The Paper
Weingart et al. Taxonomy of Key Performance Errors for Emergency Intubation
Key Factors
- All video with CMAC disposable Mac 3 or Mac 4
- RSI or DSI, not awake intubations
- Insane-level attention to preox so may not be extrapolable to places with less rigorous approach
- We don't know if the intubators were looking at the screen, the mouth, or a combination
- Most intubations performed with the bougie (and many with a manipulable-tip bougie)
- Using a mac-as-miller was not deemed an error, but a preferred move
- Almost all resident intubations
- An extensive Airway QI Program had already been implemented [See EMCrit 300]
- Videos were flagged if there was an at least one observed error on initial perusal
- ~95% FPS rate, the videos in this study were just the few non-optimal ones
- 82% Successful
List of Performance Errors
Errors of Structure Recognition during Laryngoscope Insertion:
Insertion Off Midline Leading to Esophageal Visualization
This error was coded positive when the tip of the blade overran the glottic structures by advancing to either side of the glottic structures and which led to visualizing esophagus
Overly Deep Insertion Leading to Esophageal Visualization
This error was coded when a midline insertion was used, but lack of structure recognition or too rapid an insertion led to overrunning the glottic structures and visualizing esophagus
Missed Anatomical Structure Recognition
This error was coded when the video reviewers were able to identify glottic structures during the laryngoscopy without visualization of esophagus, but the intubator did not recognize these structures as they did not correct tip position to the vallecula or glottic inlet. This error is distinct from the prior two because the intubator did not approach the esophagus
Inadequate Suction
Tube insertion was attempted without adequate structure visualization due to secretions, hemorrhage, or debris and without adequate suctioning
Most common error from this category
Errors of Vallecula Manipulation
Inadequate Lifting Force
If the tip of the laryngoscope was properly seated in the vallecula indicated by translation of the epiglottis, but inadequate lifting force was applied to expose the glottis sufficiently for bougie or tube passage
Most common error from this category
Failure to Engage Midline of Vallecula
If the tip of the laryngoscope did not engage the median glossoepiglottic fold resulting in inadequate epiglottic translation
Lost Seating in Vallecula
If during the intubation attempt, loss of vallecular engagement caused loss of glottic visualization
Not Fully Seated in Vallecula
If tip was midline in the vallecula but was not advanced forward sufficiently to cause epiglottic translation
Too Much Force/Over-Insertion at the Vallecula
Over-insertion into the vallecula causing downward movement of the epiglottis and preventing glottic visualization
Errors of Device Delivery
Bougie Delivery Issue
Bougie hung up on anterior tracheal rings without correction or intubator was unable to manipulate bougie through vocal cords
Over-Rotated Insertion (Kovacs Sign)
Unable to pass device while the glottis filled > 50% of the screen and cricoid cartilage visualized indicating over-rotation/too-deep an insertion of laryngoscope
Tube Delivery Issues
Endotracheal tube passage prevented by hangup on arytenoid cartilage without correction
Premature Withdrawal of Camera (Bougie Error)
Laryngoscope withdrawn from mouth prior to passage of the ETT over the bougie
Future Directions
- Validate on Hyper-Angulated
- Track all intubations at a program to get a true incidence
- See if errors vary by experience of the intubator
Made by my brilliant buddy, Fred Lemaire, this app will guide you through key procedures scenarios in resuscitation. It is an aide memoire and an error prevention tool. EMCrit listeners get 2 months free to check it out.
Additional New Information
More on EMCrit
- EMCrit 300 – Airway Continuous Quality Improvement and the Resus Airway Bundle
- EMCrit 176 – Updated EMCrit Rapid Sequence Intubation Checklist
Additional Resources
Professor
Nassau University Medical Center
No conflicts of interest (coi).
- EMCrit Wee (392.5) – Naughty or Nice? Bad Behavior in Healthcare with Liz Crowe, PhD - January 15, 2025
- EMCrit 392 – All Things Defibrillation with Sheldon Cheskes - January 10, 2025
- EMCrit 391 – Pericardiocentesis and Tamponade Temporization - December 27, 2024
Hey Scott Great stuff- two questions: -You mention “insane level pre-ox” I know that you talk about pre-ox strategies a lot on the podcast but what do you mean by “insane level”- I searched back and I can’t find a specific algorithm that you use but the QI post/podcast mentions end-tidal oxygen readings- unless I am just not finding what is already there- are you using end-tidal oxygen as a guide and not intubating until you reach 90%? -Can I suggest a mini-wee or post on the nuts and bolts of the “airway corner” that you run where you review… Read more »
all of that is coming!
Hey Scott. Great stuff.
Do you have the data available to stratify the frequency of error types by the type of intubation performed, i.e. RSI vs topicalized awake vs dissociated awake intubation? Would be interesting to see if certain types of errors are more common with certain types intubations, and thus one could use this knowledge to prepare for these types of errors when performing the procedure.
should have mentioned on the podcast, these were all paralyzed intubations (RSI or DSI)
Can you give a little detail about how the mac-as-miller technique is taught? Is the valeculla entered, the hypoepiglottic ligament engaged, and then the epiglottis overrun? Is this how it prevents the error of overly deep insertion? Thank you.
you get to the valeculla, engage and don’t like what you see. At this point pull back ~1 cm, let up on lifting force a bit, advance about 1.5 cm while making a scooping motion with blade tip to pick up the epiglottis. Reassert lifting force and assess view. If glottis is > 50% of the screen, pull back or derotate (with HA blade) to bring it to top 1/2 of screen
Thank you.
Will you preferentially choose a longer blade(mac 4 vs 3 for example) when you anticipate having to utilize this technique?
I do not think that that you typically can anticipate using this technique. We considered it an acceptable “plan b” technique if vallecula manipulation was too difficult. I advise our residents to start with Mac 3 as Mac 4 is rarely needed except in very tall adults and it’s size often inhibits some of the precise vallecula manipulation needed. So I think you can start with Mac 3 and if you feel that scooping the vallecula is needed but you need a longer blade, swap to the 4.
First of all, fantastic job! And kudos to everyone involved here in breaking this all down and organizing these errors into distinct categories. One comment: This is incredibly important and a largely overlooked consideration: For many years I used the reusable stainless steel CMAC standard geometry blades. They were amazing. I’d say they are just as good as regular DL blades. But for the last several years I’ve only had access to the same disposable CMAC blades used at your shop (and in your study). To compensate for the fact that they are flimsy pieces of disposable plastic, the tip… Read more »
Lots of these videos come up as blocked on multiple different computers, embedded regionally or something?
Thanks
One major category I feel was missed was errors of preparation. Was the BVM connected to oxygen? Was the oxygen turned on? Was the patient positioned properly? Was the suction connected? Was there a styler inserted before the attempt? Etc. You won’t see them on the airway cam but I see these errors all too frequently, especially when we get in a hurry.
James, are you a new listener?? This was a video study of laryngoscopy, so couldn’t address the things you mention. However, we have many episodes right up your alley! Please check out emcrit.org/300 to address the issues you raise.
Good and really amazing post. https://onlinebillexpert.com/