Could ‘safe practice’ be compromising safe practice? Should anaesthetists have to demonstrate that face mask ventilation is possible before giving a neuromuscular blocker?


  • I. Calder,

  • S. M. Yentis

A common instruction to trainee anaesthetists in the UK is that they should establish that face mask ventilation (FMV) is possible, before administering neuromuscular blocking drugs (NMBs). As far as we know the ‘no NMBs until FMV has been demonstrated’ rule is not evidence-based. The earliest reference to this policy we have found is in the chapter on tracheal intubation by Professors Gal and White in the third edition of the textbook Anesthesia [1]. Professor Gal has told us that they did not base their recommendation on published work, and did not intend to make failed FMV an absolute contraindication to NMBs (Gal TH, personal communication, 2007). Nevertheless, this policy seems to have become firmly entrenched in the minds of our trainees. This is curious, since it is barely mentioned in two standard UK texts on anaesthesia [2, 3], not mentioned in one other text [4], nor in texts dealing with airway management [5, 6]. Kheterpal and colleagues have reported recently on a study of 22 660 attempts at FMV [7], and their results suggest to us that it is time that the anaesthetic community discussed this issue.

At first sight the ‘no NMB before FMV’ rule seems an admirably sensible concept, but we perceive at least five possible defects.

Firstly, the theoretical advantage of the rule is not fulfilled in practice. Our trainees tell us that the advantage conferred is that a patient in whom FMV is impossible after induction can be woken up and an alternative technique applied. However, in real life this does not appear to be what happens. In both Kheterpal et al.’s study [7] and a previous, smaller one by Langeron et al. [8], no patient was woken. It may be that when a patient has an obstructed airway it is not feasible to await awakening; something has to be done to ventilate the lungs before catastrophic desaturation occurs. This is a different scenario to that obtaining after failed intubation when oxygenation can be performed by FMV. It is interesting that considerable attention has been given to the development of algorithms to aid the anaesthetist in decision making after failed intubation, but not to what the response to difficulty with mask ventilation before an attempt at intubation should be.

Secondly, it to some extent places anaesthetists on the horns of a dilemma. On the one hand enough anaesthetic must be given to permit FMV, but on the other hand the dose must allow a reasonably rapid return of consciousness should FMV prove impossible. It is conceivable that suspicion that a patient may need to be woken up might lead to under-dosing with induction agents, which could of itself result in difficulty with FMV due to an insufficient depth of anaesthesia. We have no evidence that the existence of the ‘no NMB before FMV’ rule leads anaesthetists to under-dose patients, but it is possible that the existence of the rule might lead an inexperienced anaesthetist to be fearful of administering adequate doses at induction. In our experience, more difficulty with FMV is encountered when the patient is too lightly anaesthetised than when too deeply anaesthetised.

Thirdly, it is our experience that in patients in whom FMV is initially awkward, FMV becomes easier once the NMB has been given. While it might be claimed that this observation is related to the deepening of anaesthesia after induction, rather than to the developing neuromuscular blockade itself, our view is supported (albeit anecdotally) by the observation that this improvement in ease of FMV seems to be related to the speed of onset of the NMB used. We do accept, though, that our claim is difficult to support with evidence, partly because of the difficulty of translating the ‘feel’ of the reservoir bag during FMV into an objective measurement [9]. So our suspicion is that trainees’ reluctance to give a NMB in case FMV proves to be difficult is more likely to result in FMV that is difficult, compounding any difficulty resulting from a relatively low dose of induction agent.

Fourthly, should difficulty occur then the absence of neuromuscular blockade may hinder rescue of the situation. In the report by Kheterpal et al. of 22 660 attempts at FMV [7], there were 313 patients in whom FMV was difficult (1.4%), and 37 in whom it was impossible (0.16%). Of the latter 37, tracheal intubation after direct laryngoscopy was successfully performed in 36 (10 were noted to have been ‘difficult’ intubations) and one patient required surgical cricothyrotomy. All 37 were given NMBs, either before FMV was found to be impossible or to aid intubation [10]. In an earlier study by Langeron et al. of 1500 patients, FMV was reported as difficult in 75 (5%) and impossible in one (0.07%) [8]. One can only speculate on the likely success of rescue techniques such as tracheal intubation in the absence of NMBs, in such cases. There is a considerable body of evidence to support the proposition that NMBs make intubation easier [11–14]. A conclusion that could be drawn from Kheterpal et al.’s report is that, if tracheal intubation is planned in a patient in whom difficulty with FMV is a possibility, the safest tactic is to give NMB drugs as soon as possible, so that conditions for intubation are optimised. A logical extension of this is to suggest that since not all such patients can be identified in advance, NMBs should be given routinely immediately after induction.

Fifthly, the rule might inhibit practitioners in situations where FMV is impossible but NMBs are unarguably a correct treatment, such as opioid-induced rigidity and laryngospasm from other causes [15].

Unfortunately, the work of Kheterpal and Langeron does not provide us with a template that we can apply to suspect patients in order accurately to identify the difficult-FMV ones. It seems that obesity is the background problem (BMI > 30 kg.m−2), and other factors that are associated with difficult FMV include beards, thick neck, age > 57 years, a history of snoring or obstructive sleep apnoea, poor mandibular protrusion, poor Mallampati grade and a thyro-mental distance of less than 6 cm [7]. However, the presence of one or a combination of these factors does not increase the likelihood of difficult FMV to an extent that makes decision-making straightforward.* Of course, in ‘ivory tower theory land’, where all anaesthetists are competent endoscopists, there are no economic restrictions and all patients are cooperative, we could simply perform awake intubations on all such patients, but in the real world we would not necessarily end up with a better outcome, as awake fibreoptic intubation itself has morbidity and mortality [16].

Anaesthetists will increasingly be presented with grossly obese patients who require general anaesthesia. It is this group that seems to be most at risk of difficult FMV, but we can take some comfort from the findings of Kheterpal et al. that, in most of them, tracheal intubation was possible after direct laryngoscopy. An LMA does not seem to have been tried in Kheterpal et al.’s investigation [7]. We suspect that many UK anaesthetists encountering very difficult FMV would attempt to ventilate via an LMA before direct laryngoscopy, and would certainly do so if laryngoscopy did not permit intubation. Another option is to leave a tube in the pharynx, inflate the cuff and seal the mouth and nose [17]. These tactics will only be successful if the glottis is open. In our opinion neuromuscular blockade is the most certain way to ensure glottic patency, so that administration of NMBs is more likely to be helpful than not in this dire situation.

So should anaesthetic instructors perform a volte-face and tell their trainees to give NMBs before demonstrating successful FMV? Our experience and Kheterpal et al.’s findings suggest to us that such a policy is more logical than the current one. However, we cannot point to conclusive evidence. Anaesthetists should consider their own abilities and the patient’s characteristics and opt for awake intubation or a regional technique when they believe that to be the best course. We are not suggesting that it is acceptable for anaesthetists to give NMBs to patients in whom it is obvious that FMV and intubation are going to be impossible. However, if general anaesthesia is (or has to be) the choice, we are concerned that the current rule may lead to half-hearted anaesthesia. Practitioners who believe that the administration of a NMB might help when difficulty with the airway is encountered should be able to exercise their judgement without fear of criticism. If anaesthetists hesitate to give NMB agents when necessary, this will be a retrograde step in patient safety. It may be that in the future we will be able to reverse the effects of NMBs promptly, so that everyone will feel comfortable with their early administration [18].


  • *

    In Kheterpal et al. the ratio of true positives to false positives identified, the likelihood ratio, was about three if two factors were present. This suggests that if the prevalence of impossible mask ventilation is 0.16% we can expect the post-test probability or positive predictive value to be about 0.5% [19].