Comparisons of respiratory mechanics in patients under different variations of noisy pressure support ventilation during ventilator weaning

Introduction

Mechanical ventilation and ventilator support are required for critically ill patients with respiratory failure. When a patient’s condition improves, he or she will be weaned from mechanical ventilation and returned to spontaneous breathing. It was reported that the time required for weaning could account for 42% of the total duration of mechanical ventilation (1). Optimizing the ventilation mode during the weaning process can improve the success rate of weaning and patient prognosis (2). Normal adults have large variations concerning the respiratory rate (RR), tidal volume (TV), inspiratory time (Ti), and expiratory time (Te) during spontaneous breathing (3-5). In patients under mechanical ventilation, a certain degree of respiratory variation might improve synchronization between patients’ spontaneous breathing efforts and the ventilator, leading to a successful weaning process (2).

Conventional mechanical ventilation only provides positive-pressure ventilation with constant parameter settings, which differs from spontaneous negative-pressure respiration and potentially results in ventilator-associated lung injury, ventilator-associated diaphragm dysfunction, and various types of patient-ventilator asynchrony. Noisy pressure support ventilation (PSV), also called variable PSV, is a relatively new mode of PSV (6). Compared with conventional PSV, noisy PSV increases the variability of TV extrinsically by randomly changing the pressure support (7). Some previous studies in animal models found that noisy PSV could improve lung recruitment in the gravity-dependent area and balance the pulmonary blood flow between the gravity-dependent and non-dependent areas to improve pulmonary function (8). Another study reported that noisy PSV improved TV variability in patients with difficult weaning and reduced patient-machine asynchronous events (9).

Despite its increasing clinical application, there are many unanswered questions concerning noisy PSV. For example, narrow pressure variation in noisy PSV might not provide adequate respiratory variation and support for the patient, whereas wide pressure variation can cause a mediastinal swing with large differences in compliance between the left and right lungs, resulting in hemodynamic instability. Wide pressure variation can also trigger alveolar stretch response, leading to abnormal switching of inhalation and exhalation and increased patient-ventilator asynchrony (10). Therefore, it is important to determine the appropriate degree of ventilation pressure variability to restore the patient’s respiratory function and facilitate weaning from mechanical ventilation.

Therefore, we investigated the effects of different levels of pressure variation during noisy PSV to identify the optimal pressure variation during the weaning process. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-78/rc).

Methods

Study design and participant selection

This is a self-controlled before-after study designed to observe the differences in respiratory mechanics among patients under different variations of noisy PSV. The aim of this study is to explore the optimal variation for patients during the weaning phase of mechanical ventilation.

The study enrolled patients in the intensive care unit (ICU) of the First Affiliated Hospital of Guangzhou Medical University (Guangzhou, China) from September 2020 to October 2022. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (Medical Research Ethics Review No. 2020-159), and all enrolled patients provided their informed consent.

The inclusion criteria were as follows: aged ≥18 and <80 years; receipt of invasive mechanical ventilation for more than 72 h; regained spontaneous breathing with the ability to tolerate PSV; and determined by the clinical physicians to be suitable for weaning. The exclusion criteria were as follows: inability to tolerate noisy PSV; contraindications for gastric pressure (Pga) measurement, including gastric or esophageal lesions and deformities, severe coagulation dysfunction, and severe gastric retention with transpyloric feeding; and any factors that could affect accurate pressure measurement, including pneumothorax, thoracic drainage tube, abdominal infection, intra-abdominal hypertension, or chest deformity.

Insertion of the pressure-measuring gastric tube

The insertion of the pressure-measuring gastric tube is performed by a qualified ICU nurse. The patient is positioned supine with the head of the bed elevated at approximately 30°. After adequate lubrication, the catheter is gently advanced through the nasal cavity. Upon reaching the pharynx (approximately 15 cm), the patient is instructed to swallow or assisted with chin tucking toward the sternum.

Localization of the pressure-measuring gastric tube

After the catheter reaches the estimated length, the transducer was calibrated and zeroed and the catheter connected to LabChart 8.0 data analysis software (ADInstruments, Dunedin, New Zealand). The following methods are employed for positioning verification:

• Pressure waveform analysis: the pressure transducer is connected to the catheter. Air is injected to inflate the esophageal and gastric balloons, and respiratory cycle-dependent waveforms of esophageal pressure (Pes) and Pga are observed.
• Abdominal compression test (for patients with minimal or without spontaneous breathing effort): gentle pressure is applied to the epigastrium, and an increase in Pga is monitored to confirm gastric placement.
• Expiratory occlusion test (for spontaneously breathing patients): an expiratory airway occlusion maneuver is performed. Synchronous negative deflection in Pes during inspiratory effort against occlusion confirms esophageal balloon position.
• Gastric aspiration and pH testing: gastric contents are aspirated via the catheter. pH verification using pH test strips is conducted if necessary.
• Auscultatory confirmation: a bolus of air (10–20 mL) is injected while auscultating the epigastrium for a gurgling sound, further confirming intragastric placement.
• Supplementary verification: bedside X-ray may be utilized if positional uncertainty persists.

Initial ventilator setting

Before the experiment, all participants received mechanical ventilation (Dräger infinity C500 ventilator, Dräger Medical GmbH, Lübeck, Germany) under the PSV mode with support pressure of 12 cmH₂O. Positive end-expiratory pressure (PEEP) and the inhaled oxygen concentration were maintained at the previous level together with a pressure rise time of 0.2 s and inspiratory peak pressure limited to within 40 cmH₂O.

Activation and parameter configuration of noisy PSV and experimental execution

The noisy PSV is a functional module superimposed on spontaneous breathing modes in Dräger ventilators. Under this module configuration, during each respiratory cycle, the ventilator delivers varying levels of pressure support randomly within the set variability range. Moreover, the pressure support level for every breath in this mode follows an approximately Gaussian distribution.

For example, with a baseline PS set at 12 cmH₂O and variation configured to 25% (pressure fluctuation: ±3 cmH₂O), the ventilator’s pressure support level will randomly oscillate between 9–15 cmH₂O, and this fluctuation conforms to a normal distribution, resulting in a pressure curve resembling “noise”.

At the beginning of the experiment, the ventilation mode for each enrolled patient was switched to noisy PSV. And the variation parameter of each enrolled patient in this mode was set in a randomized sequence to 0%, 15%, 25%, or 35%. Each parameter setting was maintained for one hour after every change. The order of these parameter settings for each enrolled patient was determined by random drawing conducted by the researchers. The ventilator parameter settings for PSV and noisy PSV modes are shown in Table 1.

After a ventilation parameter setting was maintained for one hour, the respiratory mechanics at this time were recorded by LabChart 8.0 software for no less than 3 min when the patient’s breathing status stabilized [clinicians confirmed that the patient did not have cough and the ventilator monitor displayed stable measurements for RR, airway pressure (Paw), and TV for 30 s]. Before switching to the next pressure parameter setting, the ventilator mode was always restored to conventional PSV until breathing was stable.

Data collection

We collected the patients’ baseline data, including sex, age, height, weight, underlying illnesses, Acute Physiology And Chronic Health Evaluation (APACHE) II score, and Sequential Organ Failure Assessment (SOFA) score.

Under each ventilator parameter setting, we recorded relevant data, including the rapid shallow breathing index (RSBI), RR, and total respiratory compliance (Cre). In addition, the following respiratory mechanics were monitored, calculated, and recorded by LabChart 8.0: Inspiratory trigger delay time, abnormal triggering per minute, asynchronous trigger index (AI), Paw, Pes, Pga, transdiaphragmatic pressure (Pdi), ratio of esophageal pressure to transdiaphragmatic pressure (Pes/Pdi), esophageal pressure swing (ΔPes), The mean value of the airway pressure-time product obtained by the PEAK module in LabChart 8.0 (PeakArea), Inspiratory trigger delay, pressure-time product (PTP), esophageal pressure-time product (PTPes), gastric pressure -time product (PTPga), transdiaphragmatic pressure-time product (PTPdi), ratio of ransdiaphragmatic pressure-time product to esophageal pressure-time product (PTPdi/PTPes), the coefficient of variation of esophageal pressure (CV_es), diaphragmatic electromyogram coefficient variation (CV_EMG), Ti, Te, average RR over 3 min after excluding interference (Real-RR), respiratory cycle time (Ttot), the respiratory time ratio (Ti/Te).

The calculation formulas are as follows:

AI=|NumberofAsynchronyEventsTotalRespiratoryCycles|×100%[1]

ΔPes=|Pes_end_expiration−Pes_end_inspiration|[2]

Pdi=Pga−Pes[3]

PTP=∫0TinspPmus(t)dt,Pmus=Pes−Pcw,Pcw=Ecw⋅VT[4]

PTPes=∫0TinspΔPes(t)dt[5]

PTPga=∫0TexpΔPga(t)dt[6]

PTPdi=∫0TinspPdi(t)dt[7]

CVes=StandardDeviationofPesMeanPes×100%[8]

CVEMG=StandardDeviationofEMGAmplitudeMeanEMGAmplitude×100%[9]

Patients were followed up, and their length of hospital stay, length of ICU stay, duration of mechanical ventilation, and success or failure of weaning were recorded.

Statistical analysis

The sample size was calculated in GPower software (version 3.1.9.7). From our pre-experiment data, we expected a baseline Cre of approximately 40 L/cmH₂O, whereas Cre values of approximately 40, 42, and 45 L/cmH₂O were expected for pressure variations of 15%, 25%, and 35%, respectively. To achieve 80% power (1 − β) with α of 0.05, we expected to enroll at least 16 patients. To account for a potential withdrawal rate of 25% and missing respiratory mechanics data, we aimed to enroll at least 20 patients.

Continuous variables were presented as the mean and standard deviation or median and interquartile range (IQR) depending on the normality test results. The differences in respiratory mechanics under different pressure variability levels (0%, 15%, 25%, and 35%) were analyzed. When the respiratory measurements of each group were normally distributed, one-way repeated-measures analysis of variance was used, and the model was adjusted for disease type, R (resistance), P0.1 (airway occlusion pressure), and intrinsic PEEP. If the sphericity assumption was not met, then Greenhouse-Geisser correction was applied to reflect the mean changes in respiratory mechanics among the different variability levels, and Bonferroni’s post hoc test was used for pairwise comparisons between different groups. When the respiratory measurements of each group were not normally distributed, the non-parametric Friedman test was used, and Bonferroni’s method was applied for post hoc pairwise comparisons. The categorical data were presented as numbers and percentages. Two-sided P<0.05 was considered statistically significant. All data were analyzed using SPSS 24.0 (IBM Corporation, Armonk, NY, USA).

Results

Participant characteristics

In this study, approximately 35 patients were screened for potential inclusion in the exploration. Among them, 5 patients dropped out of the study due to intolerance to noisy PSV, and 1 patient withdrew from the study because of transpyloric feeding.

In total, 29 patients, including 22 men (75.9%), were enrolled. Patients’ characteristics are listed in Table 2. The median duration of mechanical ventilation was 11 days (IQR, 8–30 days). Only 5 patients (17.2%) were successfully weaned from the ventilator.

Differences in respiratory mechanics under different pressure variation levels

After one hour of noisy PSV support under different pressure variation levels, there were significant differences in respiratory mechanics (Table 3). Noisy PSV significantly improved patients’ respiratory system compliance at variation levels of 25% and 35% (0% vs. 15% vs. 25% vs. 35%: 40 vs. 38.1 vs. 42.3 vs. 44.5 mL/cmH₂O, P=0.005). Further intragroup pairwise comparisons revealed that 25% variation was associated with higher compliance (25% vs. 0%, P=0.04; 25% vs. 15%, P=0.03). What’s more, the subgroup analysis revealed that compared with post-thoracic surgery and acute respiratory distress syndrome (ARDS) patients, Cre of chronic obstructive pulmonary disease (COPD) patients were significantly higher at 25% and 35% variation (P=0.03). At the same time, at variability levels of 25% and 35%, patients’ inspiratory trigger delay time was also significantly shorter (0% vs. 15% vs. 25% vs. 35%: 140 vs. 100 vs. 100 vs. 100 ms, P<0.001). Further intragroup pairwise comparisons revealed that 25% and 35% variation was associated with shorter inspiratory trigger delay time (25% vs. 0%, P=0.002; 35% vs. 0%, P=0.01). The PeakArea was significantly lower at variation 35% (8.8±2.5 cmH₂O·s). There was no statistical difference in the remaining indicators under the different variation levels.

Discussion

Our findings reconfirm the potential advantages of noisy PSV for mechanically ventilated patients during the weaning phase. At 25% and 35% variation levels, noisy PSV significantly improved respiratory system compliance while reducing patient-ventilator asynchrony events through shortened inspiratory trigger delay. Subgroup analysis further revealed that patients with COPD exhibited substantially greater improvements in respiratory system compliance under noisy PSV than postoperative thoracic surgery patients or ARDS patients. Our finding may provide some reference for the application and parameter setting of noisy PSV in clinical practice.

The role of noisy PSV in improving patient compliance has been mentioned in some previous studies. Boker et al. (11) found that in patients after abdominal aortic aneurysm surgery, the mean peak inspiratory pressure significantly decreased and lung compliance significantly improved after receiving noisy PSV. Guan et al. (12) used electrical impedance tomography (EIT) to monitor patients undergoing noisy PSV ventilation with different pressure variation levels, finding that noisy PSV increased gas distribution in the lung gravity-dependent area and improved lung homogeneity. However, these studies primarily focused on post-surgical and ARDS patients, without involving COPD patients. Our subgroup analysis further revealed that in COPD patients, the improvement in compliance with noisy PSV was particularly significant, suggesting that COPD patients may derive greater benefits from noisy PSV. This offers guidance for selecting target populations for noisy PSV implementation. The mechanisms by which noisy PSV improves compliance may include: The respiratory system can be considered a stochastic resonance system (5). Within a certain range, an increase in pressure can cause an increase in TV. In noisy PSV, a large TV can be generated when pressure support reaches the peak value. At the same time, because the time required for noisy PSV to achieve a larger TV is shorter than the time of alveolar derecruitment, pulmonary alveolar recruitment can be completed to reach better ventilation (13). When the variation level is relatively high (25–35%), the peak support pressure is also higher, and alveolar recruitment is more complete, which may explain the more pronounced improvement in patient compliance at this stage. In addition, the release of endogenous pulmonary surfactant and the resolution of pulmonary inflammation induced by noisy PSV might also contribute to improved lung compliance. Nevertheless, this study only collected and described respiratory system compliance over a short period. The intrinsic mechanisms by which noisy PSV improves lung compliance require further investigation (14-16).

Another finding of this study involves noisy PSV in improving patient-ventilator synchrony. In this study, it demonstrated as a significant reduction in inspiratory trigger delay time under noisy PSV support. However, discrepancies exist among previous studies: for instance, Spieth et al. (9) reported that even during short-term application (<1 hour) of noisy PSV at 30% variability in acute respiratory failure patients, patient-ventilator asynchrony events decreased significantly. Conversely, in a similar study on mild-to-moderate ARDS patients, Ball et al. (10) found that noisy PSV at 30% variation not only failed to improve oxygenation but also worsened the patient-ventilator synchrony. We speculate this discrepancy might relate to patients’ underlying pathophysiological differences. In our research, while noisy PSV significantly shortened inspiratory trigger delay time, no significant changes occurred in abnormal triggering events or the AI. This suggests noisy PSV may primarily reduce specific types of asynchrony events (e.g., trigger delay), but may not benefit patients with excessively strong respiratory efforts like ARDS. Conversely, for patients more prone to ineffective or delayed triggering (e.g., COPD), noisy PSV might offer greater advantages. This phenomenon merits further investigation.

In this study, we calculated the PeakArea. This measurement is similar to the traditional pressure-time product. It indirectly represents the work of respiratory system, but it has a longer period of data collection. We found that the Ppeak area was significantly lower when the variation was 35%. Under assisted ventilation mode, it may indicate that at variation level, the energy consumed by the patient’s respiratory muscles to overcome the ventilatory load is relatively small, and respiratory effort is relatively reduced. This phenomenon may be related to the improvements of compliance and patient-ventilator synchrony by noisy PSV mentioned above.

Notably, although this study demonstrates the advantages of noisy PSV at higher variation levels (25% and 35%), the risks associated with excessively high variation in clinical practice still worth caution. With excessively large variation, the initial movement of the patient’s diaphragm cannot fully adapt to the excessive variation, which can result in premature inhalation-exhalation switching. Meanwhile, excessively large variation can cause an intermittent high TV, thereby induce lung injury and trigger inhalation-exhalation switching through a self-protection mechanism. All of these events can cause the ventilator to stop delivering air prematurely and lead to respiratory distress. Of course, the underlying lung lesions and individual variability ultimately determine the optimal pressure level during noisy PSV. Combining EIT and other monitoring methods is required to study patients with different underlying diseases and levels of pulmonary function.

The current study still had multiple limitations. First, this was a single-center study of a small number of patients with different underlying illnesses, which could limit the generalizability of the results. Second, our patients received noisy PSV for a relatively short duration. The experimental results within the monitoring time might not fully reflect the effects of noisy PSV on the respiratory mechanics of these patients. Furthermore, because of the limited number of cases, statistical comparisons of clinical outcomes, such as the weaning success rate and ventilator-free days, under different variability rates were not performed. Finally, additional monitoring methods and measurements, such as EIT and blood gas analysis, are needed to further understand the effects of noisy PSV on patients.

Conclusions

Compared with conventional PSV, the pressure variation level of 25% or 35% might be optimal for noisy PSV to improve patient lung compliance and shorten the inspiratory trigger delay time during the weaning phase from mechanical ventilation. Further research is warranted.

Acknowledgments

We would like to thank Medjaden Inc. for its linguistic assistance during the preparation of this manuscript.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-78/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-78/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-78/prf

Funding: This study was supported by Rocket Medical Ltd. through the research project “Study on Respiratory Mechanics and Ventilatory Function through Re-titration of Respiratory Center Drive in ARDS Patients Receiving ECMO Combined with Prone Position Ventilation (PPV)” (signed on 2023-12-08).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-78/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (Medical Research Ethics Review No. 2020-159), and all enrolled patients provided their informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Fadila M, Rajasurya V, Regunath H. Ventilator Weaning. Treasure Island (FL): StatPearls Publishing; 2024.
2. Jhou HJ, Chen PH, Ou-Yang LJ, et al. Methods of Weaning From Mechanical Ventilation in Adult: A Network Meta-Analysis. Front Med (Lausanne) 2021;8:752984. [Crossref] [PubMed]
3. Bendixen HH, Smith GM, Mead J. Pattern of ventilation in young adults. J Appl Physiol 1964;19:195-8. [Crossref] [PubMed]
4. Tobin MJ, Mador MJ, Guenther SM, et al. Variability of resting respiratory drive and timing in healthy subjects. J Appl Physiol (1985) 1988;65:309-17. [Crossref] [PubMed]
5. Suki B, Alencar AM, Sujeer MK, et al. Life-support system benefits from noise. Nature 1998;393:127-8. [Crossref] [PubMed]
6. Abramovitz A, Sung S. Pressure Support Ventilation. Treasure Island (FL): StatPearls Publishing; 2024.
7. Mandyam S, Qureshi M, Katamreddy Y, et al. Neurally Adjusted Ventilatory Assist Versus Pressure Support Ventilation: A Comprehensive Review. J Intensive Care Med 2024;39:1194-203. [Crossref] [PubMed]
8. Fontela PC, Prestes RB, Forgiarini LA Jr, et al. Variable mechanical ventilation. Rev Bras Ter Intensiva 2017;29:77-86. [Crossref] [PubMed]
9. Spieth PM, Güldner A, Huhle R, et al. Short-term effects of noisy pressure support ventilation in patients with acute hypoxemic respiratory failure. Crit Care 2013;17:R261. [Crossref] [PubMed]
10. Ball L, Sutherasan Y, Fiorito M, et al. Effects of Different Levels of Variability and Pressure Support Ventilation on Lung Function in Patients With Mild-Moderate Acute Respiratory Distress Syndrome. Front Physiol 2021;12:725738. [Crossref] [PubMed]
11. Boker A, Haberman CJ, Girling L, et al. Variable ventilation improves perioperative lung function in patients undergoing abdominal aortic aneurysmectomy. Anesthesiology 2004;100:608-16. [Crossref] [PubMed]
12. Guan S, Zhang X, Zhang L, et al. Effect of different levels of pressure support variability on acute respiratory distress syndrome. Chin J Hygiene Rescue (Eletronic Edition) 2016;2:4.
13. Bellardine CL, Hoffman AM, Tsai L, et al. Comparison of variable and conventional ventilation in a sheep saline lavage lung injury model. Crit Care Med 2006;34:439-45. [Crossref] [PubMed]
14. Arold SP, Suki B, Alencar AM, et al. Variable ventilation induces endogenous surfactant release in normal guinea pigs. Am J Physiol Lung Cell Mol Physiol 2003;285:L370-5. [Crossref] [PubMed]
15. Ma B, Suki B, Bates JH. Effects of recruitment/derecruitment dynamics on the efficacy of variable ventilation. J Appl Physiol (1985) 2011;110:1319-26. [Crossref] [PubMed]
16. Spieth PM, Carvalho AR, Güldner A, et al. Pressure support improves oxygenation and lung protection compared to pressure-controlled ventilation and is further improved by random variation of pressure support. Crit Care Med 2011;39:746-55. [Crossref] [PubMed]