trainees and novices might choose the nasal route for fibreoptic intubation, unless contraindicated, under supervision of an expert physician.

Evidence summary: In most situations of difficult airway management, use of a flexible bronchoscope is a suitable solution for tracheal intubation (Supplementary Table S14, https://links.lww.com/EJA/A891). There are no alternatives if a fully visualised transnasal tracheal intubation is required or if mouth opening is severely restricted. The nasal passage is easier for less experienced trainees because of more straightforward guidance of the scope compared with the oral route.90

Suggestion: We suggest use of a rigid bronchoscope as an advanced technique when the laryngeal inlet is obstructed by swelling and in cases of upper airway stenosis or compression or in congenital or postsurgical tracheal constriction or tortuosity (2C). If necessary, a multidisciplinary team (including an otolaryngologist) should be involved.

Evidence summary: Flexible intubating fibreoptic scopes are not the single best solution for all problems with the pae-diatric airway because of its fundamental limitations: a limited field of view, interference from bleeding and secretions, and flexibility of the device as only the tip of the bron-choscope can be steered. If it is necessary to advance a bronchoscope through a narrow space, such as laryngeal stenosis or tracheal obstruction, use of a rigid endoscope can be advantageous. If preloaded with a tracheal tube, the rigid scope can be used to guide the tracheal tube beyond the stenosis. In cases of long tracheal stenosis (for example extra-luminal compression by a mediastinal mass), placement of a rigid bronchoscope might be required. This enables navigation through the narrow region while splinting the airway open, and at the same time allowing ventilation via a side port.78,91–95 The main limitation of rigid and semirigid scopes is possible difficulty in aligning the oral and tracheal axes, for example in cases of severely restricted mouth opening or severely restricted retroflexion of the head.96–98 A trial including 26 children (12 of them infants) with difficult airway reported quicker intubation with a semirigid scope than with a fibreoptic scope (52 vs. 83 s) (Supplementary Table S15, https://links.lww.com/EJA/A891).97

Clinical practice statement: After induction of general anaesthesia, when tracheal intubation fails, oxygenation and ventilation via a supraglottic airway device or face mask are severely impaired or impossible, and spontaneous breathing cannot be restored, a surgical tracheotomy should be performed. Of several techniques described, evidence is lacking for superiority of one technique over another.

Surgical cricothyroidotomy and percutaneous needle crico-thyroidotomy are not suitable options in neonates and infants; for the former because the small size of the cricothyroid membrane will likely render insertion of a tracheal tube impossible, and for the latter because of the unfavourable anatomy.

Clinical practice statement: When the expertise and equipment are available, extracorporeal membrane oxygenation (ECMO) can be considered as a rescue intervention for a difficult airway when waking the patient is not an option. However, there is lack of evidence supporting such a recommendation, and the decision and timing to proceed with ECMO is left to local guidelines.

Evidence summary: Current literature does not provide clear guidance on how and when to perform emergency front-of-neck access to the trachea as a life saving intervention in neonates and small children. The small size of the cricothy-roid membrane is not compatible with surgical or percutaneous cricothyrotomy in neonates and infants and the compressibility of the trachea makes a percutaneous approach relatively impractical. Therefore, a tracheostomy is preferable.99 Even with maximal extension of the head and neck, the combination of the high location of the larynx in the neck and the relationship between the mandible and the trachea forces percutaneous access to the airway to be approached at a very steep angle, resulting in tracheal compression and the risk of posterior wall perforation with subsequent problems leading to overall failure. A single pae-diatric case report described the need to rescue the airway with surgical access after unsuccessful percutaneous crico-thyroidotomy.100 Based on little evidence, we propose that a surgical tracheostomy represents the preferred emergency access to the trachea in neonates and infants.101,102 Access to the trachea is provided with a tracheostomy tube or a standard cuffed tracheal tube of appropriate size for age. This procedure should be carried out by the most competent physician available. A coordinated multidisciplinary team approach should be considered. Local airway leaders should prepare a standard operating procedure based on their specific expertise and availability of other involved disciplines in their institutions.

Clinical practice statement: No evidence is available supporting or refuting management of difficult airways in a particu