The view from the top. Is it worth recording for posterity?

O'Loughlin et al.'s article in the July issue of Anaesthesia 1 raises some issues about the way we consider visualisation and documentation of laryngeal anatomy in relation to tracheal intubation. Articles on airway devices often refer to the laryngeal view that can be obtained as an indicator of intubation difficulty, and for direct laryngoscopy this is a reasonable association to make. The Cormack and Lehane method of scoring laryngeal view 2 is probably the most commonly used in anaesthetic practice, and modifications of the original score from Yentis 3 and Cook 4 have added extra sub-scores in an effort to increase the precision of the information collected. The score may be used in studies to compare the view obtained under different circumstances or using different devices, and it may also be recorded on the anaesthetic chart as an indicator of future laryngeal views. However, it is not ideal as a tool for airway assessment as there may be substantial variation in interpretation of what is seen and what score is recorded 5-8. As well as inter-rater reliability, there is the question of accuracy of the scoring, that is, the proportion of assessments that were correct. However, in clinical practice, what is correct? The score is based on laryngeal view, but it is not always appreciated that the Cormack and Lehane score is based on the best view that could have been obtained. This, in turn, depends on adequate preparation for laryngoscopy, such as optimal head positioning and experienced assistance. There is also a presumption of a reasonable level of individual skill from the operator, but expertise in laryngoscopy and videolaryngoscopy may be difficult to define 9. These are variables which may differ between laryngoscopy attempts, and sub-optimal conditions will also result in an inaccurate score. A standard of accuracy can, therefore, be defined but may not always be achieved. The O'Loughlin et al. study compared the accuracy and inter-rater reliability of different scoring systems during videolaryngoscopy: the Cormack and Lehane 2; percentage of glottic opening (POGO) 10; and Freemantle 11 scoring systems. They side-stepped the skills element of obtaining a laryngeal view by asking participants to assess video recordings of laryngoscopy in patients, where the view for each score had been classified by a consensus of independent experts to establish the standard against which accuracy was then judged. Accuracy and inter-rater reliability were best with the POGO and Fremantle scoring systems, while the Cormack and Lehane score showed significantly lower accuracy and lower inter-rater reliability. At first sight it might appear that the POGO and Fremantle scores are superior assessment tools for scoring laryngeal view. However, there are some caveats for the interpretation of these results. Scoring systems for laryngeal view have been designed for different purposes, so a direct comparison is difficult and there are inevitably some compromises that have to be made. Cormack and Lehane is a descriptive system in which scores are applied to laryngeal views, resulting in ordinal ratings of 2b, 3 etc. The POGO simply describes the percentage of visualised glottic opening, whereas the Fremantle has two groupings of ordinal scores, one based on glottic view, and the other on ease of tracheal intubation. It also includes a description of the device used, to allow for variation between devices so that the view can be further contextualised. Therefore, using accuracy and inter-rater reliability to compare scoring systems is not as straightforward as it seems. These measures may also be affected by the number of elements in the score itself, and a system with a smaller number of scores may be expected to produce better accuracy and inter-rater reliability ratings than a system that uses a larger number of scores. There are three visualisation elements and three intubation elements in the Fremantle scoring system, four in the original Cormack and Lehane (but six with the extended Yentis and Cook sub-classifications), and potentially 100 percentage elements in the POGO system! In the O'Loughlin et al. study, a decision was made to define agreement for accuracy or reliability as the same score with the Fremantle and Cormack and Lehane systems, but to within 15% for the POGO to effectively reduce the number of scores with this system. These decisions appear reasonable as it seems unlikely that observers would be able to achieve finer discrimination than 15%, but such decisions have the potential to change the resulting accuracy and reliability results that were used for comparison. The main reason for documenting laryngeal view is to indicate to the next operator when there has been difficulty obtaining a view, thus making tracheal intubation more difficult. The assumption underlying the interpretation of the Cormack and Lehane score is that it also correlates with ease of tracheal intubation. Benumof defined difficult intubation and failed intubation as Cormack and Lehane 3 and 4 views with repeated intubation attempts 12. Also, as previously mentioned, the Cormack and Lehane scoring system has been extensively used as a proxy measurement for difficult tracheal intubation in many studies. It seems obvious that using direct laryngoscopy, difficult laryngeal view will make tracheal intubation more difficult. However, there may be other factors to consider such as operator skill, intubation aids, and sub-glottic factors that complicate tube placement and, for placement of endobronchial tubes, laryngeal view may be considered to be only one component of intubation difficulty 13. With videolaryngoscopy, things get even more complicated. Laryngeal visualisation seems to be easier when direct laryngoscopy is difficult 14, 15. Poor laryngeal view under direct laryngoscopy may be a predictor for difficult intubation during videolaryngoscopy 16, but a good laryngeal view at videolaryngoscopy does not necessarily correlate with ease of tracheal intubation 17. As O'Loughlin et al. point out, with videolaryngoscopy we have now introduced the new concept of ‘good view, can't intubate’. We are not quite in the situation where videolaryngoscopes have made laryngeal view redundant as a predictor of difficult intubation, but this disconnect between laryngeal view and ease of intubation creates a situation where we can no longer rely on scores that are only based on what is seen at laryngoscopy to predict intubation difficulty where indirect laryngeal visualisation techniques are concerned. Unlike the Cormack and Lehane and POGO systems, the Fremantle system was designed for use with videolaryngoscopes, and it includes a score for intubation difficulty which is, essentially, a retrospective score of intubation difficulty following videolaryngoscopy. However, the O'Loughlin et al. study required a judgement on the likelihood of intubation by examining a laryngeal view and ignoring unexpected intubation difficulties. This is a serious problem, because it assumes a strong correlation between view and difficulty with intubation, which is not necessarily the case with videolaryngoscopy. As far as we know, all the recorded intubations in this study were successful. The inclusion of failed intubations would have made a comparison of scoring systems more relevant if we consider that the main purpose for any scoring system is to help predict future intubation difficulty. There is also the problem of using a single score with different videolaryngoscopes. There may be difficulty in manipulating the tube into the larynx using an indirect view, and this may also vary depending on what type of device was used 18. Laryngeal view and intubation difficulty may vary according to whether a device is used to intubate directly or indirectly 19. Videolaryngoscopes that pre-load the tracheal tube may have an advantage over videolaryngoscopes without preloading, enabling easier tracheal intubation for a given view 20, 21. It is also known that laryngeal view varies substantially with different videolaryngoscopes 20. A C-MAC videolaryngoscope (Karl Storz, Tuttlingen, Germany) was used in this study, and this could limit the generalisability of the results. As previously discussed, there may also be unexpected sub-glottic factors that affect intubation difficulty, and the ‘intubation with aids’ score element in the Fremantle system may introduce another variable component of skills that needs to be considered when predicting difficulty of intubation by another operator. To some extent, the importance of laryngeal view has also been reduced by the increasing popularity of supraglottic airway devices. If the trachea needs to be intubated, there is the potential of passing the tracheal tube through the device, blindly or under direct vision using a fibreoptic bronchoscope. None rely on view at laryngoscopy, so it could be argued that laryngeal view is becoming increasingly irrelevant as a predictor of airway management difficulty as our use of and expertise in alternative airway equipment increases. Of course, supraglottic airways cannot be used in all cases, and there are many cases where videolaryngoscopy is difficult because of pathology or anatomical variants 22 and alternative strategies need to be considered. Direct laryngoscopy remains an important and, in many cases, vital technique and skill in anaesthetic practice. So, is laryngeal view still useful to record? If we want to continue using the Cormack and Lehane scoring system to record laryngeal view, we should be more aware of its limitations as an assessment tool and predictor of difficult intubation. It may be argued that it has never been adequately validated 6 and O'Loughlin's study confirms that there are problems with inter-rater reliability. This is probably not an inherent feature of scoring systems based on laryngeal visualisation; the POGO system (which depends on laryngeal view rather than anticipated intubation success) had similar inter-rater reliability to the Fremantle system. If the purpose of a score is to indicate previous laryngeal view and possible intubation difficulty when videolaryngoscopy is being utilised, then the Fremantle may be a better system. Given the numerous problems in predicting future intubation difficulty from a previous laryngoscopic view, why not just record intubation difficulty? If a scoring system that only quantifies intubation difficulty is required, the Intubation Difficulty Scale 23 may be worthy of consideration. There are a plethora of scoring systems and management options. Whatever scoring system is used, it is obviously important to contextualise it by recording equipment used, technique, adjuncts, difficulties with obtaining laryngeal view etc. Following laryngoscopy using whatever method, it is important to record information to help the next operator. Recording information that is inaccurate and unreliable is potentially dangerous, and recording a score that represents a single uncontextualised laryngeal view seems an increasingly inadequate approach. BJ is an Editor of Anaesthesia. No external funding or other competing interests declared.