Lung model, ventilator and pressure measurements
A commercial test model (Adult/ Pediatric Demonstration Lung Model, IngMar Medical, Pittsburgh, PA, USA) was modified to mimic the low FRC and lung compliance of most ARDS patients and provide means for measuring airspace pressure distal to the ETT during a systematic test protocol. The model was connected to an air chamber (Fig. 1a) to obtain an end-expiratory gas volume of 1400 ml, comparable to the Functional Residual Capacity (FRC) reduction found in mechanically ventilated patients with secondary lung disorder [7]. An extra weight of approximately 750 grams was placed on the bellows to simulate the reduction in compliance common in patients with acute respiratory failure (Fig. 1b).
Airspace pressures distal to the ETT (simulating airway pressures distal to an ETT in positive pressure ventilated patients and referred to as “model airways” (MA) in figures) were measured by a rapid response pressure transducer (baud rate 115200/sec) imbedded in a 5 Fr plastic tube (Reggie, Camtech AS, Høvik – Norway) inserted through an air-tight entrance port (Fig. 1c). The transducer was connected to a dedicated computer (Fig. 1d) which displayed and saved real-time pressure changes. Representative tracings from the experiments are depicted in Fig. 2.
During mechanical ventilation, the model was connected to a commonly used ICU ventilator (Servo-i, Maquet, Solna - Sweden) with ETTs of either 9 mm, 8 mm or 7 mm internal diameter (ID) (Mallinckrodt, Hazelwood, Missouri - USA) cut to lengths of 26, 25 and 24 cm, respectively, to simulate clinically relevant ETT lengths in accordance with practise in our hospital (Fig. 1e). Pressures proximal to the ETT (i.e. those detected by transducers monitoring pressures in the ventilator circuit), as well as other ventilator data, were recorded using a commercial filing system (Servo Tracker software version 4.0, Maquet, Solna - Sweden) (Fig. 1f). In addition, a video camera recorded real-time curves and parameters displayed on the ventilator screen.
Suction devices and flow rates
Bronchoscopic suctioning was performed using a 16 Fr bronchoscope (Olympus LF-TP, Tokyo - Japan) with a suction channel diameter of 2.6 mm. Catheter suctioning was performed using closed system catheters with an outer diameter of 12 Fr in 7 mm ID ETT and 14 Fr in 8 and 9 mm ID ETT, as in clinical practice.
Catheters and bronchoscopes were connected to an AGA MS-32 ejector suction device (AGA, Espoo, Finland) (Fig. 1g) with a vacuum gauge (WIKA Instrument Corporation, Georgia, USA) connected to a suction liner system (Serres Hospital Products, Kauhajoki, Finland). The suction equipment was driven by the hospital compressed air system and generated a negative pressure of −765 cm H2O (−75 kPa) (checked against a water column) when set to maximum. Experimental data to support an appropriate maximum level are lacking in the literature. Based on clinical practice, both the maximum level and a moderate level of −382 cm H2O (−37.5 kPa) were used in our investigation.
Flow rates through the different suction devices were measured by a spirometer (Vmax 22, Viasys Inc. Yorba Linda, CA, USA). The flow rate through the bronchoscope suction channel was 8.8 l/min at −382 cm H2O (−37.5 kPa) and 14.1 l/min at −765 cm H2O (−75 kPa). In 12 Fr catheters the flow rate was 9.6 and 15 l/min, respectively, and in 14 Fr catheters 9.6 and 17 l/min. These flow rates correspond with measurements in other experimental studies [3, 8].
Experimental procedures
The effect of variations in endotracheal tube size (9 mm, 8 mm and 7 mm ID), I:E ratio (1:3, 1:2, 1:1, 2:1, 3:1, 4:1) and flow (F)- or pressure (P) trigger settings (F/5, 0, P-5 cm H2O, P-10 cm H2O, P-15 cm H2O, P-20 cm H2O) were investigated with the ventilator set in either pressure control (PCV) or volume control (VCV) mode, using both a bronchoscope and a closed catheter suction system with two suction levels (864 permutations in total). Changes in ventilator tidal volume, circuit pressure and model airway pressure distal to the ETT (PMA), were measured for 30 seconds; i) before insertion of a suction catheter/bronchoscope, ii) after insertion but before suctioning, iii) during suctioning with the end of the catheter/bronchoscope positioned 5 cm distal to the ETT, and iv) after removal of the suctioning device; as depicted in Fig. 2.
For experiments with PCV, the ventilator settings were; PPEAK 20 cm H2O, PEEP 5 cm H2O, ventilator frequency 15/min and RT 5 %. During VCV, initial settings were; inspiratory tidal volume (Vt) 500 ml, PEEP 5 cm H2O, ventilator frequency 15/min, pressure rise time (RT) 5 % and pause 0 %, giving a PPEAK of 18–22 cm H2O. The model compliance, calculated for both modes, was approximately 39 mL/cm H2O. Alarms and cutoff levels for high pressures were set to maximum. For the purpose of this investigation, the level of flow trigging most commonly used in our hospital (bias flow decreased by 1 l/min, called F/5) and five other levels of pressure trigging were used (see above). Changes in ventilator rate and tidal volume were recorded continuously. Mean model airway pressure (PMEAN MA) was calculated from the tracings.
Statistics and data management
Pressures and tidal volumes measured before and after scope/catheter insertion at each given mode, I:E ratio and ETT dimension (n = 12) were analyzed using the statistical package SPSS 15.0 for Windows (SPSS Inc., Chicago, IL, USA). Wilcoxon signed rank test was used to assess possible differences between paired values. Null hypothesis were rejected if two-tailed p-value was < 0.05.
Pressures and volumes measured during suctioning were strongly influenced by the choice of ventilator trigger sensitivity and suction level (due to high frequency ventilator trigging) and therefore unique for each of the 12 permutations.