A guest post by George Kovacs and John Adam Law

These past couple of months will be remembered as significant for many reasons. (No, this is not an election commentary). To me (GK), this period was marked by three important headlines: Bob Dylan’s Nobel prize, Leonard Cohen’s death and Lewis et al.’s Cochrane review of videolaryngoscopy (VL) vs. direct laryngoscopy (DL).¹ Dylan and Cohen both share a warm place in my heart and have touched generations with their brilliance.

The subject matter of the Cochrane Review evoked a very different warm sensation for me – rather more visceral. Much as I’d love to write a song and reach a wide audience, I’ve been told I have an awful voice and no talent. News of the Lewis publication soon appeared in the social media world and interestingly, comments claimed validation of both sides of the divide between VL and DL proponents. Hallelujah (rejoice), it’s been settled…or has it?  To borrow line from Cohen’s song, “Your faith was strong but you needed proof”.  My idea to try and write an editorial on the subject of VL vs. DL (again) kept me up for almost 48 hours while I was binge listening in celebrating Dylan and mourning Cohen. Had I been writing by hand, my garbage basket would have been overflowing with crumpled-up drafts. Apparently, it took Cohen 2 years to write ‘Hallelujah’ and while I am in no way trying to compare my task to his brilliance, that’s roughly the time required from idea to publication for most peer-reviewed submissions. We did feel it important to present our views (mine and those of Adam Law)  in a more timely commentary as a ‘secondary literature’ submission (see Schriger’s editorial “Does everything need to be scientific?”) ² we are grateful for this opportunity.

The debate of VL versus DL is not a new one and most will remember several years ago when respected voices were making strong statements about the impending irrelevance of DL as they hailed a new standard intubation device, the video laryngoscope. Overwhelming supportive evidence was claimed.  My response to these claims marked my entry into an unfamiliar world of social media and FOAM. I posted a response on a blog site that was filled with strong commentaries on both sides of the debate. I subsequently sought a more traditional venue to provide my evidence-based opinion in the primary (peer reviewed) literature and I was fortunate enough to be granted that opportunity in an editorial in the Canadian journal of Emergency Medicine (CJEM) entitled Airway management: ‘‘the times they are a-changin’’.³

So, can we can differentiate proof from faith?  While peer-reviewed publications remain a critical recourse to help answer such questions and guide care, the volume and quality of data can be both under- and overwhelming. The impact and face validity of a Cochrane Review publication carries significant weight for many, us included. While I can’t say that I’ve agreed with the conclusions of every review (e.g.,. sux vs. roc⁴) it certainly causes most to take notice.  Without doubt, this particular review took a whole lot of work and represents an important contribution to the  airway management literature.

Lewis SR, Butler AR, Parker J, Cook TM, Smith AF. Videolaryngoscopy versus direct laryngoscopy for adult patients requiring tracheal intubation. Cochrane Database Syst Rev. 2016 Nov 15;11:CD011136. [Epub ahead of print] Review. PubMed PMID: 27844477

Here are the results of the primary outcomes of this review:

1. Fewer failed intubations occurred when VL was used;
2. There is no evidence of a difference in the occurrence of hypoxemia between use of VL and DL.

Here are selected results of the secondary outcomes of the review:

1. There is no evidence of improved first-attempt success with VL, compared with DL;
2. VL improves the glottic view and may reduce laryngeal/airway trauma.
3. Currently, no evidence indicates that use of VL reduces the number of intubation attempts.
4. Subgroup analyses of 4 different VL designs showed no difference between DL and the GlideScope, Pentax or McGrath (type not specified) for the outcome of failed intubation. There was a difference for the CMAC (type not specified, but presumably used with Mac blade);
5. Compared with DL, subgroup analysis showed a fewer failed intubations when VL was used for predicted difficult airway situations, but no difference when no predicted airway was reported, compared with DL;
6. There was no difference in failed intubation between VL and DL when operators were inexperienced (< 20 uses) with VL.

After posting some of the study’s outcomes on social media, I received several communications via text, email and twitter from individuals and colleagues who understandably were concerned that this was a nail in the coffin for DL, based on the  findings that statistically significantly fewer failed intubations occurred with VL.  I’ll return to this, but first, let’s deal with another study finding: that VL provides a better view than DL. Having a room with a view is great and VL capitalizes on this uncontestable fact that as an “around the corner” visualization device it will almost always produce a better view compared to direct laryngoscopy. On the face of it, by extrapolation, this should imply that (as with DL) having a good view should mean a slam dunk intubation. This new big screen glottis seemed a game changer.  But, close to 10 years later we’re still problem-solving how to get the damned tube down the hole that everyone in the room can see. ⁵(Levitan ACEP).  More about this below.

Over this same period more and more devices have appeared, all producing those seductive yet often illusive views. Our tool boxes are being filled with tools and to make room, we have been willing to sacrifice older “irrelevant technologies” such as DL. Study headlines screaming VL superiority have seemed to out-do any evidence that might suggest a more cautionary interpretation of the literature. However, what is becoming clear but generally unappreciated in the discussions of VL vs DL is that the reported superiority of VL may have more to do with the lowering of our benchmark for DL success rates, rather than any other factor.³ A recent systematic review by Park et al. reaffirms that the benchmark for first attempt success (FAS) in emergency airway management is around 84% a number consistent with historical registry data for FAS by DL.^{6 7 8} Videolaryngoscopy FAS rates in many of these studies were less than this benchmark but even more concerning has been the trend of very substandard FAS rates for DL (in the 60% or less range). All of these issues have been previously described in the CJEM editorial³ and commented on AIMEairway.ca and one of my backyard Vlogs (Safety and the Laryngoscopy Debate).

Figure 1A and 1B: Opportunities to compare various devices on manikins and in clinical cadavers

With the luxury of having access to many devices (figure 1A.) and clinical-grade cadavers (figure 1B. lifelike tissue responsiveness), at our place, we have the unique opportunity to perform many hundreds of risk-free intubations on human tissue. By comparing device performance on manikins with that in human cadavers, we have been able to evaluate how skills in using them attained on manikins transfer to real human tissue.

Figure 2: Assessing device performance and problem solving strategies using clinical cadavers as assessed using fluoroscopy.

Our cumulative conclusion has become our teaching mantra that the major determinant of success and failure is not the device but rather the hands (skill) and head (decision making) that guide the device’s use.  Each device is different and will have advantages and disadvantages but no one product can fill a void created by an unskilled user. We need a tool box customized to our skill set and built with some good tools, but not excessive redundancy. We shouldn’t give up a reasonable backup device because we have a new primary approach. It’s not plan A or B or C, it’s A and B and C. (for more see: VL or DL Doesn’t Matter- It’s about having a Plan A,B,C!)

The use of video-enhanced Macintosh blade laryngoscopy helps address patient safety when learners are involved in high acuity cases and is an incredibly valuable educational tool, allowing the teacher and learner to share the clinical experience and create an Ericssonian environment of safe, deliberate practice with real time feedback.

Figure 3: Learner performing direct laryngoscopy using a Macintosh VL (Titanium Mac) initially without view of the screen

So, let’s get back to the conclusions from the Cochrane review which used ‘failed intubation’ as its primary outcome.  While the pro VL readers may be satisfied by its major conclusion of a lessened incidence of failed intubation with VL, other findings seem to suggest equipoise in the findings of no difference in the commonly used outcome measures of FAS and complications. Admittedly, most of the included studies would have defined failure as no more than 2 or 3 failed attempts in their protocols.  Nonetheless, there is now lots of published evidence of the increased morbidity that occurs in the emergency setting with more than one attempt at intubation.^{9 10 11}  This makes first attempt success an important outcome measure:  ultimate success may be compromised if it comes at the price of a traumatized airway, hypoxemia during the attempts or aspiration of gastric contents.  Thus, although a secondary outcome, the first attempt success in the Cochrane review is an important result, and unfortunately, it showed no difference between VL and DL.

Each camp therefore has rationale to maintain status quo either for or against DL or VL for their Plan A attempt. Many clinicians, educators and academics have been hopeful that the literature on this crowded device landscape would eventually provide more clarity on the frequently conflicting reports and provide a definitive answer. Although the authors (Lewis et al.) should be congratulated for their work in addressing this complex issue, for many, the results may seem at first look to have further muddied the water.

One of the major translational problems of this review paper surrounds defining what a video laryngoscope is. In its basic form, most would agree that the term describes a laryngoscope with a distal camera near the end of the blade that provides visualization of glottic structures on a proximal display. And yet there is significant heterogeneity within this description. Broadly speaking, there are 3 classes of VL devices: ¹²

1. Video-enhanced Macintosh (aka standard geometry blade) laryngoscopy e.g., C-MACâ (Mac Blade), McGrathâ Mac (Mac blade), GideScopeâ Titanium Mac, Venner APAä(Mac blade);
2. Indirect (i.e., hyperangulated or hyper-curved blades) VL e.g., C-MACâ (D-Blade), McGrath Mac (X bladeä) standard GlideScopeâ, KingVision (non-channeled blade);
3. Channeled hyperangulated blades VL e.g., King Visionâ, Pentax AWS, Airtraqä

It is well appreciated that VL intubation success rates may differ significantly between the classes of device as defined above, and to a lesser extent there are also enough blade and camera positioning design differences that within-class performance may also vary. The Cochrane review acknowledged this device heterogeneity.  By the way, this might be a good place to make a plea for specifying exactly which blade is being studied or discussed when publishing on airway devices.  For example, the ‘C-MAC’ is a system:  with the system, you can use a Macintosh blade, the D-blade, reusable or disposable versions of both blades, or even a flexible bronchoscope. All very different devices in their indications and how they’re used. The same goes for the McGrath (McGrath Series 5 or McGrath Mac?  If McGrath Mac, with its Mac blade or X blade?), and even the GlideScope.  Still in 2016, and even in the present Cochrane review, we are made to guess and infer exactly which device is being referred to.

It’s difficult to glean more information from the body of the paper without chasing down the many papers included in this review.  However, in the discussion of the Cochrane review, the authors note that no significant difference was found for failed intubation for any of GlideScope, McGrath or Pentax compared on their own to DL. The effect only remained statistically significant for the ‘C-MAC' (presumably referring to its use with the Macintosh blade) – as well as all VL devices lumped in together. Thus, if in fact the significant device-specific effect persists only for C-MAC/Mac blade VL vs. DL, on the face of it, this seems to call in to question whether the indirect, hyperangulated-type VL blades are indeed of value in diminishing the incidence of failed intubation, compared with DL when used routinely.  Conversely, why might the video-enabled Macintosh blade VLs have an advantage over Macintosh DL, as is suggested by the Cochrane review but also by Sakles et al.?¹³ It is likely related to several issues, but perhaps most importantly it probably relates to the more direct, line-of-sight corridor afforded by the video-enhanced Macintosh type blades (as with DL).  This tends to facilitate tube passage through the oropharynx to and through the larynx and on down the trachea:  once one sees the larynx with DL, rarely does trouble ensue with tube passage – and if it does, it’s usually easily handled with a bougie. (For more see: Macintosh Video Laryngoscopy: Why does it add value?)

What then of the indirect-type VLs?  Why might some of them have under-performed the video-enhanced Mac blades with respect to failed intubation in the Cochrane review?  Well, we’ve all had the experience of having a great (e.g., Grade 1, POGO 100%) view of the cords with indirect VL, only to then struggle to get the tube to and down the trachea.  There’s nothing more frustrating. So, why does this happen?  We submit that it’s because when performing indirect VL, the clinician is mistakenly applying the same goal that’s typically sought for DL:  trying to obtain as full a view of the glottis as possible.  When doing DL, seeing a full view of the cords also implies a straight corridor (‘aligning the axes’) between the clinician’s eyeball and the larynx, down which the tube can easily be passed, so there’s a good reason to seek a maximized view.  Not necessarily so for indirect VL.  First, because of the hyper-curved or –angulated nature of the blade, some aspect of the cords will often be obtained with little effort, but without also delivering a straight, direct, line-of-sight corridor for tube passage.  Secondly, as the clinician then mistakenly attempts to improve to a full view of the cords, very often he or she will rotate the scope backwards (handle backwards/blade tip forward) to achieve that view.  This is counterproductive and only increases the angle around which tube delivery must now occur, i.e., it accentuates the bend further away from the straight, line-of-sight type corridor that makes tube delivery easy with DL.  In our experience, this can be mitigated when using indirect VL by only seeking a Grade 2 view of the larynx (i.e., only the posterior/arytenoid cartilages plus just a hint of the actual glottic opening), with the blade positioned further away from the larynx than might occur if seeking a maximized view.  This is manifested on the display screen in a number of ways:  (a) one has the impression that the camera is ‘looking down’ on the larynx, from further away, rather than ‘looking up’ at a full view, from close up; (b) the view of the larynx on the screen often ends up being in the superior half of the video display (figure 4. & 5A.), and (c) only a dark hole is visible between the posterior vocal cords:  with too good a view, one can often see the anterior tracheal wall beyond the cords figure 5B.).(Levitan ACEP)  – if you see this, you need to back away with the blade! So, why does this work? In all likelihood, it’s because in obtaining the deliberately restricted view, one ends up angling the indirect VL blade more dorsally, and probably with more tongue retraction, essentially resulting in doing more of a direct-type technique with the indirect blade.  This results in closer to a straight corridor (or at least, less of a bent corridor!), in turn making tube passage easier. If not using a channeled blade, one will still need an endotracheal tube styleted with the 60-70 degree angulation inherent in the GlideScope’s GlideRite stylet or the C-MAC Guide stylet, but no more. (Fig. 6A)

Hopefully, as clinicians learn the tips and tricks needed to consistently sink the tube with indirect VL, future comparisons of indirect-type VL blades will also show an advantage over DL in all settings.   For now, though, we know that the indirect blades will enable a better view than that offered with DL, and with appropriate clinician experience in tube delivery, they may offer utility when difficulty is predicted (or has been encountered) with direct laryngoscopy. Currently there are no well validated predictors for difficult viodeolaryngoscopy beyond a soiled or bloody airway.

Meanwhile, what should we recommend for the video-enabled Macintosh VL blades?  We would argue that these devices represent the best of both (DL and VL) worlds:  there’s no need to artificially hold back from getting a maximized view; tube delivery is generally straightforward because of the DL-type straight corridor (‘lined up axes’) that results, and because of this, the bougie will work well as an adjunct.  In addition, the videoscopic view is often slightly enhanced over the direct view obtainable with the same device (at the very least, bypassing the restrictive framing issues of lips or a moustache).

Finally, does how does this Cochrane review impact a recommendation for what should be used as a Plan A intubation technique?  First off let’s dispel the myth that you are putting the patient at risk by using a direct laryngoscope. Its optimal use with bimanual laryngoscopy alone or in combination with a simple adjunct such as a bougie is a safe and reasonable Plan A approach to airway management. However, if you have and are experienced with a video-enhanced DL device (C-MAC with Mac blade, etc.) then it may be the preferred approach over traditional DL.  Conversely, unless an expert, and/or you believe the patient has significant predictors of difficult DL, we don’t believe this review supports the routine use of indirect VL as a Plan A technique.  Either way (indirect VL or video-enhanced DL), one should always have a Plan B technique in mind:  for DL or video-enhanced DL used as Plan A, Plan B could be indirect VL;  for indirect VL used as Plan A,  Plan B could be DL or a video-enhanced DL.

Figure 4: GlideScope restricted view (50% or less POGO, glottis in upper 50% of screen)

Figure: 5A. Restricted view where blade/camera are in-line with trachea (POGO <50%, Glottis <50% Screen; trachea= black hole) 5B. Full view over-rotated, too close (anterior trachea/cricoid ring in view)

Figure 6A. Restricted view allows use of a less angulated ETT 6B. Full view requires more of a distal bend on ETT

To summarize what time, experience and evidence has taught us is that VL seems most likely to succeed when it is used like a direct laryngoscope in aligning the blade/camera with the tracheal axis. The “around the corner” best view is visually pleasing but often creates functional challenges for tube delivery. So if we are to use VL, perhaps we should be using a VL device designed like a direct laryngoscope (Macintosh VL) which may alternatively be used for DL when indirect conditions are unfavorable (e.g., blood or vomitus in the airway) or unavailable (e.g., video screen fails). The advantages of this approach are numerous as all of our well-established approaches to difficult DL such as head lift, ELM and bougie use are then interchangeable skills regardless of whether Macintosh VL or DL device is used. In contrast, experience and evidence suggests that DL and indirect VL skills are not exchangeable and in fact indirect VL dominant use likely contributes to the erosion of DL skill use.^{1 2 3} Indirect VL devices still have a significant role but for many of us the evidence reinforces their being reserved for a plan B approach in most situations.

The times will continue a-changin’ and we should rejoice in the many advancements we have witnessed in airway management. There are risks and benefits to adopting technologies, especially when there is no requirement for proof of effectiveness on humans to enter the airway device market. The 2013 Federal Aviation Authority Report warned us that the greatest threat to modern aviation may be our pilots’ dependence on automation. We need to heed a similar warning and be careful that we have the appropriate tools and skills to manage our patients. The environment will not adapt to our skill set (from Deep Survival¹⁴).

Update

George posted this video in reply to some of the comments below

References:

1. Lewis SR SA. Cochrane Database of Systematic Reviews Videolaryngoscopy versus direct laryngoscopy for adult patients requiring tracheal intubation (Review) Videolaryngoscopy versus direct laryngoscopy for adult patients requiring tracheal intubation (Review). 2016;(11). doi:10.1002/14651858.CD011136.pub2.
2. Schriger DL. Does Everything Need to Be “Scientific”? Ann Emerg Med. 2016;68(6):738-739. doi:10.1016/j.annemergmed.2016.06.043.
3. Kovacs G. Airway management?: “‘ the times they are a-changin .’” 2013;0(0):1-4. doi:10.2310/8000.2013.131106.
4. Perry JJ, Lee JS, Sillberg VAH, Wells GA. Rocuronium versus succinylcholine for rapid sequence induction intubation. Cochrane Database Syst Rev. 2008;(2):CD002788. doi:10.1002/14651858.CD002788.pub2.
5. Gu Y, Robert J, Kovacs G, et al. A deliberately restricted laryngeal view with the GlideScope® video laryngoscope is associated with faster and easier tracheal intubation when compared with a full glottic view: a randomized clinical trial. Can J Anesth Can d’anesthésie. 2016. doi:10.1007/s12630-016-0654-6.
6. Park L, Zeng I, Brainard A. Systematic review and meta-analysis of first-pass success rates in emergency department intubation: Creating a benchmark for emergency airway care. Emerg Med Australas. 2016;(September). doi:10.1111/1742-6723.12704.
7. Kerslake D, Oglesby AJ, Di Rollo N, James E, McKeown DW, Ray DC. Tracheal intubation in an urban emergency department in Scotland: a prospective, observational study of 3738 intubations. Resuscitation. 2015;89:20-24. doi:10.1016/j.resuscitation.2015.01.005.
8. Brown CA, Bair AE, Pallin DJ, Walls RM, NEAR III Investigators. Techniques, success, and adverse events of emergency department adult intubations. Ann Emerg Med. 2015;65(4):363-370.e1. doi:10.1016/j.annemergmed.2014.10.036.
9. Mort TC. Emergency tracheal intubation: complications associated with repeated laryngoscopic attempts. Anesth Analg. 2004;99(2):607-13, table of contents. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15271750.
10. Sakles JC, Chiu S, Mosier J, Walker C, Stolz U. The importance of first pass success when performing orotracheal intubation in the emergency department. Acad Emerg Med. 2013;20(1):71-78. doi:10.1111/acem.12055.
11. Natt BS, Malo J, Hypes CD, Sakles JC, Mosier JM. Strategies to improve first attempt success at intubation in critically ill patients. Br J Anaesth. 2016;117(May):aew061. doi:10.1093/bja/aew061.
12. Kelly FE, Cook TM. Seeing is believing: getting the best out of videolaryngoscopy. Br J Anaesth. 2016;117(suppl 1):i9-i13. doi:10.1093/bja/aew052.
13. Sakles JC, Mosier JM, Patanwala AE, Arcaris B, Dicken JM. The Utility of the C-MAC as a Direct Laryngoscope for Intubation in the Emergency Department. J Emerg Med. 2016;51(4):349-357. doi:10.1016/j.jemermed.2016.05.039.
14. Gonzales L. Deep Survival: Who Lives, Who Dies, and Why. W. W. Norton; 2004. https://books.google.ca/books?id=mQLDLn0GQfwC.