Article Text
PDF
Abstract
Introduction Administering supplemental oxygen is a standard of care for trauma casualties to minimise the deleterious effects of hypoxaemia. Forward deployment of oxygen using pressurised cylinders is challenging, for example, logistics (weight and finite resource) and environmental risk (fire and explosion). Oxygen concentrators may overcome these challenges. Although previous studies successfully demonstrated fractional inspired oxygen (FiO2) >0.8 using oxygen concentrators and ventilators, the systems did not fulfil the size, weight and power requirements of agile military medical units. This study evaluated whether a modular system of commercially available clinical devices could supply high FiO2 to either ventilated or spontaneously breathing casualties.
Methods As a proof of principle, we configured an Inogen One G5 oxygen concentrator, Ventway Sparrow ventilator and Wenoll rebreather system to ventilate a simulated lung (tidal volume 500 mL). Casualty oxygen consumption (gas withdrawal inspiratory limb) and carbon dioxide (CO2) production (CO2 added expiratory limb) were simulated (respiratory quotient of 0.7–0.8). Three circuit configurations were evaluated: open (supplementary oxygen introduced into air inlet of ventilator); semiclosed (ventilator replaces rebreather bag of Wenoll, oxygen connected to either ventilator or Wenoll); and semiclosed with reservoir tubing (addition of ‘deadspace’ tube between ventilator patient circuit and Wenoll). Data presented as mean and 95% reference range.
Results There were modest increases in FiO2 with increasing Inogen settings in ‘open’ configuration 0.23 (0.23–0.24) and 0.30 (0.28–0.32) (Inogen output 420 and 1260 mL/min, respectively). With the ‘semiclosed’ configuration and oxygen added directly into rebreather circuit, FiO2 increased to 0.36 (0.36–0.37). The addition of the ‘reservoir tubing’ elevated FiO2 to 0.78 (0.71–0.85). FiO2 remained stable over a 4-hour evaluation period. Fractional inspired carbon dioxide CO2 increased over time, reaching 0.005 after 170 (157–182) min.
Conclusion Combining existing lightweight devices can deliver high (>0.8) FiO2 and offers a potential solution for the forward deployment of oxygen without needing pressurised cylinders.
- ACCIDENT & EMERGENCY MEDICINE
- TRAUMA MANAGEMENT
- Physiology
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information.
This material is licensed under the terms of the Open Government Licence except where otherwise stated. To view this licence, visit https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/ or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: psi@nationalarchives.gov.uk
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Early provision of oxygen to prevent hypoxaemia improves outcomes in trauma patients.
Forward deployment of oxygen using pressurised systems is challenging, particularly for unsupported military medical and surgical teams.
Oxygen concentrators offer a possible solution but to date, no portable lightweight system capable of delivering high ≥0.8 fractional inspired oxygen (FiO2) has been reported in the literature.
WHAT THIS STUDY ADDS
This proof-of-principle laboratory study demonstrated that FiO2 of 0.8 is achievable with a modular system comprising commercially available products: a small, lightweight concentrator, a rebreather circuit and a portable ventilator suitable for both ventilated and spontaneously breathing casualties.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This modular batter-powered system, weighing in total 5.1 kg, could provide forward medical and surgical teams a safe supply of oxygen that can be adjusted to their patient’s needs.
Introduction
Early provision of oxygen (O2), administered to treat hypoxaemia, is a clinical standard of care for trauma casualties.1 Some argue that ‘O2 should be available to all trauma and medical patients in the forward operating environment’ in military settings2 and titrated to a target peripheral capillary blood saturation (SpO2) of 92% (military casualties2) or 92–96% (civilian casualties3). Supplementary O2 may benefit 48% of military trauma patients2 and up to 80% of chemical, biological, radiological or nuclear casualties.4
Efforts are made to avoid hypoxaemia in trauma patients because it increases mortality5 and morbidity,6 particularly when combined with severe traumatic brain injury.7 8 Early administration of O2 reduced mortality in trauma patients9 and significantly improved survival time in an experimental study of blast lung injury and prolonged hypotensive resuscitation.10
Forward deployment and administration of O2 in a military setting are challenging. O2 cylinders are problematic: they are heavy, pressurised, and pose both a logistical burden and inherent risk in a ballistic or constrained environment. O2 concentrators may provide an alternative source of medical-grade O2 as ‘O2-enriched air’ containing approximately 90% O2.11 12
Bordes et al13 assessed the feasibility of using an O2 concentrator in conjunction with a lightweight turbine ventilator in a military setting. They showed that it was possible to deliver a fractional inspired O2 (FiO2) in excess of 0.8 at a clinically relevant ventilator minute volume of 6 L/min. The required O2 flow output, however, would need a concentrator weighing approximately 25 kg that generally relies on an external power supply.11 A smaller, battery-powered concentrator weighing 5.5 kg provided FiO2 of 0.5–0.6 in conjunction with a range of ventilators.14 15 Further reduction of the weight burden and improved battery run times necessitate smaller concentrators with outflow rates in the approximate range of 0.8–1.5 L/min. To improve efficiency, these small portable concentrators, designed to be used in open circuit, deliver pulses of O2-rich gas synchronised with the patient’s inspiratory effort,16 17 providing an effective FiO2 of 0.25–0.3.16
Although this may be sufficient for some trauma casualties, other clinical situations demand high FiO2, for both spontaneously breathing and ventilated casualties. This is challenging for a variety of deployed military medical teams, constrained by equipment size and weight. Special forces medical teams need to be flexible, lightweight, adaptable and self-contained.15 Conventional force surgical teams operate at reach, in the congested, cluttered, contested, connected and constrained modern and future battlespaces. Therefore, the ability to operate unsupported, hold or transport casualties for extended periods of time and over long distances is essential. These requirements may also extend to some civilian and humanitarian situations.
This study aimed to demonstrate, as a proof of principle, that a lightweight modular system of commercially available clinical devices was capable of providing FiO2 in excess of 0.8 over periods relevant for military evacuation, without a need for external power supply or pressurised gas cylinders.
Methods
This study used an inanimate system to simulate the consequences of relevant aspects of respiratory activity in a casualty. As a proof of principle, we used a portable commercial O2 concentrator (Inogen One G5, Inogen, California, USA), small lightweight clinical ventilator (Ventway Sparrow VWSP-900, Inovytec Medical Solutions, Israel) combined with a medical semiclosed rebreather circuit (Wenoll-System, Emergency Oxygen System, Germany) (figure 1). A rebreather circuit (figure 1), using at least one one-way valve, recirculates the patient’s exhaled breath for subsequent inhalation, minimising the loss of O2. Expired carbon dioxide (CO2) is chemically removed during passage through a CO2 absorber (‘scrubber’) that is part of the circuit, and O2 to replace that consumed by the patient is added from an external source (usually a pressurised cylinder).
Main (left): schematic diagram indicating the Dstl test circuit. Upper inset (right) showing the original, commercially available Wenoll circuit used with oxygen from a pressurised cylinder for spontaneously breathing patients. Lower inset (right) schematic diagram of Sparrow ventilator patient connector tubing (termed ‘patient circuit’ by the manufacturer) between ventilator and Wenoll rebreather circuit. This whole figure was created with BioRender.com. CO2, carbon dioxide; EV, Sparrow exhalation valve (for more details, see 20); Exp, expiratory limb of the Wenoll rebreather circuit; Insp, inspiratory limb of the Wenoll rebreather circuit; O2, oxygen; PEEP, positive end-expiratory pressure; Vent, ventilator.
Test circuits
The circuit ventilated a simulated lung with recoil characteristics representative of a human lung (maximum volume 1000 mL; resistance 20 mbar/L/s±15% at 100 L/min; compliance 25 mL/mbar±15% at tidal volume=500 mL and positive end-expiratory pressure=0 mbar, Dräger SelfTestLung, Dräger Medical, Germany). Gas was withdrawn from the inspiratory limb of the circuit to simulate a casualty’s O2 consumption. CO2 was added to the expiratory limb to simulate metabolic production of the gas (figure 1). The rate of gas withdrawal (440–500 mL/min) was greater than the rate of CO2 addition (310–350 mL/min), simulating a casualty with a respiratory quotient of 0.7–0.8. The rate of O2 withdrawal was set to be representative of the resting O2 consumption of a large adult human (350 mL/min).
For analysis, gas was continuously sampled from the inspiratory limb, close to the simulated patient (A, figure 1), dried by passage through a desiccator tube, passed through an O2/CO2 analyser (ML206 Respiratory Gas Analyzer, ADInstruments, UK) and returned into the rebreather circuit. Data were recorded continually and displayed using a computerised data acquisition system (PowerLab 16/35 hardware and Chart 7 software for Windows, ADInstruments, UK).
Configurations
Open circuit
Ambient air was drawn into the ventilator and ‘exhaled’ air was vented into the atmosphere. Supplementary oxygen from the Inogen was introduced into the air inlet of the ventilator using a low-pressure O2 enrichment adapter (VWSP-304, Inovytec Medical Solutions, Israel) in accordance with manufacturers’ instructions. The concentrator reliably synchronised pulses of O2 at the respiratory rate (RR) set on the ventilator.
Semiclosed rebreather circuit
The ventilator was connected to the rebreather circuit in a lateral position via the relevant port on the side of the CO2 absorber (replacing the rebreather bag of the Wenoll circuit, figure 1). An adjustable pressure limiting (APL) valve was attached to the rebreather circuit overflow valve as described in the Wenoll user manual18 for conducting assisted ventilation. Neither the rebreather circuit nor the ventilator was modified beyond the scope of the relevant user manuals, with the exception that mechanical ventilation was now performed via a clinical ventilator rather than manual squeezing of a bag (from the perspective of the rebreather circuit), and that the patient connector of the ventilator was connected to a rebreather circuit rather than directly to the endotracheal tube of the patient (from the perspective of the ventilator).
When used with the rebreather circuit, the concentrator O2 outlet tubing could be connected to either the Sparrow low-pressure adapter tubing, or to the Wenoll O2 inlet port on the outflow side of the CO2 absorber. In the latter case, the concentrator reverted to ‘adaptive pulse mode’, pulsing asynchronously at a rate of 17 breaths/min.19 The overall outflow rate on the concentrator was unaffected by this. Concentrator settings 2, 4 and 6 were used (corresponding to concentrator outflow rates of 90% O2 at 420, 840 and 1260 mL/min, respectively).
Impact of reservoir tubing between the ventilator and the rebreather circuit
To maximise the potential impact of the rebreather circuit when used with the ventilator, a length of tubing (22 mm Flextube, O2 therapy tubing, Intersurgical, UK) was added between the Sparrow ventilator patient connector and the rebreather circuit (figure 2A,B). The resulting deadspace was 626 mL, which was intentionally greater than the tidal volume (500 mL) of the ventilator.
(A) Schematic diagram indicating modified ventilator/rebreather circuit configuration by insertion of reservoir tubing (orange) between ventilator connector tubing and the rebreather circuit. Created with BioRender.com. (B) Photograph of modified configuration incorporating the reservoir tubing. CO2, carbon dioxide; EV, exhalation valve; Exp, expiratory limb of the Wenoll rebreather circuit; Insp, inspiratory limb of the Wenoll rebreather circuit; O2, oxygen; Vent, ventilator.
Statistical analysis
All data were assessed for normality and subjected to transformation if necessary. Data with repeated measures over time were analysed by linear mixed-model analysis of variance using the R Statistics Package (R Studio V.3.6.3, Boston, Massachusetts, USA) using the values stabilised at 15 min as the baseline covariate. Planned comparisons between successive time points and the baseline (15 min) were made as indicated and adjusted for multiple comparisons (Tukey). Data recorded at time 0 min were not included in the analysis as the levels of O2 were changing rapidly at this point due to the flushing of the rebreather circuit with O2-enriched air. P<0.05 was taken as statistically significant and all data are presented as mean±95% reference range unless indicated otherwise.
Results
10 test runs were completed in each of open and closed-circuit configurations.
Open circuit
In the open-circuit configuration, the concentrator produced a statistically significant (p<0.0001), progressive, but modest elevation in FiO2 to 0.23 (0.225–0.235) (mean 95% reference range), 0.26 (0.247–0.271) and 0.30 (0.279–0.324), respectively, at concentrator settings of 2, 4 and 6. These values represent the combined FiO2 levels at all RRs. Changing RR in the range 12–20 breaths/min was associated with a small (p<0.0001) but clinically insignificant alteration in FiO2 (figure 3).
Effects of changing ventilator RR at constant tidal volume (500 mL) on FiO2 delivered to the simulated lung. Oxygen from the concentrator was introduced downstream into the ventilator air inlet at three progressively increasing output settings (S2, S4 and S6). Data shown as mean values (95% reference range). Where no bars to indicate the 95% reference range are visible, they are smaller than the symbol denoting mean values. FiO2, fractional inspired oxygen.
Closed circuit
Direct connection of the ventilator to the rebreather circuit (without using the reservoir tube)
Representative data recorded with a ventilator RR of 12 breaths/min are given below. When the ventilator was connected directly to the rebreather circuit in the lateral position without using a reservoir tube (figure 2A), and the output of the O2 concentrator introduced into the circuit via the low-pressure O2 adapter of the ventilator, the resulting FiO2 delivered to the simulated lung showed little change compared with the open-circuit configuration (figure 4A). However, when the O2 from the concentrator was added directly into the rebreather circuit, there was a small increase in FiO2 from 0.29 to 0.36 at the highest concentrator output (setting 6, 1260 mL/min) (figure 4A). This effect was only seen when the ventilator rate was set to 12 breaths/min (figure 4B,C).
FiO2 delivered to the simulated lung at three concentrator output settings (Inogen S2, S4 and S6). Ventilator tidal volume 500 mL, and (A–C) RRs 12, 16 and 20 breaths/min, respectively. Open circuit (no rebreather circuit): lateral, ventilator connected laterally to the rebreather circuit without a reservoir tube (see methods and figure 2A). The supplementary oxygen (O2) was delivered either into the low-pressure O2 adapter of the ventilator (O2 S) or directly into the rebreather circuit (O2 W). Data shown as mean values (95% reference range). Where no bars indicating the 95% reference range are visible, they are smaller than the symbol denoting mean values. FiO2, fractional inspired oxygen.
Addition of reservoir tubing between the ventilator patient connector and the rebreather circuit (lateral configuration)
Six test runs were completed with the ventilator/rebreather circuit in the lateral configuration after the addition of reservoir tubing (volume 626 mL) between the ventilator patient connector and the rebreather circuit (figure 2A,B). The supplementary oxygen from the concentrator was added directly into the rebreather circuit. The addition of the reservoir tubing led to an elevation of FiO2 to 0.78 (0.71–0.85), compared with 0.36 (0.36–0.37) seen with a similar configuration but without the reservoir tubing reported in the previous section. The FiO2 remained stable over a 4-hour evaluation period (figure 5). Inspired CO2 levels showed a slow steady rise, reaching 0.5% (fractional inspired CO2 (FiCO2) 0.005) after 170 (157–182) min and 1% (FiCO2 0.01) after 241 (222–253) min (median, quartile 1–3).
FiO2 and FiCO2 measured in the circuit over time. FiO2 delivered by the ventilator in the lateral position via a reservoir tube (figure 2) to the simulated lung at the highest concentrator output setting (Inogen S6). Ventilator tidal volume 500 mL and RR 12 breaths/min. Supplementary oxygen was delivered directly into the rebreather circuit. Data shown as mean values (95% reference range). FiCO2, fractional inspired carbon dioxide; FiO2, fractional inspired oxygen.
Discussion
Our proof-of-principle study shows that FiO2 of approximately 0.8 can be achieved with a single small concentrator, when combined with a ventilator and lightweight rebreather circuit. Both the concentrator and the ventilator are battery powered, with reported run times of approximately 4 hours19 20 on their own batteries. The total combined weight is approximately 5.1 kg, based on manufacturers’ information for each component (Wenoll rebreather circuit 1.6 kg, Inogen One G5 including single battery 2.15 kg and Ventway Sparrow ventilator, robust military version including battery 1.3 kg). Earlier studies showed that it is possible to deliver FiO2 in the range of 0.5–0.6 using a single O2 concentrator, weighing 5.5 kg in conjunction with several ventilators.14 15 A concern noted15 was that to provide for critically ill patients requiring higher FiO2 levels, either two concentrators or a supply of compressed O2 would be necessary. We have shown it is possible to overcome this limitation. The lightweight modular approach of our system, adaptable to the needs of a range of casualties, is particularly important for a variety of deployed small military medical teams operating at reach, where resupply and availability of external power is a major issue.
Our study examined the effects of placing the ventilator in the position occupied by the rebreather bag of the rebreather circuit, together with an APL valve over the rebreather circuit overflow valve (figure 2B). The circuit manufacturers recommend using an APL value to facilitate administration of ‘rescue breaths’ by squeezing the rebreather bag, should the patient need additional support. Replacing the bag with a ventilator simply extends this concept. However, when the ventilator’s patient circuit (connector tubing, in clinical use normally attached to an endotracheal tube) was attached directly to the Wenoll circuit, it was only possible to elevate FiO2 to approximately 0.4. Although this was an improvement to the FiO2 seen when the ventilator was used with the concentrator in open circuit, it was substantially below the target of 0.8–0.9. The likely explanation is that in this position, the exhaust valve of the ventilator patient circuit allows much of the O2-rich gas from the rebreather circuit to be lost on exhalation. A simple solution was to place a volume reservoir tube between the ventilator patient circuit and the rebreather circuit. This volume reservoir tubing was simply lightweight anaesthetic tubing. Providing the reservoir tubing volume is greater than the tidal volume, then each exhalation of O2-rich gas from the rebreather circuit will be held in the reservoir tube and pushed back into the rebreather circuit by air from the ventilator during the inspiratory cycle, thereby preserving the benefit of the rebreather circuit. Inevitably, there will be some mixing of gases at the interface between O2-rich gas and air in the reservoir tube, hence the lower FiO2 of 0.8 with the ventilator in the lateral position compared with 0.9 when it is in line within the rebreather circuit. Adopting a reservoir tube with a greater volume should reduce the impact of boundary mixing.
Conversely, it should be possible to deliberately entrain air into the rebreather circuit by reducing the volume of the reservoir tube to achieve a target FiO2 within the range of 0.3–0.8, thus titrating the amount of O2 provided to the patient to attain a target SpO2 of 92%. This approach could be used in conjunction with reducing the output of the concentrator to maximise battery run time. Under these circumstances, it would be essential to monitor the FiO2 continuously since, in theory, a hypoxic breathing gas could result from a concentrator failure. Monitoring the FiO2 continuously, with appropriate alarms, would obviate this risk.
Placing the ventilator in line within the inspiratory limb of the rebreather circuit (data not shown) did result in a further improvement in FiO2 to 0.9. However, this comes at a significant disadvantage that the ventilator would be used out of its operating specification in relation to gas intake. Additionally, it would represent a marked modification of the rebreather circuit.
The concept of using an O2 concentrator with low flow rates (0.5–1 L/min) in conjunction with a semiclosed rebreather circuit and a ventilator is not new. This concept was reported in the 1980s, although there were concerns regarding the accumulation of argon in the system,21–23 which may be obviated by increasing the fresh gas flow rate (of O2) to a minimum of 1 L/min in the rebreather circuit.24 More recently, a pilot study and a clinical trial have reported a hospital-based system to provide anaesthesia.25 26
The advantage of a modular system as described in our study is that it can be rapidly reconfigured to support either a spontaneously breathing casualty or an individual requiring ventilation. Switching between the two configurations involves replacing the rebreather bag with the volume reservoir tube and ventilator (one push-fit to disconnect/connect) and placing the APL valve over the rebreather circuit overflow valve (a second push-fit to connect/disconnect).
An obvious limitation of our study is that it is a laboratory-based investigation using an inanimate system. We did, however, emulate a patient’s O2 consumption and CO2 production at rates representative of a large adult at rest. Consequently, the study does provide a proof of principle that it is possible to deliver a high FiO2 (approximately 0.8) continuously for several hours using small, portable, lightweight concentrator, rebreather circuit and ventilator without the need for pressurised cylinders or external power, and without extensive modification of any component. The run times will of course be influenced by several factors including modes of ventilation, patient metabolic rate and ambient conditions.
In summary, pressurised O2 imposes a huge logistic burden (resupply and transport and ballistic threat). The ability to provide safe, titratable O2 to spontaneously breathing or ventilated patients for long periods of time without the weight and volume complications of the current arrangements will prove invaluable for many military, civilian and humanitarian situations.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Footnotes
Contributors The study was designed and initiated by MB, EK, CP, SW and TW. Pilot studies were conducted by EK and SW. The majority of data collection was conducted by CW. EK wrote the first draft of the paper, and all authors contributed to the refining and editing of the paper. Clinical evaluation was conducted by MB, CP and TW. Scientific evaluation was conducted by EK, CW and SW. EK is the author acting as guarantor.
Funding This work was funded by the UK Ministry of Defence.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.