A helmet is an interface that has an extremely large internal volume, up to 10 l, which constitutes a huge dead space, promoting CO₂ rebreathing [4, 13]. In fact, a helmet cannot be considered as an interface sensu stricto. It is rather a kind of hermetic tent over the patient’s head, with the aim of creating a high-pressure environment for the respiratory system. The pressure is maintained by the constant flow of gas that enters by the inspiratory limb and exits by the exhalation limb. The internal volume of a helmet is much larger than the patient’s tidal volume. The patient takes a breath from the gas located under the helmet and then exhales into this specific semi-closed atmosphere. However, the volume of a helmet does not correspond to effective dead space. Fodil et al. [14], using a mathematical model, estimated the effective dead space of a helmet by analyzing the pressure field/flow pattern. They found that effective dead space consists of only 4 % of helmet gas volume due to the streaming effect. Exhaled CO₂ is diluted in internal gas volume and inhaled gas may be only slightly enriched with CO₂. To correctly perform NIV with a helmet, sufficient gas flow has to be provided. The flow is aimed not only to maintain adequate expiratory and inspiratory pressures but also to constantly extract CO₂ from the inner space. The concentration of CO₂ in the helmet gas is dependent on PaCO₂ [6]. Thereupon, the flow must be adjusted according to the level of patient’s hypercapnia and generally should exceed 30 l/min.

Antonelli et al. [15] published one of the first studies comparing efficacy of helmet NIV and NIV with mask in hypercapnic respiratory failure. In this matched case control study, the efficacy of helmet NIV was worse in terms of reducing PaCO₂, which may confirm the important role of CO₂ rebreathing in the treatment effects. However, lesser efficacy of ventilation could partly be attributed to worse patient-ventilator synchronization in comparison with mask ventilation. In a randomized study published in 2015 by Pisani et al. [16], a new model of helmet, specifically designed to improve performance in hypercapnic patients, was tested in comparison with oronasal mask. The novel helmet (CaStar R Next, StarMed, Mirandola, Italy) has less internal volume than standard ones. Moreover, the authors used ventilator settings specifically focused on optimization of patient-ventilator synchronization: fast rate of pressurization and a cycling-off threshold between 25 and 50 % of maximal inspiratory flow. The pressures were titrated to provide a flow inside the helmet >30 l/min. The mean inspiratory and expiratory pressures were 19.6 ± 5.7 cmH₂O and 7.7 ± 1.9 cmH₂O, respectively, which was ~30 % higher than in mask ventilation. The results showed the same efficacy in improving hypercapnia and respiratory acidosis in both groups.