Is It Better to Have Seen and Failed or Never to Have Seen At All?

UPDATE:  The manuscript of the MACMAN study was simultaneously published in JAMA and presented at SCCM (Lascarrou JB, et al. JAMA. Published Online Jan 24, 2017).  While the peer reviewed primary results are unchanged, demonstrating that videolaryngoscopy improves the glottic view but does not improve the ability to pass the endotracheal tube compared to direct laryngoscopy, one of the secondary results deserves some mention.  It wasn’t presented in the brief late-breaker presentation at ESICM, so I hadn’t discussed it previously.  Lascarrou and colleagues also found a higher rate of severe life-threatening complications with videolaryngoscopy, namely severe hypoxemia, hypotension, and cardiac arrest.  While I am not entirely sure what to make of this finding, I do think it should cause us some pause.  If this is true (and it may just be a type 1 error that secondary analyses are prone to), then not only should we not use VL as the primary intubating device for all our intubations in the ICU, but we should think carefully about which ones we do use it for (and be selective in whom we use it in).  And, while these most recent studies have been informative on the risks and benefits of both VL and DL, there remain unanswered questions.  These studies do not include patients who the clinician believes they have to use VL for intubation, nor do they look exclusively at high risk populations such as morbid obesity or high Mallampati or MACHOCA scores.  I suspect additional studies will be forthcoming in the future and that this field will continue to evolve.  However, in the meantime, both of these most recent well-designed and conducted randomized controlled trials, make it apparent that we should not be using VL as the default intubation device in all our critically ill patients.


October 4, 2016;  08:55

In the last couple of decades, some of the biggest advances in medicine have involved technology.  The videolaryngoscopy represents one of these technological advances for assisting with endotracheal intubation.  FDA approval of devices simply requires demonstration that the device does what it claims to do, but does not require demonstration of improved outcomes or any clinical benefit. While there are a number of different videolaryngoscopes, they are all designed with a camera on the end of a laryngoscope connected to some form of a screen that displays the image the camera projects.  Numerous studies have demonstrated that videolaryngoscopes improve the view of the glottis and vocal cords (Cormack Lehane Glottic View Grade) over direct laryngoscopy during endotracheal intubation.  Some of these studies suggest that time to intubation may also be shorter using videolaryngoscope.  These data (probably combined with some allure of new technology) resulted in many opining that the videolaryngoscopy should be the standard of care for intubation in critically ill patients, and should be used in all intubations in the ICU, especially by inexperienced operators.  However, the studies providing the data supporting these recommendations were conducted in the operating room on patients undergoing intubation for surgical procedures.  And most of them studied experienced anesthesiologists as the operators.

Recent studies have suggested that extrapolation of these older data into critically ill patients is premature.  Our group conducted a study comparing VL with DL in critically ill patients undergoing intubation by relatively inexperienced operators (pulmonary and critical care fellows) (Janz DR, et al. Crit Care Med. 2016; Epub ahead of print, June 28).  While difficulty of the glottic view was recorded, the primary outcome was actually first pass success – or whether or not the patient was successfully intubated on the first insertion of the laryngoscope.  Any patient who was not thought to absolutely need either VL or DL, and whom the fellow had time to grab a randomization envelope from the ICU work room, was enrolled in the study, resulting in 150 patients randomized, all from the MICU.  While any videolaryngoscope was eligible for use in the study, the McGrath was the one actually used by the fellows.  The study confirmed the previous findings that VL improved the glottic view.  However, despite this improved view, VL failed to demonstrate significantly improved first pass success compared to DL (68.9% vs 65.9%; P=0.68), and this held true even when the analysis was adjusted for operator experience in using VL (OR for first pass success with VL 2.02; 95% CI 0.82 – 5.02).  Other clinical endpoints, including lowest oxygen saturation, time to successful intubation, ICU length of stay, and in-hospital mortality were not significantly different between VL and DL.  During the review process of the manuscript, many believers in VL questioned the validity of these results, believing that VL had to be better, and that these results were either an anomaly, or specific to the McGrath laryngoscope.

However, additional data have now been published that support these results and demonstrate that they are not limited to the McGrath VL.  Two recently published randomized trials comparing the C-Mac VL with DL in intubations conducted in the Emergency Department found similar results to ours (Driver BE, et al. Acad Emer Med. 2016;23(4):433-439; Sulser S, et al. Eur J Anaesthesiol. 2016 Aug 16. [Epub ahead of print]).   The Sulser study also confirmed a better glottic view using the C-Mac VL over DL.  However, in both studies, the C-Mac VL had similar rates of first pass success as DL, with no difference in clinical outcomes including time to intubation, aspiration pneumonia, or hospital length of stay.  Interestingly, the operator may not matter as much as previously thought.  Driver et al studied emergency medicine physicians and trainees as the operators while the Sulser study included mostly experienced anesthesiologists as the operators.

Now, the results of the largest randomized study of VL vs DL to date have been released.  The results of the MACMAN study were presented yesterday at the European Society of Intensive Care Medicine (ESICM) International Conference (ESICM – ICU Clinical Trials).  MACMAN is a randomized controlled, multicenter (7 centers in France) trial comparing the McGrath VL with Macintosh DL in orotracheal intubation of 370 critically ill adults (MACMAN Protocol).  Any operator was eligible to participate, although most were “inexperienced ICU physicians” (defined as fewer than 5 years of ICU experience).  Like the recently published studies, MACMAN demonstrated a better Cormack Lehane grade glottic view with the McGrath VL.  However, there was again no difference in first pass success (71% for VL vs 68.5% for DL; P=0.54) with rates very similar to our results.  During the study, the reason for failed first pass success was recorded – in the McGrath group, 70% of the failures were due to inability to pass the endotracheal tube (ETT) compared to the DL group where 70% of the failures were due to inability to visualize the trachea.  Although clinical outcomes were similar in both groups, the VL group did have more severe life threatening complications (9.5% vs 2.2%), but similar overall hospital mortality.

Given all these studies, it seems pretty clear to me that VL does improve the glottic view – but at the expense of increased difficulty in passing the endotracheal tube.  In thinking about it, I don’t think this is all that surprising.  Using DL, you are passing the ETT directly in the line that you are visualizing.  Using VL, you are trying to pass the ETT indirectly into the trachea using the video screen, and the direct line of passage is often not clear (or even very direct).  So, what is my recommendation for which device to use?  This is a like asking, “Is it better to have seen and failed or never to have seen at all?”  I think it depends.  I don’t think there is one single correct answer.  Each likely has advantages and one should use those potential advantages to better facilitate endotracheal intubation.  If you are supervising another physician and want confirmation that the ETT went through the cords, maybe VL is a better choice because it will allow you to see that.  However, I would recommend that you become proficient in using both because the more tools you have in your armamentarium, the safer your patients are going to be during your intubations.

Let me know your thoughts.



9 thoughts on “Is It Better to Have Seen and Failed or Never to Have Seen At All?”

  1. Thanks for your analyze.
    I agree with you than VL is not the magic bullet to improve intubation process in the ICU or than we have not yet test the good one and the best protocol.

  2. A fantastic piece on the continual obsession that VL MUST be better than DL, and yet really no good evidence of this. The purpose of tracheal intubation is to place an endotracheal tube into the trachea ideally at the first attempt, without delay or harm to the patient. The Macintosh blade continues to fulfill this admirably.

    Well done on a great opinion piece

  3. Hi Todd, this is Jim DuCanto from Milwaukee.
    I mainly teach Internal Medicine residents at my hospital for one of their procedural skills rotation, and I use the McGrath MAC as the instrument of choice due to economy and availability, and the familiarity of a MAC shape (I am an Anesthesiologist).
    I am an airway management educator, and teaching the airway-naive has given me many insights into routine and difficult airway management.
    1. Experienced airway endoscopists–can become naive–under severely stressful conditions. They make mistakes–simple ones–that cost them the opportunity to successfully complete the procedure. The cure for this is partially stress inoculation, but the better bet is better understanding, devices and skills for pre-intubation ventilation. Mask ventilation, lung recruitment, devices like the Oxylator–they eliminate the danger and excitement that cause competent endoscopists to become unglued, and become incompetent.
    2. Video laryngoscopes allow endoscopists to make mistakes that they can’t make with direct laryngoscopy. Those mistakes are chiefly the placement of the VL blade too close to the larynx for tube delivery. Tube delivery with VL is also difficult if the VL blade placement is sloppy, i.,e., not creating the space for tube delivery. This happens regularly with the McGrath MAC when it’s used like a King Vision or a Glidescope, i.e, going down the middle of the airway-tongue. Go down the middle with a McGrath MAC (or a Storz CMAC) and the tongue hangs over the right side of the blade, obliterating the space for tube delivery. You can’t make that mistake with DL, because if you do, you’ll immediately correct it–because you won’t see anything around the base of tongue unless you do.
    3. THE SUGGESTION: Use Mac-shaped VL’s as DL initially to avoid the two common mistakes (getting too close to larynx, not creating space for tube delivery), then take advantage of the video view. HOLD THE ETT AT THE TOP DURING TUBE DELIVERY so the tip of the ETT swings through an arc in the sagital plane to allow tip to rotate around base of tongue and around into the larynx. I call that the “One Arm-Bandit Manueuver” like you are pulling the lever on a Slot-Machine in a Casino.
    4. These mistakes perpetuate for some simple reasons–the companies that offer them for sale are focused on selling, and not on helping the customer optimize the use of the tool. WE are charged with that task as educators.
    5. Contact me by Twitter if you wish to chat about this and more, @jducanto

    1. Hi Jim. Thank you for your comments. I think the key for us was trying to determine if VL was better than DL in a pragmatic way – meaning in the way that both are used in practice. A lot of literature comments on VL demonstrating a better view – in fact, almost all studies have demonstrated this. However, practitioners need to understand that just because you get a better view with VL, that doesn’t necessarily mean it will be easier to actually intubate the patient with an endotracheal tube. There is certainly a learning curve with intubation – but interestingly, in the study that explicitly documented it, experience didn’t alter the results. Even in the most experienced operator, VL still improves the view but not the ability to intubate. This doesn’t mean that we should never use VL – part of caring for our patients better is understanding our technology, its advantages and limitations, and using those to better care for patients.

  4. Interesting blog.
    The recent Cochrane review adds strong support for VL. Though of course it’ll just kindle more discussion. To me it seems the above studies are unlikely to be large enough to identify differences in complications that should be relatively infrequent – type 2 error – unless basic level of care is poor.
    I think first pass success is a good quality assurance measure and it is important to strive to keep this rate high to reduce morbidity. However it is failed intubation and it’s sequelae that leads to deaths and too many of these start with a poorish view at DL (not an impossible one). VL has a role in making these views better and facilitating success. While I am a supporter of RCTs to explore this further they are unlikely to be the answer to this particular question – do VLs reduce major airway complications? Registries or extrapolation from other data may be a better route to the answer.

    We’ve been using VL basically as a routine tool in our dept for several years now and I struggle with this oft quoted report of good view but can’t pass the tube. As James infers is this due to poor technique and perhaps unfamiliarity? We don’t seem to see it as a significant problem.

    Finally – we are stuck with some hard science, some very poor studies looking at a heterogeneous group of devices some of which likely don’t help st all…..and a lot of opinion. Mine included. Hats off to those able to continue the good quality studies as this takes us forward and eventually we’all all get to a place where we agree. I wonder where it’ll be.

    1. Tim – thanks for your comment. The recent Cochrane review is interesting in a number of regards. Its primary conclusion is that VL improves intubation success, but then most of its secondary conclusions counter that primary conclusion. Furthermore, it does not include any of the recent trials on VL vs DL – trials that were done in the critical care population, in patients who needed to be emergently intubated. Ultimately failed intubations are one of (if not THE) biggest risk – and none of the trials are powered to evaluate that endpoint by itself (and I agree with you that they are unlikely to be large enough to do this in the future). The problem, of course, with registries is that there is an indication bias – patients who got VL (or DL) are likely different than patients who were intubated with the other device. So the comparison is difficult. Regardless, the fact that studies are being conducted in this area is a big advancement and the fact that you and I are able to have studies and information / data that we can discuss is huge. Thanks for reading the blog and for your comments – much appreciated.

      1. Thanks Todd

        I agree with everything you say. Except that the Cochrane review at any point goes back on its primary conclusion that VL increases intubation success – or rather reduces failures. It does in overall population as in predicted and simulated DTI. It makescintubatuon easier and dramatically reduced grade 3-4 views. It does this for experts too. And it reduces sequelae. It’s one of the first studies to suggest differential performance of different VLs. And this perhaps makes the overall findings more pertinent as they include some devices which frankly aren’t very good!

        Cochrane reviews are hard work – this is 217 pages. So look out for the update but enjoy this one while it’s here



  5. But what are we actually discussing, there are very many different types of VL available. My own dept uses Macgraths as its only VL and I just dont see the benefits purported by enthusiasts. The study from Dundee who are enthusiasts

    did a very good study and could only find at best equivalence with the Macintosh and importantly that the Macgrath Mac blade when used as a DL device was inferior. This is important. My issue isnt with VL use per se, but the push by some to make it default when the case is simply not made that it reliably outperforms what we already have with the standard Macintosh.

    Tim Cook above simply doesnt recognise the good view, unable to intubate problem with VL and suggests poor technique or training. This may of course be true but data exists in both the HEMS arena and critical care that demonstrates the high reliability of intubation success that we already have with the Macintosh and trained individuals in the real world.

    1. Yup
      The issue of lumping all VLs together and assuming they’re all the same is a problem. Until we have adequate vl-vl and decent vl-Mac studies we’ll all be offering opinions. Well done those like Mike Aziz who are leading the way. But the Cochrane review does also show evidence of differential performance and that’s important.
      The icu and ED data is sadly largely not informative at the moment but I hope will become clearer as more experience and more registries and trials are reported.

      An important analysis is looking at cases that go badly wrong and seeing how they got there. Too many start with an “almost saw the larynx, almost intubated situation”. VL helps with these.

      I wonder where we’ll be in five years. I know where I think we should be.

Leave a Reply

Your email address will not be published.