Comparison of clinical outcomes in critical patients undergoing different mechanical ventilation modes: a systematic review and network meta-analysis

Mengyu Wu, Xiaohong Zhang, Yu Jiang, Yun Guo, Wenjing Zhang, Hong He, Yanhua YinFrontiers in Medicine 2023 · DOI

This systematic review and network meta-analysis, published in Frontiers in Medicine, evaluates how different mechanical ventilation modes impact clinical outcomes for critically ill patients. By analyzing 28 randomized controlled trials (RCTs), the researchers found that while most modes result in similar lengths of stay, Neurally Adjusted Ventilatory Assist (NAVA) significantly reduces ICU mortality, and Proportional Assist Ventilation (PAV) improves the success rate of weaning patients off ventilators.

TL;DR

A network meta-analysis of 3,189 patients reveals that NAVA ventilation reduces ICU mortality compared to traditional pressure support, while PAV significantly increases the odds of successful ventilator withdrawal. Despite these benefits, the choice of ventilation mode does not appear to significantly alter the total duration of mechanical ventilation or hospital stay.

What Problem Does This Paper Address?

Mechanical ventilation is a cornerstone of ICU care, but it is a double-edged sword. While it maintains gas exchange, long-term invasive ventilation can cause lung injury, diaphragmatic atrophy, and infections. A major clinical challenge is "weaning"—the process of transitioning a patient off the machine.

There are numerous ventilation "modes" (control algorithms that determine how the machine responds to a patient's breath), such as Pressure Support Ventilation (PSV), Synchronized Intermittent Mandatory Ventilation (SIMV), and newer "intelligent" modes like NAVA and PAV. However, there is no global consensus on which mode is superior. Individual trials are often too small to show mortality benefits, and many modes have never been compared head-to-head. This paper uses a network meta-analysis (NMA) to synthesize existing data and provide a clearer hierarchy of effectiveness.

Methods

The authors conducted a systematic review of major medical databases (PubMed, Embase, etc.) up to November 2022. They focused on RCTs comparing at least two ventilation modes in adult ICU patients.

The "Network" Approach: Standard meta-analyses only compare "A vs. B." A Network Meta-Analysis (NMA) allows for indirect comparisons. If Study 1 compares Mode A to Mode B, and Study 2 compares Mode B to Mode C, the NMA uses Bayesian statistical modeling (via R and JAGS software) to infer the relationship between A and C.

Modes Evaluated: * PSV: The traditional standard. * NAVA: Uses a nasogastric probe to sense electrical signals from the diaphragm, triggering the ventilator via the patient's own neural drive. * PAV: Adjusts pressure support in real-time based on the patient’s instantaneous inspiratory flow and volume. * ASV, SIMV, Smartcare/PS, and PSV_ATC: Various other automated or synchronized modes.

Datasets

The study analyzed a pooled dataset from 28 RCTs involving a total of 3,189 patients. The patients were primarily admitted to the ICU for conditions like Acute Respiratory Distress Syndrome (ARDS) and Chronic Obstructive Pulmonary Disease (COPD).

Key Insights and Ideas

  • The Superiority of Neural Coupling: NAVA emerged as the most effective mode for reducing mortality (94% SUCRA ranking). Because NAVA triggers breaths based on neural impulses rather than air pressure changes, it minimizes "trigger delay" and improves man-machine synchronization. This likely prevents the respiratory muscle fatigue and lung injury associated with traditional modes.
  • PAV for Weaning: PAV was the clear winner for "ventilator withdrawal" success. By providing support that is proportional to the patient's effort, it prevents the machine from "over-supporting" the patient, which helps maintain respiratory muscle tone and makes the transition to independent breathing easier.
  • The Duration Paradox: Interestingly, none of the modes significantly reduced the duration of mechanical ventilation or the length of ICU/hospital stays compared to PSV. This suggests that while some modes are "safer" (lower mortality), the overall recovery timeline may be dictated more by the underlying pathology than the ventilation algorithm itself.

Key Results

  • ICU Mortality: NAVA significantly reduced mortality compared to PSV [OR = 0.63, 95% CI (0.43, 0.93)] and compared to PAV [OR = 0.45, 95% CI (0.21, 0.97)].
  • Ventilator Withdrawal: PAV significantly improved the success rate of weaning compared to PSV [OR = 3.07, 95% CI (1.21, 8.52)].
  • Stay Duration: No statistically significant differences were found between NAVA, SIMV, ASV, PAV, Smartcare/PS, and PSV_ATC regarding the duration of mechanical ventilation or ICU stay.
  • Ranking (SUCRA):
    • For mortality reduction: NAVA (94%) > SIMV (54%) > PSV (52%).
  • For weaning success: PAV (89%) > NAVA (58%) > ASV (40%).

Limitations and Future Work

The authors acknowledge several caveats:

  1. Risk of Bias: Many included studies did not explicitly describe allocation concealment, and blinding is inherently difficult in ventilation trials (clinicians can see the machine settings).
  2. Publication Bias: Statistical tests suggested a potential bias toward positive results in the published literature.
  3. Heterogeneity: The patient population was diverse (e.g., combining COPD and ARDS patients), and different diseases may respond better to specific modes.
  4. Need for Scale: While the NMA provides a strong signal, the authors call for large-scale, high-quality multicenter RCTs to confirm if NAVA should become the new clinical standard for reducing mortality.

Summary updated

Abstract

NAVA can reduce mortality in ICU, and PAV may increase the risk of withdrawal of the ventilator. There was no significant difference between PSV and other mechanical ventilation modes (NAVA, SIMV, AVS, PAV, Smartcare/PS, and PSV_ATC) in the duration of mechanical ventilation, duration of ICU stay, or hospital stay. Due to the limitations, more high-quality studies are needed to verify these findings.
BibTeX
@inproceedings{wu2023comparison,
  author = {Mengyu Wu and Xiaohong Zhang and Yu Jiang and Yun Guo and Wenjing Zhang and Hong He and Yanhua Yin},
  title = {{Comparison of clinical outcomes in critical patients undergoing different mechanical ventilation modes: a systematic review and network meta-analysis}},
  booktitle = {Frontiers in Medicine},
  year = {2023},
  doi = {10.3389/fmed.2023.1159567},
}
Figure