This is a lung model study, in which the authors compared the pressure transmission and CO2 elimination effects of nasal IPPV with 2 different interfaces: the neotech RAM cannula and Fisher-Paykell short binasal prongs.
The model they used was a neonatal intubation model, with a soother in the mouth to reduce oral leaks, and they measured the pressures in the distal “trachea”. A Draeger VN500 was used to deliver ventilation and the pressures were varied with a PEEP of 5 and inspiratory pressures from 14 to 34. They measured CO2 removal from the lungs by filling the balloons (“lungs”) with CO2 and then measuring how much was washed out by the nasal ventilation. They also measured the resistance of the interfaces.
One of the things that bugs me greatly, and there are many, is that a great deal of medical equipment has never been appropriately evaluated. For example, the neotech RAM cannula was introduced, and is sold, without, as far as I am aware, a single study evaluating whether pressure transmission actually occurs in newborn infants.
This study suggests that the resistance of the RAM particular system is far too high, and there is very little pressure transmission; although they don’t report on the CPAP generated in the pharynx, if any. The ventilator will of course continue to show good pressures, as the ventilator measures what is on the ventilator circuit side of the cannulae. This study shows that when you measure the pressures actually getting to the lung during simulated nasal ventilation, the pressures are much lower than you might think. Especially with the RAM cannula. When the set pip was increased from 14 to 34 cmH2O the ventilator measures pressures (in the circuit) increasing from about 12 to about 25, but the pressure in the lung goes from just under 0.5 cmH2O to just over 0.5 cmH2O.
With the Fisher Paykell prongs the ventilator measures circuit pressures which are significantly lower, due to the much lower resistance of the prongs, but much greater pressures are transmitted to the lung.
With both sets of prongs, the tidal volumes obtained were very small, but were much higher with the Fisher Paykell prongs. Also with both sets of prongs there was some CO2 washed out of the “lungs” despite the very small tidal volumes. The authors suggest that this means that maybe there are “non-conventional gas exchange mechanisms” I suggest that it shows the limitations of this kind of model, the findings are interesting, but you certainly can’t start talking about gas exchange, when you are washing CO2 out of a balloon. In the discussion the authors state:
Several factors have been suggested as determining the NIPPV-dependent gas exchange. These include an increase in tidal ventilation, improved recruitment of alveoli, a washout effect of anatomical dead space and increasing stimulation of the respiratory center. The fact that effective CO2 clearance could be demonstrated in this study despite a small fraction of delivered VT leads credence to the importance of non-VT-dependent alternate factors postulated by others.
Now increased tidal ventilation, recruitment of alveoli and stimulation of the respiratory center are clearly irrelevant in a patient that looks like this: 
This patient doesn’t have any tidal ventilation, respiratory drive or alveoli! The only one of those mechanisms which might be important is washing out the dead space.
The study clearly suggests that there is a very limited place for the neotech RAM cannulae, at least for nasal ventilation. It also suggests that the resistance of cannulae should be printed on the packaging, so that we know that what we are buying is fit for purpose.
Neonatal Research 