Overall, we have shown that the NRM and BVM were comparable in terms of the degree of denitrogenation achieved. While the final FEO2 achieved was 4.9% greater for the NRM than the BVM, this difference was not statistically significant. While the average FEO2 achieved with the BVM was comparable with the NRM, the spread in the data was greater.
Our values of FE
achieved after BVM preoxygenation were somewhat lower than those achieved in Stafford et al
who found the mean FE
to be 74.2%. However, we measured FE
at the end of an expiratory reserve breath rather than sampling end tidal gas from the mask, and this is likely to be higher due to mixing and incomplete exhalation which may account for the discrepancy. We believe our method is a more accurate measure of true end-alveolar oxygen and the actual degree of oxygen reserve obtained.
There was a statistically significant difference between the maximum and minimum mask pressures generated with the NRM and BVM (around 1–2 mm Hg given the limited resolution of our monitor). This implies a greater resistance to breathing with the BVM and, while small, is not insignificant compared to the normal pressure changes seen in eupnoea and could become important in the case of dyspnoea. That the FECO2 was higher after preoxygenation could be due to either partial rebreathing or hypoventilation. However, the increased airway pressures and subjective difficulty in breathing shown by the visual analogue scale data might give credence to the latter explanation. While the increase in FECO2 may not be clinically significant, any additional resistance to breathing is clearly undesirable and likely to reduce tolerance in anxious, disorientated or combative patients.
While it would appear that in a prehospital or emergency setting, the BVM and NRM are similarly effective, it is worth pointing out that comparison of our data with the literature confirms that both underperform an anaesthetic circuit9
for preoxygenation. It is common clinical experience that mask-sampled (tidal breathing) FE
levels approach 80 or 90% after 3 min preoxygenation with a Bain or circle system.
Our study has a number of potential limitations. One potential weakness in the study design is the integrity of the seal created by the subjects holding the BVM on themselves. However, it was felt that tightly holding a mask onto a healthy volunteer was likely to cause discomfort, which may have led them to radically alter their breathing or even withdraw from the study altogether. To mitigate this, the participants were clearly instructed to hold the mask firmly and avoid leaks and the investigators visually monitored this. In most cases, mild blanching of the skin under the mask was visible. Irrespective of this, it seems unlikely that the seal with a cushioned mask and therefore the potential for air entrainment would have been significantly worse than for the NRM design. The BVM has a valve arrangement, which may allow room air to enter the system if mask pressures are sufficiently negative and while we have not evaluated the performance of this or the effect of mask dead-space, and it is conceivable that any of the above factors may have contributed to the greater variability in average FEO2 seen with the device. At the same time it is important to remember that firm application of an anaesthetic mask to dyspnoeic or otherwise confused/obtunded patients typically increases agitation and often makes maintaining an airtight seal difficult. So, while we cannot exclude the possibility of small leaks in our volunteer study, it is doubtful if this can ever be the case in the emergency situation either.
Another limitation of the study is that critically ill patients are unlikely to have a normal breathing pattern. They may have reduced or increased minute volume as well as a reduced or increased respiratory volume. Therefore, the levels of preoxygenation achieved with the BVM and NRM may exceed the efficacy in a critically ill patient: that is, our study is likely to represent a ‘best case’. Clearly, the BVM offers the potential for ventilatory support, which the NRM does not.
In conclusion, there was no significant difference between the efficacy of preoxygenation with a BVM or an NRM. The NRM was associated with lower mask pressures in normal ventilation and was subjectively better tolerated. Its use for preoxygenation is simple although clearly it is imperative to confirm that all equipment, including a method for ventilation, be immediately available and functioning before anaesthesia is induced. Although inferior, the BVM can still be well tolerated and offers the option of assisting ventilation, which is also an important consideration in emergency care. Comparison with previous published results suggests that both devices are inferior to an anaesthetic circuit and the use of such a purpose-designed system which is valveless and has minimal resistance to flow is therefore to be encouraged wherever possible. Our results, however, emphasise the limited safety margin available in emergency anaesthesia.