Research ArticleOriginal Research

External Jet Nebulization and Measured Ventilator Performance

Jeyanthan Jayakumaran, Gerald C Smaldone and Ann D Cuccia
Respiratory Care July 2024, 69 (7) 790-798; DOI: https://doi.org/10.4187/respcare.11296

Abstract

BACKGROUND: During invasive ventilation, external flow jet nebulization results in increases in displayed exhaled tidal volumes (VT). We hypothesized that the magnitude of the increase is inaccurate. An ASL 5000 simulator measured ventilatory parameters over a wide range of adult settings: actual VT, peak inspiratory pressure (PIP), and time to minimum pressure.

METHODS: Ventilators with internal and external flow sensors were tested by using a variety of volume and pressure control modes (the target VT was 420 mL). Patient conditions (normal, COPD, ARDS) defined on the ASL 5000 were assessed at baseline and with 3.5 or 8 L/min of added external flow. Patient-triggering was assessed by reducing muscle effort to the level that resulted in backup ventilation and by changing ventilator sensitivity to the point of auto-triggering.

RESULTS: Results are reported as percentage change from baseline after addition of 3.5 or 8 L/min external flow. For ventilators with internal flow sensors, changes in displayed exhaled VT ranged from 10% to 118%, however, when using volume control, actual increases in actual VT and PIP were only 4%–21% (P = .063, .031) and 6%–24% (P = .25, .031), respectively. Changes in actual VT correlated closely with changes in PIP (P < .001; R2 = 0.68). For pressure control, actual VT decreased by 3%–5% (P = .031) and 4%–9% (P = .031) with 3.5 and 8 L/min respectively, PIP was unchanged. With external flow sensors at the distal Y-piece junction, volume and pressure changes were statistically insignificant. The time to minimum pressure increased at most by 8% (P = .02) across all modes and ventilators. The effects on muscle pressure were minimal (∼1 cm H2O), and ventilator sensitivity effects were nearly undetectable.

CONCLUSIONS: External flow jet nebulization resulted in much smaller changes in volume than indicated by the ventilator display. Statistically significant effects were confined primarily to machines with internal flow sensors. Differences approached the manufacturer-reported variation in ventilator baseline performance. During nebulizer therapy, effects on VT can be estimated at the bedside by monitoring PIP.

Introduction

During mechanical ventilation, aerosolized medications have been traditionally delivered by using jet nebulizers.1 Departments of respiratory therapy often define hospital protocols, which include device choice and methods of aerosol delivery. Factors involved in decision-making include cost, ease of use, and perception of device function. Jet nebulizers require a pressurized gas to generate aerosol. Many modern ventilators are equipped with an in-line nebulization option integrated with the ventilator flow sensor, providing control of delivered volume and pressure. However, these systems are not standardized and the combination of an unregulated pressure/flow source with a given nebulizer can affect drug delivery.2 In addition, the integrated in-line systems cannot deliver continuous nebulization. Nebulizer driving pressure and flow can be standardized by using wall air or oxygen. Clinicians experienced in aerosol delivery know that external gas added to the circuit can affect the ventilator readout on some machines, with marked elevation in the displayed exhaled VT.3,4 We hypothesized that the magnitude of this increase is inaccurate.

However, although potential effects are often mentioned, there are few sources that report actual delivered volumes, pressures, and triggering behavior. Cuccia et al,5 in a study of aerosol delivery by using a prototype jet nebulizer, the i-AIRE (InspiRx, Somerset, New Jersey), tested its effect on several ventilators by using different modes and breathing patterns. The i-AIRE was driven with wall air at a flow of 3.5 L/min 50 psig. They measured changes in delivered VT by using an ASL 5000 (ASL) test lung (Ingmar Medical, Pittsburgh, Pennsylvania) and found small changes in delivered VT, much less than the indicated values on the ventilator monitor.5 It is important to note that the study by Cuccia et al5 tested the nebulizer in a “breath actuated” condition, which can reduce effects of added gases. Other reports, some in abstract form, report variable changes in delivered volumes and pressures.6

Li et al,7 in a recent comprehensive study, measured delivered volume and pressure by using several ventilators. They tested a single breathing pattern that represented a patient with COPD when using the ASL breathing simulator, which served as both a test lung and a ventilator trigger (similar to Cuccia et al5). In the volume control mode, they reported increases in VT as well as increases in inspiratory pressures. The ASL also provided a measure of time to reach the lowest airway pressure after a patient-initiated effort (time to minimum pressure). Li et al7 reported that the time to minimum pressure was increased and statistically significant, with the implication that this will further affect patient ventilation.

The present paper is designed to further explore this problem by following a similar protocol to that used by Li et al.7 We tested a broader range of adult breathing patterns designed to reflect commonly used modern ventilator settings for different patient conditions and additional measures of patient-triggering, measuring changes induced by adding external flow to the circuit, duplicating the clinical use of continuous nebulization.

QUICK LOOK

Current Knowledge

During invasive ventilation, ventilators equipped with an in-line nebulization option are not standardized. The combination of an unregulated pressure/flow source with a given nebulizer can affect drug delivery and cannot deliver continuous nebulization. External flow jet nebulization allows for standardization of driving pressure and flow by using wall air or oxygen; however, this results in increases in displayed exhaled tidal volume. Interactions between continuous external jet flow and mechanical ventilators are not well defined.

What This Paper Contributes to Our Knowledge

Actual tidal volume changes were significantly lower than those displayed by the ventilator. Analysis of our findings suggests that these changes are best monitored through associated changes in peak inspiratory pressure.

Methods

Three adult breathing patterns were evaluated to simulate normal physiology, COPD, and ARDS. Ventilators tested included Dräger V500 critical care ventilator (Dräger, Telford, Pennsylvania), Servo-i (Maquet, Getinge, Solna, Sweden), Bella Vista (Vyaire Medical Inc, Mettawa, Illinois), and Hamilton-C1 (Hamilton Medical, Bonaduz, Switzerland). The tested Dräger V500 and Servo-i ventilators use an internal flow sensor, whereas the Bella Vista 1000 and Hamilton-C1 machines use an external flow sensor located at the distal Y-piece of the circuit near the test lung.

The following ventilator modes were tested for each patient breathing pattern: VC-CMV (volume control continuous mandatory ventilation), VC-IMV (volume control intermittent mandatory ventilation), PC-CMV (pressure control continuous mandatory ventilation), PRVC (pressure regulated volume control), and CSV (continuous spontaneous ventilation), with APRV tested only in the ARDS patient model for all ventilators. VC-CMV and VC-IMV were not available for the Hamilton ventilator. Ventilator settings and patient breathing patterns were chosen to simulate common scenarios that could reasonably present to a clinician in the clinical setting (Table 1 & 2). VT was chosen to simulate lung-protective ventilation of a 70-kg ideal body weight adult. Clinically relevant inspiratory times for patients who are critically ill and require invasive ventilation were selected. Pressure support settings were adjusted to achieve the target VT. Default bias flow for each ventilator was used: 2 L/min for the Dräger V500 and Servo-i, 3 L/min for the Hamilton-C1, and 6 L/min for the Bella Vista 1000. Complete ventilator settings are detailed in the supplementary file (see the supplementary materials at http://www.rcjournal.com).

View this table:
Table 1.

Experimental Conditions: Tested Ventilator Modes and ASL 5000 Settings

View this table:
Table 2.

Effect of Added Nebulizer Flow Separated by Ventilator Flow Sensor Type

The ASL test lung served to both monitor ventilatory parameters and simulate a variety of relevant patient scenarios, which consisted of set compliance, resistance, and rate. Different from other test lungs, the ASL can trigger the ventilator, through generation of negative inspiratory force that can be set and modified by the operator. Settings for each patient scenario in Table 1 were based on manufacturer recommendations as well as previously published literature.8,9 Settings for the normal example were chosen from the ASL library. For the normal settings, inspiratory muscle force was decreased from the default settings to avoid double-triggering of the ventilator at the chosen set VT.

Each ventilator was connected directly to the test lung, without active humidification. Two nebulizers were tested. They were chosen based on recommended flow and pressure settings. The experimental setup is illustrated in Figure 1. The i-AIRE, placed at the ventilator outlet port, operated at 3.5 L/min and air at 50 psig, and the Hudson RCI (Teleflex Medical, Research Triangle Park, North Carolina), placed in the inspiratory limb proximal to the Y-piece position, operated at 8 L/min and air at 50 psig. All the nebulizers were run with dry gas only. The ventilator and ASL were monitored during testing with and without added flow from the test nebulizer. Steady-state (after 5–10 breaths) readings were recorded of exhaled VT measured by the ventilator and VT delivered (actual VT) measured by the ASL. Other recorded parameters included peak inspiratory pressure (PIP) from the ventilator readout as well as time to reach minimum pressure (time to minimum pressure, a parameter reported by Li et al,7 from the ASL. To evaluate triggering, 2 maneuvers were performed; the first was tested by reducing inspiratory muscle force (muscle pressure, Pmus) (with ventilator sensitivity fixed at its default setting) until the ventilator defaulted to its backup rate, the second trigger function was tested by gradually increasing sensitivity (with the chosen Pmus fixed) until auto-triggering was observed.

Fig. 1.
Fig. 1.

Experimental setup for different nebulizer locations with only one nebulizer placed in the circuit at a time. The i-AIRE breath-enhanced jet nebulizer prototype located at the inspiratory outlet of the ventilator powered by 3.5 L/min air from flow meter. Hudson nebulizer placed in the inspiratory line proximal to the patient Y-piece powered by 8 L/min air from flow meter. External flow sensor shown at the distal Y-piece position for tested ventilators (Bella Vista and Hamilton).

The statistical significance of the effects of added external flow was assessed by using the Wilcoxon test (GraphPad Prism for Mac OS X, GraphPad Software, San Diego, California) to compare baseline measurements with no additional flow to those with added flows of 3.5 L/min and 8 L/min. Comparison data were grouped based on mode type across the tested ventilators and patient scenarios. Differences were reported between ventilator and ASL measurements with and without added flow. Data were separated for the internal and external distal Y-piece flow sensor ventilators. Data were reported as mean ± SD.

Results

A complete data set for all measurements is available in the supplementary files. Summary data are reported in Table 2 and in the online supplement (see the supplementary materials at http://www.rcjournal.com). For internal flow sensor ventilators (Table 2) across volume control modes, changes measured by the ASL (actual VT) were much lower than the changes in exhaled VT displayed on the ventilator. exhaled VT changes displayed on the ventilator ranged from 45 ± 11% to 118 ± 32%, whereas actual VT delivered increased from 4 ± 4% to 11 ± 2% with the introduction of 3.5 L/min of external flow and 12 ± 9% to 21 ± 9% for 8 L/min of added flow. PIP changes were of similar magnitude (6 ± 8% to 12 ± 0% with 3.5 L/min, and 14 ± 14% to 24 ± 8% with 8 L/min). In pressure control modes, actual VT decreased slightly (−3 ± 2% to −9 ± 10%) for both 3.5 L/min and 8 L/min added flow. All changes, although small, were statistically significant (P = .031). With added flow, PIP did not change significantly, with P values that ranged from P = .031 to P > .99. Despite the small, measured decreases in VT, as assessed by actual VT in pressure controlled modes, the ventilator monitor indicated marked false elevations in exhaled VT of 41 ± 11% to 110 ± 33%. In APRV, VT delivered decreased slightly, –3 ± 5%, with 3.5 L/min of added flow and –13 ± 5% with 8 L/min of added flow. These measured changes during APRV were different in magnitude and direction than the mean ± SD 10 ± 17% and 23 ± 38% increase in exhaled VT reported on the ventilator monitor for 3.5 L/min and 8 L/min added flow, respectively.

Ventilators with the external flow sensors located in the distal Y-piece position behaved differently. As shown in Table 2, the measured changes in all the parameters were less than their internal flow sensor counterparts and largely statistically insignificant (outside of the changes noted with 8 L/min of added flow in PRVC with an average increase of 1% [P = .031]). Across the volume controlled modes, exhaled VT usually mirrored actual VT. With 3.5 L/min of external flow, exhaled VT increased by 0 ± 1% to 9 ± 3%, while actual VT increased by 0 ± 1% to 9 ± 3%. With 8 L/min of external flow, exhaled VT increased by 1 ± 1% to 11 ± 4% and actual VT increased by 2 ± 2% to 12 ± 3%. PIP changes were of similar magnitude, with an increase of 2 ± 3% to 6 ± 0% with 3.5 L/min, and 2 ± 3% to 10 ± 7% with 8 L/min.

In the pressure controlled modes, actual VT ranged from –2 ± 6% to 2 ± 2% with added flow. These changes were similar to those displayed on the ventilator, with exhaled VT that ranged from –3 ± 9% to 2 ± 3%. PIP did not change significantly in pressure controlled modes (–3 ± 8% to 3 ± 5%). In APRV, volume delivered decreased slightly with 3.5 L/min added flow, –3 ± 1% and increased slightly with 8 L/min added flow, 8 ± 6%. These changes were different than those displayed on the ventilator, with a negligible decrease in exhaled VT of 0 ± 5% with 3.5 L/min and decreased by 5 ± 1% with 8 L/min.

Changes in actual VT and PIP, for 3.5 L/min and 8 L/min added external flow when using volume controlled modes and internal flow sensor ventilators are described in Figure 2. For all data, the relationship is described by the linear correlation, y = 17.21x + 13.51 (P < .001, R2 = 0.68). The coefficient of determination indicates that 68.2% of the variability is explained. The same analysis for changes in exhaled VT versus changes in PIP is illustrated in Figure 3. The regression, although significant (P = .02), poorly explains the data (R2 = 0.16).

Fig. 2.
Fig. 2.

Change in tidal volume delivered (Δ actual VT) (mL) vs change in peak inspiratory pressure (ΔPIP) (cm H2O) during volume control modes, internal flow sensor ventilators; 3.5 L/min external flow, 8 L/min external flow; y = 17.21x + 13.51, P < .001, R2 = 0.68.

Fig. 3.
Fig. 3.

Change in displayed exhaled tidal volume (VT) (mL) vs change in peak inspiratory pressure (PIP) (cm H2O) during volume control modes, internal flow sensor ventilators; 3.5 L/min external flow, 8 L/min external flow; y = 39.9x + 243.3; P =  .02; R2 = 0.16.

The time to minimum pressure (Table 3) increased between 1 and 8% with 3.5 L/min external flow. At 8 L/min, unusually large changes were seen on the Dräger ventilator when using the normal patient ASL settings (eg, time to minimum pressure change from baseline 54 ms to 2,046 ms in VC-CMV); however, the ventilator continued to trigger with no dropped breaths or change in ventilator frequency. For all other measurements at 8 L/min flow, with the Dräger findings excluded, time to minimum pressure increased from 1 to 8%.

View this table:
Table 3.

Changes in Time to Minimum Pressure Reported as Percentage Change from Baseline*

View this table:
Table 4.

Muscle Pressure at Failure to Trigger (initial set point 8–12 cm H2O)

Results for measures of ventilator triggering via changes in patient muscle strength are summarized in Tables 4. Muscle strength (Pmus) settings were lowered until the ventilator failed to trigger, defined by initiation of the backup rate. With added flow, differences in muscle strength (Pmus) required to trigger the ventilator were small, <1 cm H2O with 3.5 L/min and at most by ∼1 cm H2O with 8 L/min. The independent measure, ventilator sensitivity, assessed by adjusting sensitivity to the most sensitive setting for both pressure and flow triggers, rarely induced auto-triggering, with failure observed 3 times with 8 L/min and 4 times with 3.5 L/min across a total of 104 trigger tests per flow.

Discussion

Although changes in ventilator performance with added flow to the circuit are real, they are predictable and often minimal. Changes in PIP, as shown in Figure 2, which can be measured without the ASL, paralleled the changes in actual VT and indicate that a clinician at the bedside can easily assess real changes in delivered volume by looking at the magnitude of changes in PIP. For all settings, the changes in delivered volume measured by the ASL were markedly different from the readouts on the ventilator monitors for machines that used an internal flow sensor. The changes in exhaled VT recorded by the ventilator did not correlate well with PIP (Fig. 3). Of note, during testing of the Dräger V500, a flow measurement alarm would be triggered with the addition of external flow. Despite the alarm, the ventilator continued to trigger and deliver the volumes reported. External flow sensors placed in the distal Y-piece readily compensated for any added external flow. Our measures of ventilator triggering (changing ventilator sensitivity and the ASL index for muscle activity) were minimally responsive to added flow. Preliminary data from earlier studies predicted that many results will be statistically significant because expected changes, no matter how minimal, will be in one direction and, therefore, significant, but small changes may not be clinically important.

Review of manufacturer specifications indicated that many repeated measurements of function might vary as much as 10% (8–10% volume, 4–6% pressure).10,-,14 The ventilator manufacturer specifications for volume and pressure accuracy are summarized in Table 5, and the changes observed in this study are illustrated in Table 3. Changes to the actual VT delivered with the introduction of nebulizer flow are reflected in the columns Δ actual VT 3.5 L and Δ actual VT 8L. With the introduction of 3.5 L/m nebulizer flow, the internal flow sensors showed a volume change of 4–11% in volume controlled modes; however, all other modes were ≤4%. External flow sensors showed a change of ≤9% across all modes. With the introduction of 8 L/m nebulizer flow, internal flow sensors showed a change of 12–21% in volume controlled modes and external flow sensors showed a change of ≤12% across all modes. Changes in PIP with the introduction of nebulizer flow are reflected in the columns: ΔPIP 3.5 L and ΔPIP 8 L. With the introduction of 3.5 L/m nebulizer flow, internal flow sensors showed a pressure change of ≤12%, while external flow sensors showed a change of ≤4% across all modes. With the introduction of 8 L/m nebulizer flow, internal flow sensors showed a change of ≤24% and, with external flow sensors, a change of ≤10% across all modes.

View this table:
Table 5.

Ventilator Manufacturer Specifications for Volume and Pressure Accuracy10,11,13,14

Li et al7 also examined the influence of external flow, testing one breathing pattern (COPD) across several ventilators. Six ventilators were tested, 5 with internal flow sensors and one distal Y-piece sensor. Our findings for actual VT and PIP for internal flow sensor ventilators were comparable with theirs. However, Li et al7 recommend that the therapist should lower VT settings based on the exhaled VT displayed on the ventilator. This recommendation seems inconsistent with both our observations and the data reported by Li et al.7 That is, the exhaled VT changes displayed on the ventilators tested were much larger than the true increases in VT. Making ventilator adjustments based on exhaled VT could potentially lead to hypoventilation of the patient. Analysis of our results suggests that clinicians should simply monitor the PIP. For ventilators with internal flow sensors, measured changes in exhaled VT are not accurate. True changes in delivered VT can be assessed by observing PIP, which reflects actual volume changes delivered to the patient, and PIP is not subject to flow artifacts. This is best seen in Figure 2, which describes a strong linear relationship between changes in actual VT and changes in PIP. For example, an addition of 3.5 L/min generally increased PIP by ∼2 cm H2O. The greatest changes (eg, 6 cm H2O) were measured with the addition of 8 L/min in models with low lung compliance (ARDS model). Due to the weak correlation between exhaled VT and PIP, it may be unsuitable to make ventilator adjustments based on this parameter.

Li et al7 also examined “triggering performance” through the evaluation of variables displayed on the ASL, one of which was the time to reach the lowest airway pressure (time to minimum pressure). They reported significant increases and concluded that these changes may affect a patient’s ability to reliably trigger the ventilator. Although we found much smaller changes in time to minimum pressure across most ventilators tested (except for Dräger at 8 L/min), no faulty triggering or dropped breaths were noted, even for Dräger, which suggested to us that this parameter is not related to measures of trigger function. measures of trigger function and changes in muscle force and ventilatory sensitivity were minimally affected.

Our study has limitations. Bench studies, including ours, may not reflect actual patient behavior but guide the initial approach to therapy as well as future directions in studies on patients who are critically ill. We did not evaluate all commercially available ventilators, nebulizer positions, and clinically relevant situations. In this study, we specifically used an adult patient model, it is unclear what the effects of external flow would be on smaller patients, such as pediatric or neonatal patients, by using lower VT.

Conclusions

Based on our findings, clinicians should be aware that external flow affects ventilators differently based on the location of the flow sensor. Ventilators with the flow sensor in the distal Y-piece position include external flow into their exhaled VT calculation after a few transient breaths, compensating for external flow with a quick return to baseline. For ventilators with internal flow sensors, external VT readings are inaccurate. Our bench study indicates that, to estimate VT changes, it is more reliable to monitor PIP, which increased in parallel to actual changes in VT.

Acknowledgments

The authors thank Stony Brook University Hospital Respiratory Care Department.

Footnotes

  • Correspondence: Ann D Cuccia MPH RRT RRT-NPS RPFT AE-C FAARC, School of Health Professions, Stony Brook University Respiratory Care Program, 100 Nicolls Road/HSC Level 2, Room 410, Stony Brook, NY 11794–8203. E-mail: ann.cuccia{at}stonybrook.edu

  • The location of the study was the Pulmonary Mechanics and Aerosol Research Laboratory, Division of Pulmonary, Critical Care and Sleep Medicine, Health Sciences Center (HSC) T17-040, Stony Brook University Medical Center, Stony Brook, New York.

  • The State University of New York at Stony Brook holds patents in the fields of nebulizer development and inhaled drug delivery, which have been licensed to InspiRx.

  • Dr Smaldone is a consultant to InspiRx and is a member of the Advisory Board; Ms Cuccia serves as a consultant to InspiRx; and Dr Jayakumaran has disclosed no conflicts of interest. Hamilton Medical loaned equipment used in this study.

  • Internal funding supported this study.

  • Supplementary material related to this paper is available at http://www.rcjournal.com.

  • See the Related Editorial on Page 907

REFERENCES

  1. 1.↵
    1. Ehrmann S,
    2. Roche-Campo F,
    3. Sferrazza Papa GF,
    4. Isabey D,
    5. Brochard L,
    6. Apiou-Sbirlea G
    ; REVA research network. Aerosol therapy during mechanical ventilation: an international survey. Intensive Care Med 2013;39(6):1048-1056.
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. McPeck M,
    2. O’Riordan TG,
    3. Smaldone GC
    . Choice of mechanical ventilator: influence on nebulizer performance. Respir Care 1993;38(8):887-895.
    OpenUrl
  3. 3.↵
    1. Ari A,
    2. Fink JB,
    3. Dhand R
    . Inhalation therapy in patients receiving mechanical ventilation: an update. J Aerosol Med Pulm Drug Deliv 2012;25(6):319-332.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Dhand R
    . How should aerosols be delivered during invasive mechanical ventilation? Respir Care 2017;62(10):1343-1367.
    OpenUrlAbstract/FREE Full Text
  5. 5.↵
    1. Cuccia AD,
    2. Ashraf S,
    3. McPeck M,
    4. Samuel J,
    5. Smaldone GC
    . Wet-side breath-enhanced jet nebulization: controlling drug delivery during mechanical ventilation. Respir Care 2020;65(8):1077-1089.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Wang L,
    2. Guan C,
    3. Qin X,
    4. Qu Y
    . Effects of aerosol inhalation on respiratory mechanical parameters under different ventilation patterns and ventilator parameters. Zhonghua Wei Zhong Bing ji Jiu yi Xue 2018;30(11):1036-1040.
    OpenUrl
  7. 7.↵
    1. Li X,
    2. Tan W,
    3. Zhao H,
    4. Wang W,
    5. Dai B,
    6. Hou H
    . Effects of jet nebulization on ventilator performance with different invasive ventilation modes: a bench study. Front Med (Lausanne) 2022;9:1004551.
    OpenUrl
  8. 8.↵
    1. Peng Y,
    2. Dai B,
    3. Hu C-X,
    4. Su J,
    5. Tan W,
    6. Zhao H-W,
    7. Kang J
    . Which nebulizer position should be avoided? An extended study of aerosol delivery and ventilator performance during noninvasive positive pressure ventilation. Respiration 2018;95(3):145-153.
    OpenUrl
  9. 9.↵
    1. Perinel-Ragey S,
    2. Baboi L,
    3. Guerin C
    . Variability of tidal volume in patient-triggered mechanical ventilation in ARDS. Respir Care 2017;62(11):1437-1446.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    Instructions for Use Infinity Acute Care System: Evita Infinity V500 Ventilation Unit: Drager Medical; 2016.
  11. 11.↵
    Hamilton C-1 Operator's Manual. Reno, Nevada: Hamilton Medical AG; 2019.
  12. 12.↵
    1. Garnier M,
    2. Quesnel C,
    3. Fulgencio J-P,
    4. Degrain M,
    5. Carteaux G,
    6. Bonnet F,
    7. et al
    . Multifaceted bench comparative evaluation of latest intensive care unit ventilators. Br J Anaesth 2015;115(1):89-98.
    OpenUrlCrossRefPubMed
  13. 13.↵
    Servo-i Ventilator System V8.0, User’s Manual. Solna, Sweden: Getinge; 2018.
  14. 14.↵
    User Manual bellavista Ventilator. Mettawa, Illinois: Vyaire Medical; 2020.
Back to top

In this issue

 

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Related Articles

Cited By...

Keywords