Preoxygenation should be performed using a tight-fitting face mask with a reservoir and ≥15 L/min of oxygen in order to provide a high (>0.9) fraction of inspired oxygen (FiO₂) during inspiration. In hospital this is done using either a standard anaesthetic circuit or a disposable Mapleson C (sometimes called a ‘Waters’ circuit) with adjustable pressure limiting (APL) valve (Fig. 3.3). This also allows an element of continuous positive airway pressure (CPAP) to be applied, as long as the face mask is tight-fitting. These systems require a constant high flow of oxygen and a tight seal to allow manual ventilation, neither of which can be guaranteed in the pre-hospital setting. The default system in the emergency situation is a self-inflating bag-valve-mask (BVM), of which there are many varieties, including many single-use versions. Some older versions of BVMs did not have one-way exhalation valves near the mask and therefore allowed entrainment of air into the face mask at lower oxygen flow rates. These would deliver relatively low FiO₂ (<0.4) during spontaneous ventilation (Mills et al. 1991; Nimmagadda et al. 2000). This is not the case with the majority of BVM systems currently in use, however, it is always important to check the type of BVM being used.

Even when a BVM has a one-way exhalation valve, there is variation in the FiO₂ delivered by different BVM systems in common use. As an example, two single-use BVMs with exactly 15 L/min of oxygen attached, were compared for both ease of breathing and delivered FiO₂ (using a calibrated oxygen analyser in a t-piece behind the face mask).

The Ambu Single Patient Use Resuscitator (SPUR) II (Fig. 3.4a) has an extremely low resistance inspiratory/expiratory single-shutter (one-way) valve at the patient end (Fig. 3.4b) which provided only minimal resistance to breathing. The valve prevents entrainment of air during inhalation, so only oxygen from the bag is inhaled if the mask is correctly sealed on the patient’s face. Even during large inspiratory breaths, the FiO₂ remained above 0.9 at all times and at normal tidal volumes was 0.95. The large safety inspiratory valve at the reservoir bag end of the BVM assembly (Fig. 3.4c) is designed to allow air to be entrained if the reservoir bag is completely empty, to prevent asphyxiation. This only opened a little even during large tidal volumes; hence the FiO₂ only dropped slightly due to indrawn air mixing with the oxygen.

The Marshall Classic Manual Resuscitator (Fig. 3.5a) has a Laerdal type ‘duck-billed’ inspiratory valve and a flat silicone ring one-way expiratory valve at the patient end (Fig. 3.5b). The ‘duck-billed’ valve is subjectively more difficult to breathe through. The additional negative pressure generated, along with the small size and lightweight safety inspiratory valve (Fig. 3.5c), means that air is entrained through the valve even during relatively small volume breaths. This reduced the measured FiO₂ from 0.9 at normal tidal volume down to 0.75 during larger inspiratory breaths. It is useful to know that both the ease of breathing and the inspired FiO₂ can be improved significantly by gently squeezing the bag in time with the patient’s inspiration.

Even more important than the design of the valves is a good seal between the mask and the patient’s face. Gas will always take the path of least resistance. During inspiration with a poor seal, gas will be entrained around the edges of the mask rather than drawing from the oxygen in the bag and reservoir via the one-way inspiratory valve. The requirement for a good face seal also applies to disposable non-rebreathe reservoir masks; the path of least resistance is also around the edge of the mask rather than through the one-way valve (although the resistance through the valve does tend to be less than in a BVM).

With a good seal and a patient with reasonable inspiratory effort, a BVM with a reservoir bag and >15 L/min will provide a high FiO₂ (>0.9) and avoids the requirement to change face mask during the RSI (Nimmagadda et al. 2000). It also means that ventilation can be easily assisted if inspiratory breaths are shallow or if SpO₂ drops whilst waiting for the muscle relaxant to work.

BVM preoxygenation does, however, require a competent additional person to maintain the tight face seal whilst the team is preparing for RSI. It may be more practical (and also possibly less daunting for the patient), to use a well-fitting ‘non-rebreathe’ face mask with reservoir bag for preoxygenation. This is held in place with an elastic strap placed behind the head. Basic models are not ideal as they have a thin plastic edge that may not create a good seal, and an open hole in one side of the mask that allows air to be entrained if the oxygen supply fails (Fig. 3.6). Both of these factors can result in a relatively low FiO₂ (0.6–0.7), even at very high oxygen flow rates. Newer rebreathe masks have softer cushioned edges, which improve the seal, and have no open holes in the sides of the mask, thereby increasing the chance of delivering an FiO₂ closer to 0.9 whilst freeing up an extra person (Fig. 3.7). With additional nasal cannula oxygen, as described below, this can increase the FiO₂ further (Fig. 3.8). Clearly if the patient requires assistance to maintain airway patency or needs assisted ventilation, this benefit of freeing a person is lost.

Three minutes of normal tidal volume breaths is acceptable preoxygenation for most patients. This preoxygenation can be improved by asking the patient to fully exhale and inhale (vital capacity breaths), although most patients in the pre-hospital setting are unable to co-operate with such a request. Analgesia prior to preoxygenation may improve the effectiveness by reducing pain from chest trauma and increasing inspiratory volumes.

In agitated patients it may be necessary to use sedation to facilitate preoxygenation. Small doses of Midazolam (1–2 mg) or Ketamine (10–40 mg) should be titrated to effect. Weingart et al (2015) reported the successful use of Ketamine 1 mg/kg (+/− further 0.5 mg/kg aliquots) to facilitate preoxygenation in agitated patients. This then allowed 3 min of pre-oxygenation, before full sedation and a muscle relaxant were given. Although based on a relatively small convenience sample of patients unable to tolerate preoxygenation due to delirium, they demonstrated the ability to increase mean pre-laryngoscopy SpO₂ from 89.9 to 98.8 %, and reported no complications such as pre-muscle relaxant apnoea, peri-intubation emesis, cardiac arrest or death, from doing so. Preoxygenation was achieved with a high flow non-rebreathe oxygen mask, or non-invasive CPAP if SpO₂ <95 %. They commented that facilitating preoxygenation in this manner avoids the risks of proceeding to RSI without an adequate oxygen reserve in the patient, and the potential risks of peri-intubation BVM ventilation that may be required to prevent desaturation; such as gastric insufflation and aspiration. They termed this ‘delayed sequence intubation’ although it is essentially just sedation to facilitate preoxygenation. This concept has been advocated on the PHA course for several years.

SpO₂ will fall precipitously during the 45–60 s of apnoea whilst awaiting onset of muscle relaxation, if starting SpO₂ is already low. In these patients, if already receiving oxygen through a BVM, it may be advisable to gently augment the patient’s own respiratory effort prior to induction in order to maintain a safe oxygen saturation throughout the procedure. CPAP can also be applied to improve oxygenation if it is possible to fit an external positive end expiratory pressure (PEEP) valve to the BVM.

If ventilation is continued following induction and apnoea, cricoid pressure will help to avoid the associated gastric insufflation and reduce the risk of aspiration, although this may make ventilation more difficult in some cases. BVM ventilations should aim to be delivered slowly with both low volume (6–7 mL/kg) and low rate, in order to avoid inspiratory pressures sufficient to overcome the lower oesophageal sphincter pressure.

There is evidence that head-up positioning delays the time to desaturation during apnoea (Lane et al. 2005; Ramkumar et al. 2011). It may also reduce the risk of passive regurgitation. This may be achieved by using a semi-recumbent position on a stretcher trolley or, in trauma patients requiring spinal immobilisation, a reverse-trendelenburg position, with the head of the spinal board/scoop stretcher/trolley raised 30° above the feet.

Although described in the medical literature for over a century, the application of apnoeic oxygenation to pre-hospital airway management procedures is relatively recent (Weingart et al. 2011). It utilises the physiological principle that even without respiratory effort, approximately 250 mL/min oxygen is taken from the alveoli into the bloodstream. If oxygen is supplied to a patent upper airway, a mass flow of gas from the pharynx to alveoli maintains this diffusion of oxygen into the bloodstream, and in doing so, maintains oxygenation without ventilation. The Greater Sydney Helicopter Emergency Medical Service (HEMS) has included apnoeic oxygenation in their PRSI standard operating procedure (SOP) since 2011. During the preoxygenation phase, whilst a non-rebreather mask or BVM set at 15 L/min oxygen is used to preoxygenate, nasal cannulae are applied and set to 5 L/min oxygen, as tolerated (Fig. 3.8). After induction drugs are administered, the flow rate of the nasal cannulae is increased to 15 L/min until the endotracheal tube is secured (Wimalasena et al. 2015). During the 22-month period prior to apnoeic oxygenation, 22.6 % of patients had a desaturation episode below 93 % during RSI, compared to 16.5 % of patients in the 22-month period after introduction of apnoeic oxygenation to the SOP. They suggest this simple and inexpensive technique is of clear clinical benefit to critically ill and injured patients undergoing PRSI.

Preassessment should be easily remembered with a familiar <C>ABCDE approach. It is a similar, but much more rapid assessment, to that which occurs before any anaesthetic in the hospital environment.