“To see clearly, it is often enough to change the direction of your gaze.” – Antoine de Saint-Exupéry
Intubations on intensive care are recognized to constitute a more ‘difficult airway’ than cases performed in theatre. This is due to comorbidities, oxygen-dependence, time for preparation/positioning, laryngeal oedema etc. Multiple case-series have suggested some advantage in the use of videolaryngoscopes over traditional direct laryngoscopy (particularly in less-experienced intubators).
Does use of the McGrath videolaryngoscope (VL) improve the frequency of successful first-pass orotracheal intubation by comparison with the Macintosh direct laryngoscope (DL)?
This was a prospective, non-blinded multicenter RCT carried out in 7 ITUs in France between May and December 2015. On a baseline success rate of 65% (using DL) and an expected improvement of 15% (using VL) the trial aimed to recruit 370 patients to give 90% power at a significance level of 5%.
All ITU patients that needed intubation were considered for inclusion. Those who were less than 18 years old, pregnant/breastfeeding, under state guardianship, had a contraindication to oral intubation, lacking health insurance or had previously been involved in a trial with intubation as primary outcome were excluded. In addition, in cases such as cardiac arrest where there was insufficient time to randomize, these patients were excluded.
All eligible patients were included in assessing the primary outcome measure without the need for consent as both interventions were considered acceptable practice. Following inclusion, patients or relatives could withdraw consent to be included in secondary outcome analysis.
All participating physicians had been fully trained in the two techniques of intubation. Physicians were subdivided into expert (either two years anaesthetics training and one year ICM training; or five years ICM training) or non-expert (all others). At all intubations, an expert intubator was present but did not necessarily do the intubation.
Preoxygenation technique was up to the physician present, as was choice of induction drugs (in the majority of cases, etomidate/sux were used). Intubation was confirmed by four ETCO2 breaths. Following a failed attempt at intubation, the physician was allowed to either continue with the laryngoscope as randomized, or change to the other. If intubation could not be achieved as above, national failed intubation guidelines were followed.
The primary outcome was the proportion of patients successfully intubated first time.
Secondary outcomes looked at complications surrounding the intubation, quality of glottic view achieved, and ultimate outcomes of the patients.
Four hundred and eighty-nine patients were screened and 371 patients were ultimately included in the trial. The Macintosh and McGrath groups were evenly matched.
There was no difference in first-pass intubation rates between the two groups, 67.7% of those in the VL group and 70.3% of those in the DL group were intubated first time (OR 1.12).
The VL group achieved better glottis visualization.
Every patient was intubated successfully although one patient from the VL group was intubated by DL due to confusion at the time of randomization.
In each group, 7.5% of patients were described as having a difficult intubation. In the DL cohort 70% of the time this was due to failure to visualize the glottis. In the VL group, 70% of the time was due to inability to intubate the trachea despite having a good glottic view.
In the VL group there were 17 episodes of life-threatening complications including four cardiac arrests, one death and six episodes of severe hypoxaemia (SpO2 <80%). In contrast, there were no cardiac arrests or deaths, and only one episode of severe hypoxaemia in the DL group.
The majority of first-attempt intubations were by ‘non-experts’. The majority of subsequent intubations were by ‘experts’. Where ‘experts’ performed the first-attempt, intubation success rate was 91.7% – regardless of the laryngoscope used.
This was a well-designed, well-matched RCT using an objective outcome measure. It addressed a relevant clinical question and used intention to treat analysis with complete follow up for their primary outcome.
Unfortunately, they tested a single VL that a lot of UK clinicians have not used – in defence of the trial, they chose this VL as it is highly portable and quick to set up with the hope that it could be more transportable around a hospital.
The trial reports a surprisingly low baseline intubation rate although this may be due to a high proportion of medical vs anaesthetic backgrounds (58-66% vs 17%).
There were some questionable exclusion criteria given that both interventions were accepted standards of treatment.
This was a well-designed trial with strong internal validity. VL did not improve frequency of first-pass intubation although seemingly contributed to a concerning number of life-threatening complications.
The background of the doctors in the study brings into question its external validity to the UK, where a high proportion of ITU staff have greater anaesthetic experience.
Overall, the trial suggests experience trumps technique.
If my ‘go-to’ device was a videolaryngoscope, and I could always have access to it, I would still use it. If my backup device was a videolaryngoscope, I would still consider it.
Summary by Dr Michael Hoy