This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.
Abstract
BACKGROUND: The Oxylator is an automatic resuscitator, powered only by an oxygen cylinder with no electricity required, that could be used in acute respiratory failure in situations in which standard mechanical ventilation is not available or feasible. We aimed to assess the feasibility and safety of mechanical ventilation by using this automatic resuscitator in an animal model of ARDS.
METHODS: A randomized experimental study in a porcine ARDS model with 12 pigs randomized to the Oxylator group or the control group (6 per group) and ventilated for 4 h. In the Oxylator group, peak pressure was set at 20 cm H2O and PEEP was set at the lowest observed breathing frequency during a decremental PEEP titration. The control pigs were ventilated with a conventional ventilator by using protective settings and PEEP at the crossing point of collapse and overdistention, as indicated by electrical impedance tomography. Our end points were feasibility and safety as well as respiratory mechanics, gas exchange, and hemodynamics.
RESULTS: After lung injury, the mean ± SD respiratory system compliance and PaO2/FIO2 were 13 ± 2 mL/cm H2O and 61 ± 17 mm Hg, respectively. The mean ± SD total PEEP was 10 ± 2 cm H2O and 13 ± 2 cm H2O in the control and Oxylator groups, respectively (P = .046). The mean plateau pressure was kept to < 30 cm H2O in both groups. In the Oxylator group, the tidal volume was transiently > 8 mL/kg but was 6 ± 0.4 mL/kg at 4 h, whereas the breathing frequency increased from 38 ± 4 to 48 ± 3 breaths/min (P < .001). There was no difference in driving pressure, compliance, PaO2/FIO2, and pulmonary shunt between the groups. The mean ± SD PaCO2 was higher in the Oxylator group after 4 h, 74 ± 9 mm Hg versus 58 ± 6 mm Hg (P < .001). There were no differences in hemodynamics between the groups, including blood pressure and cardiac output.
CONCLUSIONS: Short-term mechanical ventilation by using an automatic resuscitator and a fixed pressure setting in an ARDS animal model was feasible and safe.
- artificial respiration
- mechanical ventilators
- respiratory insufficiency
- ventilator-induced lung injury
- respiratory distress syndrome
- rescue ventilation
Introduction
The incidence of ARDS has increased drastically in recent years with the COVID-19 pandemic,1–4 which highlights an emerging need for simple, safe, and low-cost mechanical ventilators to be applied in case of surge.5–8 The Oxylator (CPR Medical Devices, Markham, Ontario, Canada) is an automatic resuscitator cleared by the United States FDA, developed for prehospital emergencies. It has been recently proposed for use in acute respiratory failure as a temporary bridge in situations in which standard mechanical ventilation is not feasible. The device does not use electricity and is a hand-held device but does not provide any monitoring.9,10
In a previous bench testing and a short-term porcine lung injury model, the Oxylator delivered stable ventilation, with tidal volumes (VT) within a clinically acceptable range in lungs with low compliance, showing a breathing pattern dependent on the set pressure and compliance, which provided information on the effect of the settings on breathing frequency.10 However, this automatic resuscitator device has not been tested in an extended invasive mechanical ventilation in an ARDS model compared with conventional mechanical ventilation.
The main objectives of this study were to assess the feasibility and safety of mechanical ventilation when using an automatic resuscitator in a porcine ARDS model and to compare its effect with conventional mechanical ventilation on respiratory mechanics, gas exchange, and lung injury. We hypothesized that mechanical ventilation by using the Oxylator would be feasible and safe in terms of VT, and estimated driving pressure and plateau pressure in a situation comparable with human ARDS over several hours.
QUICK LOOK
Current Knowledge
The COVID-19 pandemic has highlighted an emerging need for simple, safe, and low-cost mechanical ventilators. The Oxylator is an FDA-cleared automatic resuscitator that has been recently proposed for use in acute respiratory failure as a temporary bridge in situations in which standard mechanical ventilation is not feasible.
What this Paper Contributes to Our Knowledge
Mechanical ventilation using an automatic resuscitator in a severe porcine ARDS model, was feasible and safe. The resuscitator was able to deliver a relatively protective mechanical ventilation in terms of VT, estimated plateau pressure, and driving pressure.
Methods
This was a randomized experimental study. Experiments followed the Canadian Animal Care guidelines, and the protocol was approved by the Animal Care Committee (reference AUP 46420) of the Peter Gilgan Centre for Research and Learning from The Hospital for Sick Children, Toronto, Canada. Twelve Yorkshire pigs (∼35 kg) were randomized between 2 groups: Oxylator or control (conventional mechanical ventilation) by allocating 6 pigs per group. All the pigs were anesthetized with akemezine and pentobarbital, and paralyzed with rocuronium (see the supplementary materials at http://www.rcjournal.com), and a pulmonary artery catheter thermodilution catheter and an arterial line were placed. An electrical impedance tomography (EIT) belt was placed around the thorax, just below the front legs.
Acute lung injury was established by using a 2-hit model, including surfactant depletion and high stretch ventilation, as previously described.11 Briefly, lungs were repeatedly lavaged (30 mL/kg warm 0.9% saline solution intratracheally) until the arterial oxygen level was < 100 mm Hg. Lungs were then ventilated by using conventional mechanical ventilation (Evita Infinity V500, Dräger, Lübeck, Germany) with PEEP of 1–13 cm H2O and peak pressure of 40 cm H2O for 30 min until reaching pre-specified levels of oxygenation and compliance.
The Oxylator
The Oxylator has been described elsewhere.10,12 Briefly, the device is a light weight (0.25 kg) portable resuscitator (diameter, 57 mm; length, 108 mm), designed for out-of-hospital use during cardiac arrest resuscitation, powered by a compressed source of gas (compatible with oxygen supply tanks and regulators that supplies between 45 and 80 psi inlet pressure). Breaths are pressure triggered and cycled, and can be patient-triggered or machine triggered. A constant inspiratory flow of ∼30 L/min is delivered on reaching a pre-set maximal inspiratory pressure.
This leads to passive exhalation until the expiratory pressure diminishes to a level of 2 to 4 cm H2O (auto-PEEP generated by the device's valve10) and no longer opposes the weight of a valve that is then magnetically pulled into the inspiration position, which triggers the next breath. A PEEP valve can be added in series, which alters the triggering pressure to 2 to 4 cm H2O above PEEP. The pressure cycling value above PEEP (a value that leads to the end of an inspiration cycle) is adjustable between 15 and 30 cm H2O and set by the user. Breathing frequency, inspiratory time, and VT are not directly adjusted by the user but are determined primarily by the patient's respiratory system compliance and airway resistance. The device has no monitoring system. An external gas blender can be used to control , and the oxygen consumption will vary depending on set
.
PEEP Titration
All pigs were submitted to 2 decremental PEEP titrations before randomization. The first one was performed on a conventional mechanical ventilator, on pressure-controlled ventilation, with a driving pressure of 15 cm H2O. PEEP was stepwise decreased, from 22 to 4 or 0 cm H2O (2 cm H2O/step, 30 s in each step). We recorded respiratory system compliance, percentage of lung collapse, and lung overdistention (measured by EIT) at each level of PEEP. In this PEEP titration, the best PEEP was defined as the crossing point between collapse and overdistention, as previously described.13 Before each PEEP titration, a recruitment maneuver was performed on pressure-controlled ventilation, with a driving pressure set at 15 cm H2O and PEEP at 15, 20, and 24 cm H2O (30–60 s in each step).
The second PEEP titration was performed by using the Oxylator, attached to a spring-loaded PEEP valve (AMBU S/A, Ballerup, Denmark) (Supplementary Fig. 1, see the supplementary materials at http://www.rcjournal.com). We set the peak pressure at 20 cm H2O, which gives a driving pressure of ∼15 cm H2O (anticipating some resistive pressure to be subtracted, as well as auto-PEEP). We started PEEP titration dialing at 15 cm H2O when considering that the device generates an “auto-PEEP” of 2–4 cm H2O and then progressively decreased dialed PEEP on the external spring-loaded PEEP valve to 12.5, 10, 7.5, 5, 2.5 and 0 cm H2O (30 s in each step). The total applied PEEP was derived from the airway pressure waveform.
We recorded the breathing frequency at each level of PEEP during the decremental PEEP titration on the Oxylator. The inspiratory and expiratory time, as well as the breathing frequency, are not pre-set on the Oxylator. The inspiratory time depends on how long it takes to reach the pressure limit during the constant flow insufflation, and the expiratory time depends on how long it takes to come back to the expiratory pressure. This is dictated by the time constant of the respiratory system that corresponds to the product of respiratory system compliance and airway resistance14,15; when assuming that, during PEEP titration, the change in PEEP does not result in a remarkable change in airway resistance, the highest compliance will result in the highest time constant and, therefore, the longest inspiratory and expiratory times, and the lowest breathing frequency.10 We used the observed breathing frequency to select the PEEP to achieve the highest compliance. After PEEP titration, we also tested the presence of airway closure pressure with a low-flow inflation and measured the recruitment-to-inflation ratio during a PEEP drop from 15 to 5 cm H2O (https://crec.coemv.ca)16 in all the pigs while they are on conventional mechanical ventilation.
Interventions
After randomization, the pigs were ventilated for 4 h in one of the groups. In the control group, the pigs were ventilated on a conventional mechanical ventilator by using volume-controlled ventilation with a VT of 6–8 mL/kg and a PEEP at the crossing point between lung collapse and overdistention (measured by EIT), and a breathing frequency that aimed for acceptable . In the Oxylator group, the pigs were ventilated with the Oxylator by using a set peak pressure at 20 cm H2O and a PEEP at the lowest breathing frequency obtained during PEEP titration on the Oxylator. In both groups,
was set and adjusted to maintain
between 92 and 95%. In the Oxylator,
was controlled with an external gas blender. During the experiment, if we observed an increase in breathing frequency > 20% in the Oxylator group or an increase in driving pressure > 20% in the control group, we performed a new decremental PEEP titration and changed PEEP if necessary. Continuous infusions of pentobarbital (15 mg/kg/h), for general anesthesia, and rocuronium (0.02 mg/kg/h), for neuromuscular blockage, were maintained during the entire experiment.
Measured Variables
We continuously monitored respiratory and hemodynamic variables, which were recorded hourly. During all measurements, in both groups, a pneumotachograph was placed between the endotracheal tube and the Oxylator for the Oxylator group or the ventilator circuit for the control group. Airway pressure and flow were acquired at 1 kHz in LabChart (ADInstruments, Sydney, Australia) and were used to measure and/or estimate respiratory mechanics variables offline. VT was derived by integration of the flow signal over time, and plateau pressure was measured by subtracting resistive pressure from the peak pressure (plateau pressure = peak pressure – [airway resistance × airway flow]). Airway resistance was measured on a conventional ventilator before randomization.
Total PEEP in the Oxylator group was estimated as the airway pressure at zero flow (applied PEEP plus dynamic intrinsic PEEP) by using the Maltais-Gottfried technique.17 To ensure that our measurements were accurate, we did an additional bench analysis on a 2-chamber Michigan test lung (Michigan Instruments, Grand Rapids, Michigan) and the difference between the airway pressure and the pressure inside the chambers (alveolar pressure) at zero flow, in a scenario without airway obstruction, is 0.8 ± 0.5 cm H2O (see the supplementary materials at http://www.rcjournal.com).
EIT data were acquired by using the electrical impedance tomograph PulmoVista 500 (Dräger Medical GmbH, Lübeck, Germany), which produces 50 images/s when using 16 electrodes. The image pixels changes represent the impedance changes and are proportional to changes in the lung volume.18,19 We used EIT data to calculate the total and regional (dependent and non-dependent) ventilation distribution to estimate changes in end-expiratory lung volume after 2 and 4 hours (compared with the beginning of the intervention, time 0) and to estimate the amount of lung collapse and lung overdistention in each level of PEEP during a decremental PEEP titration.
We also collected arterial and mixed venous blood samples and measured cardiac output every 2 h. Pulmonary shunt was automatically calculated by using a blood gas analyzer (ABL800, Radiometer Medical ApS, Copenhagen, Denmark).20 Arterial blood samples were used for blood gas and cytokines analyses. After euthanizing the pig, 2 portions of the lung (dependent and non-dependent regions) were taken for wet-to-dry weight measurement.
Statistical Analyses
Categorical variables were presented as proportions, and continuous variables were reported as mean ± SD or as median and interquartile range, according to their distribution. We used the Shapiro-Wilk test to assess if continuous variables followed normal distribution. Comparisons of continuous data between the groups were done by using the t test or Mann-Whitney test and mixed analysis of variance for repeated measures. We used the Bonferroni test for post hoc analysis. Comparisons of categorical data were done by using the chi-square test or Fisher exact test. The statistical analyses were performed by the software R Programming (cran.r-project.org) and were considered statistically significant if P < .05.
Results
There was no difference with regard to respiratory system compliance and /
after lung injury between the groups at the start of the ventilation period (Table 1). Both groups were considered recruitable, but the pigs randomized to the Oxylator group presented a lower recruitment-to-inflation ratio after lung injury (Table 1). The main ventilatory variables are shown in Figure 1. All the pigs in both groups (no. = 12) survived the 4-h end point. Plateau pressure was slightly lower in the control group at the beginning of the intervention (time zero), and PEEP total was ∼2 cm H2O higher in the Oxylator group in the first 3 h of intervention. All pressures, including plateau, peak, and total PEEP, were similar between the 2 groups after 4 h of mechanical ventilation.
Ventilatory variables. Data are presented as mean ± SD. *Significant values and/or differences (P < .05) for between-group comparisons at a specific time point.
Pig Characteristics Before and After Lung Injury
The mean plateau pressure was kept < 30 cm H2O during the entire period of mechanical ventilation in both groups, and it was feasible to deliver a relative lung-protective strategy when using the automatic resuscitator. The mean driving pressure was ∼15 cm H2O in the Oxylator group, and it was not different between the groups after 4 h. The mean VT per body weight was always < 8 mL/kg in the control group. It was only transiently > 8 mL/kg in the Oxylator group at the beginning of the intervention.
Respiratory system compliance decreased over time in the Oxylator group, but it was not different from the control group after time zero. Because of the working principle of the Oxylator, the decrease in compliance resulted in increased breathing frequency over time and also a decrease in VT and minute ventilation over time but mainly in the first hour. However, these variables were not different from the control group after 4 h of ventilation. There was also no difference with regard to regional distribution of ventilation, measured by EIT (Supplementary Table 2, see the supplementary materials at http://www.rcjournal.com). Dorsal ventilation after 4 h was 57 ± 7% in the Oxylator group and 51 ± 12% in the control group (P = .34). End-expiratory lung volume decreased over time, but it was not different between the groups (Supplementary Table 2, see the supplementary materials at http://www.rcjournal.com). The change in end-expiratory lung volume after 4 h was –0.4 ± 0.2 L in the Oxylator group and –0.3 ± 0.4 L in the control group (P = .31).
In each group, 3 pigs (50% per group) needed an additional PEEP titration based on the pre-defined criteria. The main difference in PEEP after a second PEEP titration in the animals that needed that intervention was –1 ± 2 cm H2O in the Oxylator group and 1 ± 3 cm H2O in the control group (P > .99). Hemodynamic variables are summarized in Table 2. Cardiac output decreased over time in both groups and did not differ between the groups. We did not observe differences with regard to heart rate, arterial blood pressure, central venous pressure, and mean pulmonary artery pressure. Pulmonary vascular resistance increased over time in each group, but it was not different between both groups.
Hemodynamics
As planned, was > 92% during the entire ventilation period, and there was no difference between both groups. There was also no difference with regard to
/
and pulmonary shunt (Table 3).
decreased over time in the control group, whereas it slightly increased in the Oxylator group after 2 h of ventilation. Arterial pH increased over time in each group, with no difference between the groups after 4 h (Table 3).
Gas Exchange
A: Comparison of PEEP titration by using the resuscitator versus conventional ventilator. The black line represents the breathing frequency, according to total PEEP, during PEEP titration when using the resuscitator, and the gray line represents respiratory system compliance (CRS) during PEEP titration on conventional ventilation. Data are presented as mean ± SD. B: Correlation, and C: A Bland-Altman plot of optimal PEEP by breathing frequency on the resuscitator and optimal PEEP by CRS on conventional ventilator. D: Correlation, and E: A Bland-Altman plot of optimal PEEP by breathing frequency on the resuscitator and optimal PEEP by electrical impedance tomography (EIT) on conventional ventilator, based on lung collapse and lung overdistention.
The wet-to-dry ratio in the dependent and non-dependent parts of the lungs was used as an index of pulmonary edema formation (Supplementary Fig. 4, see the supplementary materials at http://www.rcjournal.com). In the dependent region, the wet-to-dry ratio was 8.2 ± 1.2 for the control group versus 9.3 ± 0.8 for the resuscitator group (P = .11). In the non-dependent region, the mean ± SD wet-to-dry ratio was 7.5 ± 1.0 for the control group versus 7.9 ± 0.7 for the Oxylator group (P = .50). We did not observe differences with regard to plasma cytokine concentration between the groups, either at the start of the intervention (time zero) or after 4 h (Supplementary Table 2, see the supplementary materials at http://www.rcjournal.com).
We performed a PEEP titration both on conventional ventilation (measuring respiratory system compliance, lung collapse, and lung overdistention with EIT) and on ventilation with the automatic resuscitator in all the pigs before randomization. Setting PEEP based on the breathing frequency with the Oxylator was feasible and reliable. The lowest breathing frequency during PEEP titration on the automated resuscitator corresponded to the best respiratory system compliance during the PEEP titration on conventional mechanical ventilation, as shown in Figure 2A. We compared the optimal PEEP estimated by the Oxylator and the optimal PEEP estimated by EIT (based on relative overdistention and collapse) at the beginning of the experiments, and we observed a moderate correlation (Fig. 2 D and E). The Oxylator tended to give a slightly higher PEEP level, by 2 cm H2O.
Discussion
We demonstrate that mechanical ventilation by using the resuscitator, in a porcine model of severe ARDS, was feasible and safe over several hours. In the pigs ventilated with the automatic resuscitator, ventilation was kept safe in terms of VT and estimated plateau pressure and driving pressure. Respiratory mechanics was stable, and VT and breathing frequency varied according to respiratory system compliance. Gas exchange, lung injury and hemodynamics were not different from the pigs ventilated on a conventional ventilator.
Peak pressure, plateau pressure, and total PEEP were stable during the entire experiment, whereas the VT and breathing frequency varied over time in the Oxylator group. However, minute ventilation was stable and similar to the control group after 4 h. The observed variability in VT and breathing frequency was expected when considering the device characteristics of constant flow and pressure cycling. As shown in a previous bench and animal study,10 the maximum airway pressure when using the Oxylator is reached after an inspiratory time that depends on compliance. Thus, based on the time constant concept,14 the inspiratory time (at any particular setting of maximum pressure and PEEP) will be passive consequence of respiratory system compliance and airway resistance, and it will vary together with VT (same direction) and breathing frequency (opposite direction).
The total PEEP was slightly higher in the Oxylator group (∼3 cm H2O) most of the time. This is expected when considering that the Oxylator generates an intrinsic PEEP of 2–4 cm H2O. It is important to highlight that the Oxylator does not allow the user to perform an expiratory hold, so, even with an external pneumotachograph, the direct measurement of static intrinsic PEEP is not possible. As described in our methods section, for this study propose, we estimated the total PEEP as the airway pressure at zero flow. We considered that dynamic intrinsic PEEP is similar to static intrinsic PEEP in scenarios without airway obstruction, as observed in previous studies.17,21
The Oxylator was able to deliver a relatively protective mechanical ventilation. Numerous strategies have been proposed to ensure protective mechanical ventilation to reduce ventilator-induced lung injury, especially in patients with ARDS. These strategies include a plateau pressure < 30 cm H2O,22 VT of 6–8 mL/kg,23 and a driving pressure limited to 15 cm H2O.24 In the Oxylator group, setting a maximum inspiratory pressure at 20 cm H2O kept plateau pressure and VT within recommended values with a driving pressure of ∼15 cm H2O. It is possible to reduce the maximum inspiratory pressure on the Oxylator and, consequently, the driving pressure. However, it would result in a VT < 6 mL/kg, which is very low for the pigs when considering the severity of the lung injury model.
We assessed the hemodynamics associated with the resuscitator over a long period in an acute ARDS model. Mechanical ventilation by using the automatic resuscitator did not result in hemodynamic abnormalities, and there was no difference compared with conventional mechanical ventilation. In a previous study during cardiac arrest and chest compressions, ventilation by using the Oxylator did not result in hemodynamic complications.25 We did not observe a difference in /
between the Oxylator group and the control group. Because
/
is an important variable to define the level of lung injury,4 analysis of our results suggests that ventilation with the Oxylator did not increase lung injury. This is also supported by the absence of differences in blood cytokines and lung wet-to-dry ratio.
Previous study reported no significant changes in end-tidal CO2 levels when using the Oxylator in a model of cardiac arrest for short periods.25 In our study, increased over time in the Oxylator group. This may be explained by the reduction in minute ventilation over time in the Oxylator group, which was expected when considering that respiratory system compliance decreased over time. However, arterial pH was not different from conventional mechanical ventilation. This may be acceptable, at least for relatively limited periods.26
As suggested in a previous study,10 our results support the idea of performing PEEP titration based on breathing frequency. One of the most common ways to titrate PEEP in clinical practice is to select the best PEEP that results in the best respiratory system compliance. We observed that the level of PEEP at the lowest breathing frequency on the Oxylator (a procedure that requires 1–2 min and no added equipment) was very similar to the level of PEEP at the best compliance obtained with conventional mechanical ventilation.
Supporting the reliability of PEEP titration with the Oxylator, we also observed a moderate correlation between the Oxylator established optimal PEEP and the optimal PEEP estimated by EIT. The automated resuscitator tended to give a slightly higher PEEP level of 2 cm H2O. EIT is a sophisticated approach to individualize PEEP,18 based on regional lung collapse and lung overdistention, and previous studies show that the EIT-optimal PEEP may be slightly different from the optimal PEEP at the best compliance.13,27
Our study has several limitations. First, we tested the Oxylator for only 4 h, and we cannot extrapolate our results to longer durations of ventilation. Second, our porcine ARDS model was recruitable (recruitment-to-inflation ratio > 0.50) in all the pigs. It may be useful to investigate the use of the device in a non-recruitable ARDS scenario. Third, because all the pigs were paralyzed during the entire protocol, our results represent the use of the device only during controlled mechanical ventilation. Fourth, although ventilation with the Oxylator may be predictable, we highlight that it does not have alarms as do most traditional mechanical ventilators and therefore requires vigilance to ensure patient safety, with breathing frequency and oxygenation being the basic variables to monitor. If available, additional monitoring tools, such as a proximal flow or end-tidal CO2 sensors, can be used to improve patient safety.
Conclusions
Mechanical ventilation by using an automatic resuscitator in a recruitable severe ARDS animal model was feasible and safe in terms of VT, estimated plateau pressure, and driving pressure. We suggest that the resuscitator could be used to ventilate patients as a bridge to more adjustable ventilation for short periods as long as there is continuous monitoring by a professional or in constrained or limited recourse scenarios.
Footnotes
- Correspondence: Laurent J Brochard MD, Keenan Research Centre - St. Michael's Hospital, Unity Health Toronto, Li Ka Shing Knowledge Institute, 4th Floor, Room 411, 209 Victoria Street, Toronto, ON, M5B 1T8 Canada. E-mail: laurent.brochard{at}unityhealth.to
The authors have disclosed no conflicts of interest.
This work was supported by a grant from the University of Toronto (Toronto COVID-19 Action Initiative – Connaught/ISI (Dr Dorian), a MITACS Accelerate Training Grant (Dr Dorian), and Canadian Institutes of Health Research (PJT 156336 to Drs Post and Brochard).
Supplementary material related to this paper is available at http://www.rcjournal.com.
- Copyright © 2023 by Daedalus Enterprises
REFERENCES
Pay Per Article - You may access this article (from the computer you are currently using) for 1 day for US$30.00
Regain Access - You can regain access to a recent Pay per Article purchase if your access period has not yet expired.