Ethics approval
This study is a completely technical simulation with no participants. Thus, no ethical approval was required.
Experimental setup
We developed a y-adapter fitting to a modern ventilation device according to ISO 5356-1 (Figure 1 A). This adapter can be used both in the in- and the expiratory branch of the ventilation tubes. Accordingly, we developed flow limitation devices with a diameter of 2-9mm that could be inserted in the ventilator tubes (Figure 1B).
Both pieces consisted of carbon fibre (Multec Carbon, Multec, Illmensee, Germany), and were printed on a 3D printer (Multec Multirap M800, Multec, Illmensee, Germany). The average printing time was 4 hrs; the adapter could be disinfected and sterilised. For lung and airway simulation, we used a double test lung with different lung compliances and airway resistance that are adjustable in each test lung separately (dual Adult TTL â - Model 5600i, Michigan Instruments, Grand Rapids, MI). Each side was set to represent airway characteristics of a patient; in result the dual lung represented airway characteristics of two patients. For determining minute ventilation, we included a respirometer Ferraris, Hertford, UK) in the tubes connecting the ventilator with each of the separated test lungs simulating one patient.
We adjusted the y-adapter to a ventilator (Hamilton C6, Bonaduz, Switzerland), allowing parallel ventilation of both test lungs (Figure 2).
In a pilot setup, we used standard double-lumen ventilation tubes distally of the y-adapter enabling gas to flow back to the ventilator in both designs. The flow limiters were adjusted on side B directly behind the inspiratory y-adapter. (Figure 2A). In a second setup, we integrated the flow limiter in circuit B again via a standard double-lumen connection tube, allowing expiratory gas to flow back to the ventilator from lung B; to the double-lumen connection tube to test lung A, we adjusted distally the double-lumen tube a valve of an AMBU Mark V bag-valve device (Ambu Glostrup, Copenhagen, Denmark) thus directing expiratory gas flow directly into ambient air through an HME-filter, and not back to the ventilator via the double-lumen tube. As a result, only expiratory gas of test lung B flowed back to the ventilator. The flow sensor was adjusted to the double-lumen side on circuit B (Figure 2B). In a third setup, we connected test lung A with a standard double-lumen connection tube, allowing expiratory gas to flow from lung A back to the ventilator. The double-lumen connection tube was connected to test lung B; we adjusted distally the flow diameter and the expiratory valve of an AMBU Mark V bag-valve device (Ambu Glostrup, Copenhagen, Denmark) thus directing expiratory gas flow from lung B directly into ambient air. The flow sensor was adjusted to the standard double-lumen side on circuit A (Figure 2C).
Experimental procedure:
In total, six experiments were performed in the aforementioned setups; each experiment was performed for 1 min and repeated six times. In the first experiment (Figure 2A), both lungs were ventilated first in volume-controlled mode with a compliance of 30mL∙mbar-1 simulating moderate ARDS in both lungs (respiratory rate 12∙min-1, tidal volume 1200mL, pmax 20mbar, PEEP 5mbar, FiO2 0.21, I:E 1:2). The first measurement was without a flow limiter. Subsequently, flow limiters were integrated into the inspiratory tube connected to test lung B. The flow limiter diameter was decreased with each attempt, starting with 9mm and ending with 2mm. The second experiment was done in pressure-controlled mode in the same way (respiratory rate 12∙min-1, Pmax 20mbar, PEEP 5mbar, FiO2 0.21, I:E 1:2). In all pressure-controlled experiments, peak pressure was adjusted to achieve a tidal volume of 600mL in the unobstructed lung A. Since measurements were done with different compliances, this was adjusted for each experiment.
In the third experiment, the pressure-controlled experiment two in the second setup (Figure 2B) was repeated. The peak inspiratory pressure to achieve tidal volumes in test lung A of 600 mL was adjusted. Compliance of test lung A was 70mL∙mbar-1 simulating a healthy lung. The fourth experiment was pressure-controlled ventilation in analogy to experiment 2 and 3 in the third setup (Figure 2C) with a compliance of 70mL∙mbar-1 in both lungs. In the last two experiments, the last setup (Figure 2C) was also used and compared test lung A with 50 and 30 mL∙mbar-1 to 70mL∙mbar-1 in test lung B, respectively. Pressure for ventilation was always adjusted to achieve tidal volumes of 600mL.
Statistical analysis:
Statistical evaluation was performed with Sigmaplot 14 (Systat, San Jose, CA) using Student's t-test after Shapiro Wilk Analysis for normality and Equal Variance Test (Brown-Forsythe). If a normality test failed, the Mann-Whitney rank-sum test was applied. Data are given as mean ± standard deviation. P values <0.05 were considered to be significant.