To the Editor:—
We would like to comment on the article by Meissner et al. 1introducing a manual, Venturi-based respiration valve that assists transtracheal ventilation through a small lumen catheter, even in case of complete airway obstruction. Both in an in vitro setting and in an in vivo trial, the effect of this valve was studied. Undoubtedly, the use of assisted expiration to achieve an adequate minute volume through a small lumen catheter is a promising principle. But, as demonstrated in the calculation below, the results of this paper cannot be interpreted accurately, as the effective oxygen flow ranged from 27% above to 41% below the set flow rate. The systematic reason for these errors is an unreliable method of testing and measurement.
The pressure of a volume of 1,000 ml of an ideal gas is at sea level 1,000 mbar (excluding extreme weather conditions). If, at a set compliance of 100 ml/mbar, the pressure of a compressed volume of 1,000 ml is 1,010 mbar, the decompressed volume will then be 1,010 ml following Boyle's law (p x V = constant given isothermic conditions). The same initial volume of 1,000 ml, now at a compliance of 10 ml/mbar, will expand to 1,100 ml at atmospheric pressure. To insufflate an additional, compliance-related volume of 10 ml at a flow of 6 l/min will take 0.1 s. Thus the calculated inspiration time for 1,000 ml will be 10.1 s at a compliance of 100 ml/mbar, and 11 s at a compliance of 10 ml/mbar. Of course, at a flow of 12 l/min these times will be halved (5.05 and 5.5 s, respectively).
A recalculation of flows based on the reported inspiration times at compliances of 100, 30, and 10 ml/mbar and a resistance of 2 mbar/l-1/s-1 reveals flow errors ranging from 27% above to 41% below the set flow rate (table 1).
Given a properly working lung simulator, the resulting differences are probably caused by the use of a nonpressure-compensated flowmeter. Conventional, nonpressure-compensated flowmeters are not capable of maintaining a chosen flow against high(er) resistance, and are therefore very sensitive to a reduced compliance or any small bore tubing or cannula attached in-line. On the contrary, pressure-compensated flowmeters built into a pressurized housing, are calibrated to work at a pressure of several bars, and will therefore continue to deliver a chosen flow even against high resistance.2Unfortunately, no information regarding the flowmeter is provided in the article, but looking at the inspiration times of the in vitro tests, we presume that a nonpressure-compensated flowmeter was employed. If a nonpressure-compensated flowmeter was used in the in vivo trials as well, the effectively injected volume was smaller than the calculated volume. This might be an additional explanation for the reported hypoventilation of the animals.
A higher flow improves the efficacy of a Venturi device. Because of the acceptable flow error of just - 4.8% for the 12-gauge cannula at an oxygen flow of 12 l/min, a compliance of 100 ml/mbar, and a resistance of 2 mbar · l−1· s−1(which resembles a situation where in this study the efficacy of the device can be assessed best), the given data still allow for an estimation of the overall effect of the device. Only an increase of the minute volume by 700 ml was achieved under these conditions. This absolute increase of the minute volume shows that the efficacy of this Venturi-based respiration valve is far below the point where normoventilation in an adult patient can be achieved.
Nonpressure-compensated flowmeters are widely spread and will almost universally be used for oxygen therapy, but as explained above, the use of these flowmeters for scientific purposes can produce unreliable data. Therefore, when dealing with flow against higher resistances in experimental settings, it is mandatory to use a properly calibrated, pressure-compensated flowmeter.
*Maastricht University Medical Centre, Maastricht, The Netherlands. a.hamaekers@mumc.nl