The Change of Central Vein Oxygen Saturation Level during Spontaneous Breathing Trial as a Predictor of Successful Extubation in Intensive Care Unit

1. INTRODUCTION

Mechanical ventilation is essential in managing critically ill patients in the intensive care unit (ICU). However, prolonged use increases the risk of complications such as ventilator- associated pneumonia, tracheal injury, stress ulcers, and sepsis, leading to higher morbidity and mortality [1]. Timely weaning is crucial, and predicting successful weaning poses a significant clinical challenge.

To determine readiness for weaning, a spontaneous breathing trial (SBT) is performed using a T-tube with minimal pressure support, assessing the patient’s ability to breathe independently after the acute phase of illness resolves. Success in SBT is evaluated through hemodynamic stability, mental status, and oxygenation efficiency. However, a successful SBT does not guarantee successful extubation, with failure rates as high as 30% in the first 2-3 days. More accurate predictors of extubation success are needed [1, 2].

The Rapid Shallow Breathing Index (RSBI) is the most commonly used parameter for predicting extubation success due to its simplicity and high sensitivity. An RSBI below 105 breaths/min/L is a common threshold indicating likely extubation success. However, a recent meta-analysis revealed moderate sensitivity (83%) but low specificity (58%), highlighting the need for additional predictive tools [2, 3].

Global tissue oxygenation status during SBT is another potential predictor. Increased breathing effort raises oxygen demand, reducing tissue oxygenation. Mixed venous oxygen saturation (SvO₂) and central venous oxygen saturation (SvO₂) are key indicators. While SvO₂ is more representative, it is invasive. SvO₂, requiring only a central venous catheter, is less invasive and more practical. Studies suggest that greater decreases in SvO₂ during SBT may indicate extubation failure, while smaller decreases suggest success. Despite this, SvO₂ is not widely used due to limited data on its optimal cut-off values and prognostic accuracy [1, 2].

This study aims to evaluate whether changes in SvO₂ during SBT can predict successful extubation in ICU patients.

2. METHODS

3. RESULTS

Forty-two eligible patients were involved in this study, with a mean age of 42.2±15.34 years. The majority of patients (37/42; 89.1%) were successfully extubated. Detailed demographic and clinical baseline characteristics are shown in Table 1. There was no significant difference in baseline characteristics between successfully extubated patients and those who failed, except for higher survival status in successfully extubated patients (100% vs. 60%, p=0.012).

There was a significant difference in RR, SaO₂, and MAP between successfully extubated patients and those who failed at the first minute and after 30 minutes of SBT (Table 2). The baseline PaO₂ and ScvO₂ were also not significantly different between successfully extubated patients and those who failed. However, when the change of PO₂ level (ΔPaO₂) and change of ScvO₂ level (ΔScvO₂) after 30 minutes of SBT were compared between groups, patients who were success- fully extubated significan- tly had a lower decrease of PaO₂ (ΔPaO₂ -2.54±1.88 vs -8.6±9,42; p=0.016) and ScvO₂ (ΔScvO₂ -2.89±1.63 vs -8.20±4.27; p=0.049) (Table 3). Thus, a decrease in ScvO₂ could be used to discriminate successfully extubated patients from patients more likely to fail.

Further analysis using the ROC curve (Fig. 1) showed that the most optimal ΔScvO₂ cut-off to predict successful extubation in critically ill patients was 4.5%. Thus, a decrease of ScvO₂ less than 4.5% is associated with a higher probability of successful extubation (Table 4). This cut-off had an area under curve (AUC) of 0.897, sensitivity of 81.1%, specificity of 60%, positive predictive value (PPV) of 93.8%, and negative predictive value (NPV) of 70% (Table 5).

4. DISCUSSION

A decreased ScvO₂ level by less than 4.5% during SBT could predict successful extubation with high sensitivity (81.1%) and moderate specificity (60.0%). Both groups also had comparable comorbidities, mechanical ventilation duration, and even similar blood gas analysis and other laboratory parameters at baseline. This indicated that the change in ScvO₂ level during SBT could discriminate successfully extubated patients from patients more likely to fail even if they had similar clinical characteristics. Thus, our study suggests that a decrease in ScvO₂ level of less than 4.5% during SBT could be utilized in clinical practice as a guide for the mechanical ventilation weaning process in critically ill patients.

Previous studies also proved that ΔScvO₂ during SBT could predict successful extubation in difficult-to-wean patients [4]. Ashmawi et al. also reported that a decrease of ScvO₂ of more than 3.8% is significantly associated with extubation failure. Even though their cut-off was slightly different, this study showed that a decrease in ScvO₂ level had a high sensitivity (89.74%), specificity (90.91%), and PPV (97.22%) but with moderate NPV (71.43%) [5]. Mallat et al. also found that ΔScvO₂ was an independent predictor of extubation success in patients with good SBT tolerance. A decrease of ScvO₂ ≥5.4% had an excellent NPV and a moderate PPV in predicting extubation failure [6]. Teixeira et al. also found that a greater decrease in ScvO₂ level during a T-tube trial could predict extubation failure in 86% of cases. Meanwhile, patients who were successfully extubated had minimal change in ScvO₂ level [7]. They found that the extubation failure and re-intubation requirement rate significantly increased when ScvO₂ decreased to more than 4.5% (sensitivity 88%, specificity 95%, PPV 0.93, and NPV 0.90) [8]. Shalaby et al. also stated that the risk of re-intubation was significantly higher when the decrease in ScvO₂ level >5% (sensitivity 87%, specificity 90%, PPV 0.78, and NPV 0.93) [7].

Oxygen consumption (VO₂) depends on the delivery of oxygen (DO₂) and oxygen extraction ratio (O₂ER). Therefore, increased DO₂ will increase VO₂ levels. During the transition from mechanical ventilation to spontaneous breathing, the respiratory muscle will work harder, increasing VO₂ [9, 10]. Thus, CO and DO₂ will also increase as compensation to offset the increased demand for respiratory muscle VO₂. Increased VO₂ demand during this transition process will cause decreased O₂ supply to cells, resulting in decreased global tissue oxygenation muscle [11-13]. In addition to the effects of increased VO₂, other factors, such as laryngeal oedema and low muscle endurance, are some of the potential causes for extubation failure, which are not directly related to oxygen consumption and use, so these results may be, to some extent, difficult to explain [4]. These confounding factors (mental status alteration and inadequate cough) were excluded from our study.

The respiratory rate, MAP and SaO₂, in the successful and unsuccessful group during the first minute and after 30 minutes show a significant difference statistically, but clinically it’s difficult to precisely predict the failure of extubation using only a small difference in respiratory rate (16.30±3.09 vs. 20.20±4.27), SaO₂ (99.7±0.62 vs. 98.20±2.17) and MAP (84.14±11.86 vs. Line 187-19196.40±12.22). An Intensivist should use not one but several parameters, from patient’s vital signs, clinical conditions, and laboratory parameters that show tissue perfusion to predict the failure of extubation.

ScvO₂ is used as a surrogate of SvO₂, which is more invasive, high-risk, and expensive because it requires pulmonary catheter insertion. Now, ScvO₂ has become part of standard care for critically ill patients. The previous study showed a significant correlation between ScvO₂ and SvO₂ [14, 15]. Therefore, ScvO₂ measurement is also a reliable indicator for global tissue oxygenation status, but it is easier, safer, and affordable in clinical practice.

The outcomes of successfully extubated patients and patients who failed significantly differed. A higher survival rate was found in groups of successfully extubated patients (100% vs. 60%, p=0.012). This emphasizes that proper timing of extubation in patients ready for weaning is critical, and failure of extubation significantly affects the morbidity and mortality of critically ill patients. Therefore, we highly suggest using the change in ScvO₂ level as a guide to predict successful extubation in critically ill patients because successful SBT does not always mean successful extubation.

This study also had limitations. It was a non-randomized study with a limited number of participants, which may limit the generalization of its results. However, it involved both surgical and medical critically ill patients. Another limitation is the absence of hemodynamic measurements to assess CO and also no SvO₂ measurement as a gold standart test, due to the limited use of right heart catheterization. Nevertheless, the results of this study still add important points regarding the role of ScvO₂ measurements during SBT to guide clinicians in selecting patients who are likely to be successful or unsuccessfully extubated.

CONCLUSION

The change of ScvO₂ during SBT was a good predictor of successful intubation in critically ill patients in the ICU. A decrease of ScvO₂ less than 4.5% after 30 minutes of SBT is associated with a higher probability of successful extubation, with a sensitivity of 81.1%, specificity of 60%, positive PPV of 93.8%, and NPV of 70%. An Intensivist should use not one, but several parameters, from patient’s vital signs, clinical condition, laboratory parameters that show tissue perfusion to predict the failure of extubation. We suggest using this new parameter as an additional judgment in selecting patients in the ICU who are likely to be successfully extubated.

REFERENCES