The airway assessment has come of age—or has it?

Goodness, the airway assessment is a complex beast. Most of us think we can do one from 10 m away in a train station, and when we do, it is usually with a sense of relief that the subject of interest is getting on a train, not entering one's operating theatre. A more challenging pastime is to reflect on the actual components of a complete airway assessment. What in the history or physical examination should we be assessing, or screening, if you will? And even if the screening test works reasonably well to predict difficulty (if that is the outcome of interest), how should we then alter our management of the airway? In an ideal world, the ‘unanticipated difficult airway’ would never occur. Bed-side screening would be 100% sensitive in its ability to predict difficulty with all airway interventions while also being very specific in predicting when there will be no difficulty. Unfortunately, we are not there yet. According to many studies and now meta-analyses of those studies, we are not even close. But before casting the airway assessment on the scrap heap, we must ask if we are being unfair to it. Are we screening the correct parameters? Are we screening for all potential threats to patient safety during the airway management process or only some? And ultimately, if published sensitivities for screening single anatomic parameters are poor, might combining many parameters (i.e. what we actually do in clinical practice) into a multifactorial screening test be better? Despite the unanswered questions, most of us continue to assess the airway, if only to pass time in train stations. For patients undergoing general anaesthesia (or who may lose consciousness as a complication of regional anaesthesia or procedural sedation), in assessing the airway, we are seeking to predict if we can manage what patients normally do for themselves: maintain gas exchange, airway patency and protection of the lower airway against aspiration of foreign material. When difficulty is predicted in maintaining these vital functions once the patient is unconscious and often apnoeic, these predictions should help inform the clinician's choice of intended primary and ‘rescue’ approaches in managing the patient's airway. Airway assessment also helps determine if any extra preparations are needed during the implementation of the intended approach and is equally necessary before extubation of the intubated patient, particularly in the patient who is at risk due to pre-existing or new predictors of a difficult re-intubation 1. Robust data are lacking on whether patient outcomes are improved by performing structured airway assessment. Notwithstanding, we have all been taught to do one and studies on airway-related morbidity indicate the hazards of not doing so (or ignoring its findings) 2. The lack of a documented airway assessment has been cited in medicolegal cases as falling beneath the standard of care 3. Perhaps most importantly, performing an airway assessment can be considered a cognitive forcing strategy to get one thinking about the possibility of encountering a difficult airway in any patient. It is reasonable to start an airway assessment by screening for predictors of difficulty or failure with one's intended technique – most often, tracheal intubation or supraglottic airway device (SAD) use. Thus, if one's initial plan is tracheal intubation facilitated by direct laryngoscopy, screening the patient for anatomic predictors of difficult direct laryngoscopy is logical. Fortunately, due to publications such as the meta-analysis in this issue of Anaesthesia by Roth et al. 4, we have a good idea of which anatomic predictors of difficult direct laryngoscopy, difficult tracheal intubation and difficult facemask ventilation are typically assessed and how well they work to predict the outcome of interest. Many of the studies included in this meta-analysis sought to determine the sensitivity of screening anatomic features in isolation, yet rarely would we examine only a single parameter in clinical practice; rather, airway assessment is multifactorial. Furthermore, we should also screen for whether an alternative device (e.g. a videolaryngoscope) might succeed if the first fails. Unfortunately, although some data exist 5-7, there is less published literature on the predictors of difficulty with tracheal intubation facilitated by videolaryngoscopy. If a supraglottic airway device (SAD) is the intended technique, screening should begin by seeking predictors of difficulty or failure using these devices. Again, with some exceptions 8-11, there are few published studies that address anatomic predictors of difficult SAD use. Common sense can be applied to an extent (e.g. mouth opening is required), but as with video laryngoscopy, more study is needed. The airway assessment should also screen for difficulty with pathways other than the intended technique. Thus, for a case using a SAD, one should also screen for difficulty with facemask ventilation and tracheal intubation. Equally, when tracheal intubation is the intended technique, one should screen for difficulty with facemask ventilation and SAD use. Although overlap exists between predictors of difficulty among pathways 12, there are also predictors unique to a particular technique (e.g. mask seal issues for facemask ventilation), which is why each should be considered separately and deliberately during airway assessment. Some espouse evaluating all patients for potentially difficult emergency front-of-neck airway access; others only when difficulty with other pathways is evident. Not all anatomic predictors of difficulty are easily visible. For the patient presenting with known or suspected obstructing upper airway pathology, the airway assessment may require enhancement by pre-operative nasendoscopy to determine the current state of the pathology 13. Similarly, as part of a pre-extubation airway assessment, the at-risk patient may require assessment of the larynx by videolaryngoscopy or flexible endoscopy for the presence of oedema. Subglottic pathology may require review of imaging studies 14. Beyond predictors of technical difficulty with airway instrumentation, airway assessment should also address whether any out-of-the-ordinary physiological threats to patient safety exist that might impact a decision on how to proceed 15. Intolerance of apnoea due to compromised functional residual capacity, increased oxygen consumption or high minute ventilation requirements are examples, as is a full stomach or haemodynamic instability. Finally, the clinical context can similarly impact decision making, such as the lack of ready access to skilled assistance, clinician inexperience or lack of the desired device due to reprocessing or repair 16. Patient history, examining the notes and old anaesthetic records can only add to the information available and should be done for all cases. Arguably, if one can shop at a multinational clothing store and have a sales clerk list one's past purchases made in a different country, then the clinician who manages airways should have access to records detailing remote episodes of airway management. Organisations such as the MedicAlert Foundation are trying to address this 17. Ultimately, then, the airway assessment is multifaceted and should extend substantially beyond simply seeking isolated anatomic predictors of technical difficulty with one's intended technique. Whether the average clinician routinely systematically evaluates all of the foregoing factors with every airway assessment is currently unknown. There is little point in performing a screening test unless prepared to act on its results, and the airway assessment is no different. Predicted difficulty based on the airway assessment might result in one or more of a number of responses. For example, an additional margin of safety might be afforded from instrumenting the airway in the awake, spontaneously-breathing patient, under topical airway anaesthesia. The safety benefit accrues from having the patient maintain gas exchange, airway patency and protection of the lower airway against aspiration of foreign material during airway instrumentation. The need for awake tracheal intubation is suggested by findings such as anticipated difficult or impossible direct laryngoscopy or videolaryngoscopy, either alone or in conjunction with other predictors of difficult airway management. The latter might include predicted difficult facemask or SAD ventilation, apnoea intolerance, a high risk of pulmonary aspiration or a lack of ready access to desired equipment or skilled help 16. Alternatively, the results of the airway assessment may suggest the need to revisit the originally planned technique. For example, a case originally planned with a SAD or regional anaesthesia might be more safely managed with controlled, tracheal intubation (awake, or after the induction of general anaesthesia, with appropriate preparation) if difficulty with airway management is predicted. Conversely, if airway assessment suggests difficulty, one could revisit whether a case originally planned under general anaesthesia could feasibly and safely be performed with a regional technique if the right conditions prevailed. Finally, although perhaps stopping short of indicating the need to move to an awake approach or to change the originally planned technique, the airway assessment might simply suggest the need for extra care during the implementation of the originally planned technique. This might include reverse Trendelenburg patient positioning, use of high-flow nasal oxygen and having extra equipment and a colleague standing by in the operating theatre. Without a doubt, the sometimes disappointing results of clinical trials of the standard airway assessment in predicting difficult or failed direct laryngoscopy-aided tracheal intubation and facemask ventilation 18 can leave one cynical about the value of airway assessment. However, such studies often exclude participation by patients already identified as overtly requiring awake tracheal intubation. For this subset, many of whom would be impossible to intubate using standard techniques (e.g. those with very limited mouth opening or a fixed flexion deformity of the neck), we submit that existing published predictors of difficulty are likely to be highly sensitive in identifying difficulty. The more problematic issue relates to the overtly normal-looking patient requiring tracheal intubation or SAD use who, despite a reassuring airway examination, presents difficulty after the induction of general anaesthesia. The so-called ‘unanticipated difficult airway’ appears to occur with distressing regularity. So, what should happen to address this problem? First, we should continue to work on developing multifactorial ‘tools’ of predictors. Many existing studies of airway assessment efficacy, such as those included in the Roth et al. 4 meta-analysis, studied how well a single anatomic feature predicted difficulty or failure of tracheal intubation or facemask ventilation in apparently normal patients. But that is not what we do in our clinical practice; rather, we examine multiple features. ‘Open your mouth.’ ‘Stick out your tongue.’ ‘Bite your upper lip with your lower teeth.’ ‘Extend your head.’ Although some studies have indeed looked at the effectiveness of examining multiple anatomic parameters in predicting difficulty 19-24, they are in the minority and not all have performed well to date. Many fall short of the sensitivities reported for analogous tools used in other specialities, such as the Ottawa ankle rules Ottawa ankle rules, used very successfully to exclude an ankle or mid-foot fracture based on clinical signs 25. Perfecting such tools for airway management will not be an easy endeavour. We will have to keep working on determining which anatomic predictors of difficulty to include, how to weight them and we will have to determine the score that serves as a trigger to alter one's management. Clinical trials and meta-analyses such as the Roth et al. study 4 will help to determine which anatomic features should populate the tool. At a minimum, tools should be developed to predict technical difficulty or failure with each of the three pathways of airway instrumentation, that is, tracheal intubation, SAD use and facemask ventilation. Physiological risk factors should perhaps be included. Secondly, we should continue to seek predictors of difficulty or failure with contemporary techniques. Many studies have addressed the predictors of difficult direct laryngoscopy; fewer have addressed difficult and failed facemask ventilation; fewer still have addressed predictors of difficult and failed SAD use or tracheal intubation facilitated by videolaryngoscopy. More study is required. Third, we must define relevant outcome measures. As an outcome measure, ‘difficulty’ might best be defined as the need for more than one attempt at the intended technique, whereas ‘failure’ of the intended technique might best be defined as not having succeeded after three attempts. Many national airway guidelines use three failed attempts at a technique as an indication to pause, reconsider options and take a different pathway 26-28 to help avoid perseveration with an unsuccessful technique. For a higher level tool seeking to predict overall hazard to the patient rather than the technical failure of a particular technique, study outcome measures might be based on the occurrence of physiological trespasses such as hypoxaemia, haemodynamic instability, aspiration of foreign material, brain damage or death 29. Fourth, newly developed tools would need to be prospectively validated in a study population different from that in which they were developed, representative of the population in whom they would be seeking to predict their outcome of interest 30. Finally, with newly validated tools to predict difficulty with airway management, we must provide guidance on how to use the information. Ultimately, the information gathered during an airway assessment should be used to determine how most safely to proceed with airway management. Unfortunately, little guidance appears in existing national airway guidelines on decision making once the airway assessment has predicted difficulty. A simplistic adage of ‘if it looks difficult, do it awake’ is not practised, practical, or in many cases, necessary. Future iterations of airway guidelines need to address such decision making, rather than only advising on the management of the already-unconscious patient. In the interim, we must continue to be prepared for difficulty with all cases. Unless or until we have good tools proven to be more effective than existing screening tests for predicting difficulty in all patients, we will have to continue to be prepared for the unanticipated difficult airway. This includes having a plan for difficulty in all patients, sharing the plan with the assembled team, having immediate availability of a good array of equipment and knowing where and when to obtain skilled help 31. In our opinion, the airway assessment has yet to come of age. Many clinical trials still focus on the effectiveness of a single anatomic predictor in predicting difficulty, and although a good a starting point, it is simply not what we do in clinical practice. There remains a paucity of information on what predicts difficulty or failure with other commonly used modalities such as SADs or videolaryngoscopy. We are only starting to recognise the need to assess physiological or contextual predictors of hazard to the patient during airway assessment. And then, even to the individual who conscientiously seeks predictors of difficulty or failure in these many facets, few national airway guidelines have yet forayed into providing meaningful guidance on how most safely to proceed when difficulty is predicted. Should we give up on formal airway assessment, though? No. We can at least teach a structured approach to airway assessment, indicate the factors that should be considered, and yes, we should continue to study whether the combination of some or all of these factors into multifactorial tools will more successfully predict difficulty with airway management. Meanwhile, enjoy airway spotting at the train station.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.