Over the last decade, supraglottic devices have become popular to manage the airway of anaesthetized patients undergoing minor surgical procedures [1,2]. After introduction of the laryngeal mask airway (LMA), new alternative devices have been proposed, such as the PAXpress (PAX) (Vital Signs Inc, Barnham, UK) . The PAX consists of an anatomically-curved tube with an inflatable circular cuff in the midsection, a non-inflatable gilled conical tip at the distal end, and an anterior-facing, rectangular window opening between the cuff and the conical tip, which is placed in close proximity to the glottis.
We conducted a prospective, randomized study to evaluate efficacy, side-effects and sealing pressures of the PAX compared to the LM in healthy, mechanically ventilated, anaesthetized patients. Based on previous investigations [2,3] with the two supraglottic devices, 35 patients per group were required to detect a difference of 20% in the incidence of sore throat, accepting a two-tailed α error of 5% and a β error of 20%.
Ethics Committee approval and written informed consent was obtained: 70 ASA I-II patients, aged 18-65 yr, without expected difficult airway and receiving general anaesthesia for extra-abdominal extremity procedures of short duration were studied. After induction of general anaesthesia with intravenous propofol (2.5 mg kg−1) and fentanyl (1 μg kg−1), lung ventilation was manually assisted through a facemask but without an oral airway. Using computer-generated numbers, the patients were randomly allocated to receive either an LMA (Group LM, n = 35) or a PAX airway (Group PAX, n = 35). All supraglottic airways were placed and managed according to the manufacturers' recommendations by the same two anaesthesiologists, who had adequate and similar experience with both devices. General anaesthesia was maintained with sevoflurane.
The cuff of the LM was inflated with 20-30 mL, according to its size (3 or 4). For the PAX, an initial bolus of 30 mL was injected, followed by 10 mL increments (maximum 60 mL) until an adequate seal was achieved: cuff volumes and pressures at adequate sealing were recorded. The airway sealing pressure was defined as the pressure at which gas leakage into the mouth occurred after closing the expiratory valve of the circuit at a fresh gas flow of 10 L min−1. The time required for successful placement, need for any additional manoeuvre to place the airway, the number of attempts, incorrect positioning, air leakage or other adverse events (vomiting, coughing at the positioning, gastric insufflation (by auscultation) or laryngeal spasm) were also recorded. After 20 min of stable lung ventilation arterial blood was drawn and the arterial-to-end-tidal CO2 partial pressure gradient and PaO2/FiO2 ratio were determined. At the end of surgery, the sevoflurane was stopped and, after spontaneous ventilation resumed, the cuff of the device was deflated and the supraglottic airway removed. Coughing at removal, presence of blood on the device and the occurrence of postoperative sore throat in the recovery area and at 6 h after the end of surgery were recorded.
In two patients of Group PAX (5.7%), the device had to be removed (P = 0.493). In one patient, the cuff has been damaged during insertion and it was replaced with another PAX. In the second case, adequate lung ventilation was impossible. In this patient an LM was successfully placed with adequate airway control. Both these patients were excluded from further analysis. During the first attempt successful placement was obtained in 29 patients with the LM (83%) and 24 patients with the PAX (69%) (P = 0.235). In the remaining cases, successful airway control was obtained at the second attempt. The median (range) time required to achieve adequate ventilation was 14 (10-180) s for the LM Group and 28 (15-300) s for the PAX Group (P = 0.004). Cough after insertion was reported in two LM patients (6%) and four PAX patients (11%) (P = 0.673). Three patients of Group LM (8%) and two patients of Group PAX (6%) required further manoeuvres, such as head tilting, chin lift, jaw thrust or neck rotation to obtain correct lung ventilation (P = 0.99). However, after achieving an adequate placement, no further problems were observed during mechanical ventilation, and no signs of gastric insufflation or regurgitation were reported in any patient.
Cuff volume and pressure at adequate sealing, leakage pressure and side-effects after airway removal in the two groups are shown in Table 1. There were no differences in the arterial-to-end-tidal CO2 partial pressure gradient (PAX 0.55 ± 0.53 kPa and LMA 0.56 ± 0.35 kPa, P = 0.94), or PaO2/FiO2 ratio (PAX 56 ± 21 kPa and LMA 59 ± 17 kPa, P = 0.51) after 20 min of stable lung ventilation.
The higher incidence of cough, presence of blood on the device, and immediate postoperative sore throat in Group PAX compared to Group LM suggests that the insertion of the PAX could be more traumatic than the LM. This can be reasonably related to the anatomical structure of the airway, since the conical gilled tip of the PAX can potentially induce trauma to the hard and soft palates during the insertion manoeuvre. The incidence of these side-effects with the LM was similar to that reported in a large observational study . Sore throat was short lasting, and 6 h after the end of surgery no differences were found between the two groups. In agreement with our findings, Ahmed and colleagues  reported blood on the device and postoperative sore throat more frequently in patients receiving the PAX as compared to the LM.
Despite the larger volume of air injected into the cuff of the PAX, the cuff pressure was lower than that measured with the LM; this indirectly indicates that the airway trauma observed with the PAX is related to the insertion manoeuvre rather than to the pressure exerted on the mucosa. On the other hand, the higher leakage pressure observed with the PAX together with the lack of an increased incidence of gastric insufflation could suggest a better performance of this new device during mechanical ventilation. The cuff pressure measured with LM was surprisingly high; this was probably related to the insufflation of a fixed volume of air according to the size of the mask, and suggests that an incremental insufflation technique should be preferred.
The arterial-to-end-tidal CO2 tension difference can be used as an index of physiological and alveolar dead space . We found no differences between the PAX and LM Groups in this parameter after 20 min of stable ventilation, suggesting similar effects on physiological dead space. The values reported with the LM are similar to those observed in previous investigations .
We conclude that PAX, in mechanically ventilated patients, provides similarly effective airway control with better sealing and lower cuff pressure as compared to the traditional LM, but requires a longer time to achieve successful ventilation and produces sore throat more frequently than the LM.
Department of Anaesthesiology; Vita-Salute University of Milan; Milan, Italy
1. Brimacombe J. The advantages of the LMA over the tracheal tube or face mask: a meta-analysis. Can J Anesth
2. Casati A, Fanelli G, Casaletti E, Cappelleri GL, Magistris L, Aldegheri G. Do we need new supraglottic devices? Minerva Anestesiol
3. Mondello E, Casati A, Italian PAXpress
Group. A prospective, observational evaluation of a new supraglottic airway: the PAXpress. Minerva Anestesiol
4. Ahmed SM, Maroof M, Khan RM, Singhal V, Rizvi KA. A comparison of the laryngeal mask airway and PAXpress
for short surgical procedures. Anaesthesia
5. Brimacombe J. Analysis of 1500 laryngeal mask uses by one anaesthetist in adults undergoing routine anaesthesia. Anaesthesia
6. Hardman JG, Aitkenhead AR. Estimating alveolar dead space from the arterial to end-tidal CO2
gradient: a modelling analysis. Anesth Analg
7. Casati A, Fanelli G, Torri G. Physiological deadspace/tidal volume ratio during face mask, laryngeal mask, and cuffed oropharyngeal airway in anesthetized adult patients. J Clin Anesth