Role of a successful spontaneous breathing trial in ventilator liberation in brain-injured patients

Zhonghua Shi, Annemijn H. Jonkman, Pieter R. Tuinman, Guangqiang Chen, Ming Xu, Yan-Lin Yang, Leo Heunks, Jian-Xin ZhouAnnals 2021 · DOI

TL;DR

This study investigated whether physiological parameters measured during a spontaneous breathing trial (SBT) could predict successful ventilator liberation in brain-injured patients. The findings suggest that while SBTs are standard practice, they may not be sufficient predictors in this specific population, as a significant portion of patients still require re-ventilation. This highlights the need for a more comprehensive approach to weaning in brain-injured individuals.

What Problem Does This Paper Address?

Critically ill patients often require mechanical ventilation, but prolonged use can lead to complications. Therefore, timely liberation from the ventilator is crucial. Spontaneous Breathing Trials (SBTs) are a common method to assess a patient's readiness for extubation. However, brain-injured patients present unique challenges. Their indications for intubation and ventilation differ from other critically ill patients, and the effectiveness of SBTs in predicting successful liberation in this group is less established. Previous studies have indicated higher extubation failure rates in brain-injured patients compared to general ICU populations. This paper aims to explore whether detailed physiological measurements during an SBT can differentiate between successful and failed ventilator liberation in patients with brain injury.

Methods

The study employed a prospective design, enrolling adult patients with brain injury who had been mechanically ventilated for over 48 hours. These patients underwent a 30-minute SBT. During the SBT, several key respiratory variables were continuously monitored: airway pressure, airflow, esophageal pressure, and diaphragm electrical activity (EAdi). These measurements were taken at baseline (before the SBT) and at various time points during the SBT. From these raw data, parameters such as respiratory rate, tidal volume, inspiratory muscle pressure (ΔPmus), and neuromechanical efficiency (NMEoccl) were calculated breath-by-breath. The study then compared these physiological parameters between two groups: those who were successfully liberated from the ventilator within 48 hours of the SBT and those who failed liberation (requiring re-ventilation).

Datasets

Not discussed in the paper. The study collected data from patients directly.

Key Insights and Ideas

  • SBTs may be less predictive in brain-injured patients: Despite a successful SBT, a substantial number of brain-injured patients still require re-ventilation, suggesting that the physiological readiness assessed by SBTs might not fully capture the complexities of weaning in this population.
  • Neurological factors are critical: The paper emphasizes that factors beyond respiratory function, particularly neurological status, play a significant role in ventilator liberation success for brain-injured patients.
  • Standard physiological parameters are not discriminative: The detailed analysis of breathing patterns and respiratory muscle effort during the SBT did not reveal significant differences between successful and failed liberation groups, challenging the utility of these specific metrics for prediction in this cohort.
  • Brainstem involvement doesn't necessarily predict failure: Interestingly, the study found no association between the anatomical location of the brain injury (including brainstem involvement) and the rate of liberation failure.

Key Results

  • Out of 46 patients who completed the SBT, 37% required invasive ventilator support within 48 hours, and 61% within 7 days.
  • There were no significant differences in baseline characteristics between the successful and failed liberation groups.
  • In-depth analysis of respiratory physiological parameters (breathing pattern, muscle effort) did not show differences between the groups during the SBT.
  • The Glasgow Coma Scale (GCS) score at the time of ventilator disconnection was similar in both successful and failed liberation groups, and even patients with a GCS <8 had a reasonable chance of successful liberation.
  • Arterial HCO3- was significantly higher in the successful liberation group compared to the failed group at the start of the SBT (28.4±2 vs. 26.1±2.7 mmol/L, P=0.001).

Limitations and Future Work

The authors acknowledge several limitations:

  • Single-center study: This may limit the generalizability of the findings.
  • No formal sample size analysis: The study's explorative nature means the sample size was comparable to similar studies, but future research might benefit from a formal calculation.
  • Definition of liberation: Not all patients were extubated immediately after the SBT; some remained on a T-tube circuit. The study acknowledges this distinction but found comparable liberation failure rates between early extubated and non-extubated patients.

Future work should focus on:

  • Exploring non-respiratory parameters: Given the high failure rate and the limited predictive value of respiratory parameters, further studies should investigate other factors, particularly neurological parameters, to predict liberation success.
  • Developing new assessment tools: There's a need for novel tools to assess underlying neurological conditions more effectively before making liberation decisions.
  • Investigating specific SBT strategies: The study used CPAP only, and future research could explore the impact of different SBT strategies (e.g., pressure support) on outcomes in brain-injured patients.

Summary updated

Abstract

The trial was registered at ClinicalTrials.gov (No. NCT02863237).
BibTeX
@article{shi2021role,
  author = {Zhonghua Shi and Annemijn H. Jonkman and Pieter R. Tuinman and Guangqiang Chen and Ming Xu and Yan-Lin Yang and Leo Heunks and Jian-Xin Zhou},
  title = {{Role of a successful spontaneous breathing trial in ventilator liberation in brain-injured patients}},
  journal = {Annals},
  year = {2021},
  doi = {10.21037/atm-20-6407},
}
Figure