A novel Venturi system to generate high flow with titratable FiO₂

Background and rationale

Noninvasive ventilatory support, based on free-flow continuous positive airway pressure (CPAP) delivered by helmet or face mask or high-flow nasal oxygen (HFNO), is increasingly used to treat patients with acute respiratory failure (1). During the recent Coronavirus-2019 (COVID-19) pandemic, the use of free-flow system massively spread also outside the intensive care units (2-5). The main advantage of free-flow systems is that an active mechanical ventilator is not required; a flow generator connected to the oxygen pipe or tank can feed a helmet or a face mask to provide CPAP through a positive end-expiratory pressure (PEEP) valve (6), or to nasal cannulas to provide HFNO (7). Flow generators can be either turbine-based or Venturi-based systems.

Venturi systems are based on the principle that a high-pressure gas (typically oxygen at 4 atm) is delivered through a small-bore nozzle, reaching a very high speed. Due to Bernoulli’s law, once the high-speed gas moves into a larger conduit, it generates a decrease of the lateral pressure, which becomes subatmospheric and entrains another gas (air) from the external atmosphere. In this way, starting from a pure flow of oxygen (Q_O2), it is possible to achieve a much higher flow of a mixture of gas (Q_tot), at a variable FiO₂. For each Venturi system, it is hence possible to calculate the flow of gas entrained (Q_entr) for a given Q_O2, where the ratio . For example, R_vent = 3 indicates that for each 1 L/min of oxygen flowing through the nozzle, three additional liters/minute of gas is entrained, resulting in a total flow of 4 L/min. The final flow (Q_tot) depends on Q_O2 and R_vent according to the following formula:

The FiO₂ of Q_tot (if the gas entrained is air) equals

which, combined with Equation [1] and simplified becomes:

Hence, if R_vent does not change, varying Q_O2 to the nozzle will not vary the FiO₂ of Q_tot, but only Q_tot itself.

R_vent depends mainly not only on the diameter and shape of the nozzle orifice but also on the pressure downstream, as positive pressure decreases the entrainment ability of Venturi.

From a clinical standpoint, this is relevant, since when delivering free-flow CPAP using a Venturi system to generate the flow, any PEEP increase will cause an FiO₂ increase, which might be misinterpreted as a patient’s response to PEEP. Finally R_vent tends to decrease for low Q_O2, due to the loss of entrainment at lower gas speeds. A typical behavior of a Venturi system is displayed in Figure 1.

Description of the system

We herein describe and characterize a novel Venturi system (EasyVee flowmeter, Levate, Bergamo, Italy), coupled with a variable area oxygen flowmeter, connected to a high-pressure (4 bar) O₂ source. The Venturi nozzle (with a diameter of 1.1 mm) is characterized by a nominal R_vent of approximately 4.5 when operating without downstream pressure. A novel solution is that the size of the aspiration port (open to ambient air) can be modulated, by means of a nonius, from full opening to complete closure (Fig. 2). The excursion progresses continuously, but six precalibrated marks are put in order to facilitate Q_entr (and hence FiO₂) titration with the use of a table or an app.

Because of the size of the nozzle, the maximum achievable flow from a standard 4-bar oxygen wall port equals 38-40 L/min, even if the flow meter is opened to full scale. Therefore, when the nonius is completely closed to deliver pure oxygen (FiO₂ 100%), the total flow might be inadequate to provide noninvasive CPAP support; a novel solution is the presence of a switch to bypass the Venturi nozzle, achieving up to 70 L/min oxygen flows.

Methods

In this bench study we evaluated the performance of the novel Venturi system, in terms of total flow generated and FiO₂ for different Q_O2, nonius position, and PEEP loads, to simulate different clinical scenarios. Total flow and FiO₂ were measured by VT Plus HF (Fluke Biomedical, USA). FiO₂ was measured additionally by PICK-O2 (Medizintechnik Juergen K. Kranz GmbH) and values were averaged to obtain the measured FiO₂ (FiO_2,meas). Theoretical FiO₂ (FiO_2,th) was calculated as:

and compared with FiO_2,meas to verify internal data consistency.

For each condition R_vent was calculated as:

Results

As expected FiO_2,meas and FiO_2,th were tightly correlated (Fig. 3).

Figure 4 shows the performance of the novel Venturi system for different Q_O2 and position of the nonius, for a PEEP value of 7.5 cmH₂O. As expected, progressively closing the nonius leads to a decrease of the air entrained by the Venturi (decreasing R_vent) and hence Q_tot (upper panel), allowing to achieve higher FiO₂ (lower panel), in the range between 35% and 80%.

As shown in Figure 5, the performance of the novel Venturi system, measured as capacity to entrain air (R_vent), remained rather stable over a wide range of PEEP, showing a sensible impact only for PEEP levels above 12.5 cmH₂O, particularly if a low Q_O2 is used (upper panel).

Discussion

In this report we described the pneumatic performance of the EasyVEE system, a Venturi flow generator equipped with an adjustable air entraining port and a single extended-range flowmeter. The novel system was very flexible, in generating high-flow gas mixture, with a simple control to titrate FiO₂ and a constant performance over a wide range of PEEP and Q_O2.

As mentioned, an adequate flow of fresh gas is necessary in order to efficiently deliver CPAP or HFNO. This is often achieved by Venturi systems, which drop the FiO₂ proportionally to their efficiency in gas entrainment. Standard Venturi systems are typically built with a fixed (not adjustable) port for air entrainment; therefore, a second flowmeter is needed to increase FiO₂ by addition of oxygen to the gas mixture generated by the Venturi. At variance from these, the EasyVEE foresees a progressive reduction of the air opening, so that the net result is an adjustable R_vent and an increase in FiO₂.

Another crucial point of the Venturi systems is the loss of performance for increasing back pressures, typically due to PEEP increase. In this respect, the Easy-VEE maintains a relatively constant performance over a wide range of PEEP levels, albeit the use of a direct FiO₂ measure or custom tables for the device is recommended to determine the exact FiO₂ after any therapeutic adjustment.

In conclusion the EasyVEE system appears to be a simple, flexible, and reliable solution to generate continuous flow for noninvasive respiratory support interfaces.

Disclosures

Conflict of interest: Both authors are coinventors on patents and receive consultancy fees from Flow-Meter SPA (Levate, BG, Italy) for a matter related to the topic of the article.

Financial support: None.