Evaluation and Characteristic of Modied Ventilator Under Hyperbaric Conditions During Volume-controlled Ventilation

Purpose(cid:0)The stability of the modied ventilator (Shangrila590, Beijing Aeonmed Company, Beijing, China) was evaluated under hyperbaric conditions during volume-controlled ventilation in this study by Michigan test lung (5601i, Grand Rapids, MI, US). Methods(cid:0)Experiments were performed inside the multiplace hyperbaric chamber at 1.0, 1.5 and 2.0 atmospheres absolute (ATA). The modied ventilator placed inside the hyperbaric chamber was connected to the test lung. During volume-controlled ventilation (VCV), data for the test lung were collected by a personal computer outside the hyperbaric chamber. The preset tide volume (V Tset ) of the ventilator (400-1000 ml) and the resistance and compliance of the testing lung were adjusted before the experiments at every ambient pressure. With every test setting, the tide volume (V T ), inspiratory airway peak pressure (Ppeak) and minute volume (MV) displayed by the ventilator and the test lung were recorded by the computer. We compared the ventilator and test lung data under 1.0, 1.5 and 2.0 ATA to evaluate the stability of the modied ventilator. changes ambient obviously, ambient condition. Ppeak increased obviously because of the raised airway resistance due to high ambient pressure.


Introduction
Hyperbaric oxygen (HBO) therapy involves treatment with a high fraction of inspired oxygen (FiO 2 ) at higher than atmospheric pressure, whereby increased pressures depend on guidelines and indications.
HBO therapy is widely used in critically ill patients in the intensive care unit (ICU), such as for acute carbon monoxide poisoning, decompression sickness, arterial gas embolism, radiation-induced tissue injury, and acute traumatic ischemic injury [1,2]. HBO therapy remains among the safest therapies used today under the common range of the 2-3ATA environment [3]. Arterial partial pressure of oxygen (PaO 2 ) can be improved through increased pressure of oxygen (PO 2 ) in inspired breathing gas physiologically, especially during HBO treatment. According to the gas law, PO 2 in inspired breathing gas can be elevated not only by increasing FiO 2 but also by increasing the pressure of breathing gas. Under a normal baric environment, we can only limitedly raise PO 2 by increasing FiO 2 [4]. Alternatively, we can elevate PO 2 by increasing ambient pressure and FiO 2 , which can raise the oxygenated e ciency of pulmonary tissue in the hyperbaric chamber. When the ventilator is placed in the hyperbaric chamber with the same parameters as outside the chamber, pulmonary gas exchange will be markedly improved if mechanical ventilation can provide stable pulmonary ventilation. However, the ordinary ventilator used in the ICU does not work well during HBO therapy because of the high ambient pressure. This limits mechanical ventilation carried out during HBO therapy. Overall, broadening the eld of mechanical ventilation may bene t patients on mechanical ventilation who need HBO therapy.
Not all medical devices can be subjected to hyperbaric chambers; indeed, most life support technologies, such as hemo ltration, electrical de brillation and extracorporeal membrane oxygenation, are at present incompatible with the hyperbaric environment [5,6]. Pneumatical ventilators can operate in hyperbaric environments safely, but they cannot provide stable V T , respiratory rate (RR) or MV with raised ambient pressure [7,8]. Stable operation of a ventilator should be the guarantor for patients on mechanical ventilation, not only in the ICU but also in the hyperbaric chamber. Modern ventilators tend to be electropneumatical and electronic. In recent years, a series of bench tests have been carried out on ventilators under hyperbaric conditions during different ventilation models, such as VCV and pressure-controlled ventilation (PCV) [9,10]. Nevertheless, most electropneumatical ventilators cannot function well in hyperbaric chambers. In the early stage, researchers focused on empirically predicting changes in ventilation parameters under speci c high pressures and then adjusted the parameters of the ventilator to manually compensate for the changes. With the understanding of respiratory mechanics in hyperbaric environments and the development of modern ventilator technology, modi ed ventilators for hyperbaric conditions have been developed.
Modi ed ventilators can adjust the compensation by itself automatically, precisely and rapidly when ambient pressure changes, for example, Siaretron 1000 Iper (Bologna, Italy) [10,11]. However, the type of ventilator for hyperbaric conditions is limited globally. In China, we do not have our own ventilator for hyperbaric conditions with independent intellectual property rights. In brief, we rst modi ed the Shangriala590 ventilator (Beijing, China) and then performed a series of bench tests in a hyperbaric chamber before putting it to use clinically. In this study, we detected the stability of V T and MV by a test lung during VCV in a hyperbaric chamber, and we clari ed the characteristics of hyperbaric ventilation according to the respiratory mechanics parameters.

The ventilator
The Shangrila590 ventilator is an electropneumatic ventilator from Beijing Aeonmed Company that is commonly used in the ICU in China. According to the safety regulations of medical hyperbaric chambers in China [12,13], the pneumatic part was placed in the chamber, and the electronic part was assembled out of the chamber. The two parts of the ventilator were connected with wire through the chamber without leakage, allowing doctors to operate the ventilator outside the hyperbaric chamber. Ventilator engineers modi ed the algorithm and replaced some components of the ordinary ventilator to make the ventilator work reliably and safely in a hyperbaric environment.

The test lung
We used the Model 5601i Adult/Infant PneuView System (Michigan Instrument, Grand Rapids, MI, US) to detect ventilation parameters. The detection system contains Michigan test lung and PneuView data collection software. Model 5601i houses an electronic interface module, which converts the pressure signal from the test lung to digital data and transfers these data to a personal computer. PneuView data collection software serves as the link between the test lung and the computer.
The critical care multiplace hyperbaric chamber It has been proven that a multiplace hyperbaric chamber is better suited for HBO treatment of critically ill patients than a monoplace hyperbaric chamber because it permits appropriate ICU equipment to be used inside the chamber by staff [14]. The critical care hyperbaric chamber [GY3800-A (GY3800 M2-D), Yantai Hongyuan Oxygen Industrial Inc., Yantai, China] is a multiplace hyperbaric chamber with an automated operation system equipped with electrocardiogram monitors, ventilators, transcutaneous oxygen (O 2 ) and carbon dioxide (CO 2 ) tension monitors, syringe drivers, and infusion pumps, among others, to ensure the continuous treatment of ICU patients. Our chambers have three compartments, two ICU chambers and a prechamber between them, and have the capacity for 24 seated people or 8 gurneys.

The experimental con guration
We calibrated the ventilator and the test lung at atmospheric pressure before the experiments. The test lung was located inside the hyperbaric chamber and connected to the pneumatic part of the ventilator. The digital data detected by the test lung were passed by electrical penetration and wires through the bulkhead to the personal computer outside ( Figure 1). According to the parameters shown in the Calibration Speci cation for Ventilators [15], the ventilator was adjusted by the doctors outside, and the resistance and compliance of the test lung were regulated by the staff inside according to the test settings in Table 1.

Results
During VCV with every V Tset , V T displayed by the ventilator itself was higher than that detected by the test lung in the same ambient pressure, especially for the V Tset 800 ml groups and V Tset 1000 ml groups ( Table   2). With increased ambient pressure, V T decreased not only in the test lung groups but also in the ventilator groups. Changes in V T in the test lung and ventilator at different ambient pressures seemed stable, and the differences were statistically signi cant ( Figure  2). Compared with V Tset , downtrends of V T displayed by the ventilator were 0.5-1 % for the V Tset 400 ml group, 1-2.6 % for the V Tset 500 ml group, 4-6.5 % for the V Tset 600 ml group  (Table 3).
With increased ambient pressure during VCV at a xed level of V Tset , we observed increased Ppeak displayed by the ventilator but stable Ppeak detected by the test lung ( Figure 4). Statistical analyses showed that Ppeak displayed by the ventilator was higher than that detected by the test lung for the same V Tset group and ambient pressure (

Discussion
Advanced investigation showed that ordinary ventilators used at normal atmospheric pressure cannot maintain stable V T during VCV when operated at high atmospheric pressure. Inspiratory ow provided by the ventilator will decrease with increasing ambient pressure [7,8,9,10]. The reason is that during HBO therapy, the ambient pressure is raised by compression air, which results in a high breath gas density and has no in uence on breath gas viscosity. High breath gas density results in more turbulent ow in peripheral airways according to an increased Reynold's value (>1000, Figure 5). To obtain the same inspiratory ow, turbulent ow produces higher airway resistance than laminar ow. Hence, more driving pressure (△P) must be provided by the ventilator to overcome the higher airway resistance; otherwise, it will lead to decreased inspiratory ow. Unless this phenomenon is technically compensated for, hypoventilation may occur due to decreased V T and MV [11,17]. To maintain adequate VT and MV, we need to increase the inspiratory ow as the chamber pressure increases through manual regulation or automatic compensation [7,8,9,10,16].
Evaluating V T during VCV at high ambient pressure During VCV, the modi ed Shangrila590 ventilator can provide more △P to overcome the greater resistance and achieve constant V T and MV, even though V T and MV decreased within a narrow range. In the high V Tset group (800-1000 ml), the decline in V T as the ambient pressure increased was greater than that in the low V Tset group (400-600 ml) (Table 2, Figure 2). However, the degree of decline was smaller than that in previous research at the same ambient pressure scale, which resulted in a 20-56 % decline in V T [7,9]. Nonetheless, compared with the hyperbaric ventilator Siaretron IPER 1000, a 6.5-20 % increase in V T during VCV under 1.0 ATA to 2.2 ATA ambient pressure occurred, which is CE-certi ed for hyperbaric use in Europe [10]. When we focused on the accuracy of V T displayed by the ventilator, it was not exactly equal to V T detected by the test lung. For V Tset between 400-600 ml, the V T displayed by the ventilator may overestimate the actual V T Otherwise, for V Tset between 800-1000 ml, the V T displayed by the ventilator may underestimate the actual V T . The difference in V T displayed by the ventilator and by the test lung was narrow during V Tset between 400-600 ml, but it was wide during V Tset between 800-1000 ml. However, the accuracy of the test lung in previous studies was controversial at high ambient pressure [7,9]. We used a water tank-simulated lung in pre-experiments, which can roughly re ect the true value of ventilator V T . The data between the ventilator and the water tank-simulated lung seemed good, but the numerical precision of the water tank-simulated lung was low. For statistical analysis, we decided to choose the Michigan test lung (5601i) for the test equipment, according to previous research [7].
During VCV, the goal of maintaining constant V T is to take in enough O 2 and ensure expiration of CO 2 .
Factors affecting the expiration of CO 2 include not only Vt but also dead space and high PaO 2 . During HBO therapy, PaO 2 is much higher than that under normobaric conditions. Under hyperbaric conditions, respiratory resistance leads to decreased breath gas ow and enlarged dead space. These may reduce exhalation of CO 2 [18,19]. In addition to the stable operation of ventilators, it is essential to monitor the expiratory volume, arterial partial pressure of carbon dioxide (PaCO 2 ), transcutaneous carbon dioxide tension (P TC CO 2 ), or end-tide carbon dioxide partial pressure (P ET CO 2 ) [11,20].

Changes in Peak during VCV at high ambient pressure
Our data showed that the Ppeak displayed by the ventilator increased obviously during VCV with xed V Tset in the process of ambient pressure rise (Table 4, Figure 4). As Ppeak can re ect inspiratory resistance, the ventilator can provide more △P to overcome increased airway resistance to maintain stable V T . Our data shown in Figure 4 supports this mechanism. However, changes in Ppeak detected by the test lung were gentle because of the different detected positions of the ventilator and the test lung. The breathing gas ow was buffered when detected by the test lung.
Side effect of HBO in pulmonary system and preventive measures In general, there is no risk of pulmonary barotrauma (PBT) in patients with normal lungs during HBO therapy. Based on Boyle's Law, there is potential for PBT due to lung overin ation during decompression when disease is present, such as asthma or chronic obstructive pulmonary disease (COPD) with active bronchospasm, mucous plugging, and bullous lung disease. Additionally, pneumothorax (PTX) is a potentially life-threatening phenomenon, especially given the increased risk of tension PTX during decompression. All candidates for HBO therapy must be screened for pulmonary disease to avoid increasing the risk of PBT and PTX [21].
Continuous exposure of the lungs to elevated PaO 2 , either at normobaric or hyperbaric pressure, leads to toxic effects of O 2 . Pulmonary O 2 toxicity can be avoided if O 2 is provided at the proper dose [5,22]. When  [22].
In the hyperbaric chamber, we can obtain higher PO 2 by increasing ambient pressure with lower FiO 2 to avoid absorption atelectasis. In rats, HBO exposure caused signi cant oxidative stress in the rst 24 h. However, these effects were resolved at the end of the tenth day of HBO treatment [23]. There are two pathways for the development of CO 2 intoxication. PCO 2 is increased in inspired breathing gas or expiration of produced CO 2 is insu cient [17]. Increased PCO 2 in inspired gas may occur when gas exchange occurs in the hyperbaric chamber. To prevent raised PCO 2 levels, hyperbaric chambers must be ushed continuously with breathing gas. Increased breathing resistance in the hyperbaric chamber may decrease the expiration of produced CO 2 . Maintaining stable pulmonary ventilation and monitoring PaCO 2 by blood gas, P ET CO 2 , and P TC CO 2 must be established during HBO therapy [11,17,21].
Work of breathing in hyperbaric environments is also a concern. Combined with the breathing equipment itself, the work of breathing will be increased compared to breathing the same gas in a normobaric environment [6]. For patients on mechanical ventilation, the endotracheal tube diameter is critical with regard to its effect on airway pressure and work of breathing [18]. During HBO therapy, we must consider that a high breath gas density induces high airway resistance, which cannot be avoided in hyperbaric chambers. In addition to sputum aspiration and exchange for large endotracheal intubation, we can decrease airway resistance by prolonging the inspiratory time appropriately and using a helium oxygen mixture to decrease the gas density. Additionally, we can reduce the high airway resistance and breathing work by downregulating ambient pressure or upregulating the support pressure of the ventilator.

Conclusions
In summary, during HBO therapy with raised ambient pressure from 1.0 ATA to 2.0 ATA, a modi ed Shangrila590 ventilator made in China can provide stable V T and MV during VCV with V Tset from 400 ml to 1000 ml. During VCV, Ppeak increased obviously because of the raised ambient pressure. We will evaluate other ventilator models, such as PCV, in the future. Advanced detection of this ventilator operating in an environment of more than 2.0 ATA will be carried out in our further study.  Tables   Due to technical limitations, table 1 to 4 is only available as a download in the Supplemental Files section. Figure 1 The experimental con guration Page 12/14

Figure 2
Changes in tide volume during volume-controlled ventilation at different ambient pressure Figure 3 Changes in minute volume during volume-controlled ventilation at different ambient pressure  Type of breathing gas ow

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Table.pdf