Acute respiratory failure (ARF) is characterised by hypoaxemia, which may be accompanied by hypoventilation, hypercapnia, or isolated acute hypoxaemic respiratory failure (AHRF). This condition results in an inability to deliver adequate oxygen to organs and remove CO2, leading to acid-base imbalances. Although supplemental oxygen can sometimes alleviate these issues, advanced respiratory support may be necessary for patients with higher ventilatory demands.
Non-invasive respiratory support (NRS) can help manage ARF by meeting patients’ needs without the need for invasive mechanical ventilation (IMV), potentially avoiding intubation. However, delaying IMV in critically ill patients may worsen their condition due to increased respiratory effort and oxygen consumption.
A recent review explores the physiological effects, indications, and considerations for NRS in ARF. It emphasises the importance of selecting appropriate devices and interfaces, understanding determinants and implications of NRS failure, and using monitoring tools to guide clinical decisions regarding ventilator settings and intubation.
NRS includes High-Flow Nasal Oxygen (HFNO), Continuous Positive Airway Pressure (CPAP), and Non-Invasive Ventilation (NIV). NRS failure is defined epidemiologically as the need for endotracheal intubation.
The link between delayed intubation following NRS failure and worse clinical outcomes is not clearly understood. Three possible pathways may explain this connection: (1) direct harm to the lung and respiratory muscles from excessive breathing effort (e.g., patient self-inflicted lung injury - P-SILI - and myotrauma), (2) progression of the underlying illness due to inadequate respiratory muscle unloading and low oxygen delivery, and (3) complications from intubation and IMV, such as sedation, immobilisation, diaphragm atrophy, and ventilator-induced lung injury.
To minimise the risk of NRS failure in patients with AHRF, a pragmatic approach to monitoring before and during NRS is recommended. The goals of respiratory monitoring during NRS are to (1) assess treatment response, (2) guide adjustments in ventilator settings to meet patients’ ventilatory demands and breathing patterns (e.g., increasing or decreasing support, correcting asynchronies), and (3) identify early which patients may require IMV. A multimodal approach should incorporate the patient’s initial clinical characteristics, illness severity, respiratory effort, gas exchange, and lung imaging to guide clinical decisions.
Dyspnoea and discomfort should be regularly assessed during NRS. Dyspnoea is often linked to high respiratory drive and has been independently associated with a higher risk of intubation and mortality in spontaneously breathing patients with ARF, requiring close monitoring. Discomfort related to the NIV interface is a key factor in NIV failure, emphasising the importance of proper interface selection and fitting.
Monitoring expired tidal volume (Vte) during NIV and CPAP can provide insights into respiratory effort and lung stress. A high Vte (≥9.5 mL/kg of predicted body weight) is linked to NIV failure and worse outcomes in patients with moderate-to-severe AHRF. However, clinical interpretation of a high Vte should consider factors beyond respiratory effort, such as high pressure support (PS) and respiratory system compliance, which can also lead to increased Vte.
Respiratory rate (RR) is a key variable in NRS but can be influenced by factors such as systemic inflammation, anxiety, and discomfort, making it a late sign of high respiratory drive. A high RR and a lack of initial decrease during NRS often predict failure, particularly with HFNO, but decisions should not be made based solely on RR. Finally, higher minute ventilation (Vte × RR) is also associated with NIV failure in AHRF and mild ARDS.
Oxygenation changes are crucial indicators of NRS success or failure in patients with ARF since they reflect both disease severity and response to therapy. Oxygenation is monitored continuously by pulse oximetry and intermittently via arterial blood gases, with a lower baseline PaO2/FiO2 ratio and lack of improvement over time predicting NRS failure. A lower SpO2/FiO2 ratio also correlates with HFNO failure within 24 hours. However, technical factors such as skin colour, perfusion, and anaemia can affect pulse oximetry readings.
Monitoring partial pressure of carbon dioxide (PaCO2) is essential, with transcutaneous monitoring offering a convenient alternative to blood samples for tracking trends in acute hypercapnic respiratory failure. Direct monitoring of breathing effort using esophageal pressure (Pes) provides valuable insights into ventilatory demands, respiratory muscle unloading, and the risk of patient self-inflicted lung injury (P-SILI).
Diaphragmatic ultrasound is a non-invasive method to monitor inspiratory effort by assessing diaphragm thickening fraction (Tfdi), with values below 36% suggesting early diaphragmatic dysfunction and a higher likelihood of NIV failure. The central venous pressure swing (ΔCVP) is another technique for identifying strong respiratory effort and titrating pressure support (PS). Nasal pressure swing (ΔPnose) is a minimally invasive alternative for monitoring inspiratory effort. A ΔPnose above 5.1 cmH2O during HFNO has been shown to predict the need for escalation in ventilatory support with high diagnostic accuracy.
The HACOR score (heart rate, acidosis, consciousness, oxygenation, and respiratory rate) is a tool developed to predict NIV failure in hypoxaemic patients. A score higher than 5 after 1 hour of NIV indicates a high likelihood of failure. The ROX index (SpO2/FiO2 ratio divided by respiratory rate) is another clinical score used to predict NRS failure. For patients on HFNO in AHRF, a ROX index below 4.88 suggests an increased risk of failure within the first 12–24 hours.
Other indices, like the VOX index (SpO2/FiO2 ratio to tidal volume) and the BREF score (base excess, respiratory rate, and PaO2/FiO2), may be useful for estimating inspiratory effort in settings with fewer monitoring tools, though they still require further validation in larger studies.
Lung ultrasound (LUS) is a non-invasive method that can quantify lung aeration and serve as a predictor of NRS success or failure. Electrical impedance tomography (EIT), though not widely available, provides a non-invasive way to assess lung aeration, estimate tidal volume, and monitor regional ventilation distribution.
Patients with ARF who receive NRS strategies and avoid intubation generally have better outcomes than those who require IMV, underscoring the importance of attempting NRS in appropriate patients. However, delaying intubation when NRS is not effective can increase mortality.
Improvement within the first 1-2 hours or clear worsening during NRS can help predict success or failure. Key indicators for escalation include high secretion burden, a low Glasgow Coma Scale score (<10), and failure to achieve proper interface fit despite adjustments. In cases of early improvement without further change within 24 hours, regular monitoring (e.g., every 4-6 hours) is recommended. Distinguishing between “delayed intubation” and “late intubation” is crucial, as patients who fail NRS and are intubated after 24–48 hours typically experience worse outcomes.
Non-invasive respiratory support has shown varied effectiveness in preventing the negative effects of invasive ventilation and improving outcomes for patients with acute respiratory failure. A careful evaluation of the patient’s initial condition, along with continuous physiological monitoring to adjust treatment settings, is crucial for personalising care and minimising the risk of harm or failure. In critically ill patients, particularly those with new-onset acute respiratory failure, it is essential to balance avoiding invasive ventilation with ensuring timely intubation when necessary.
Source: Critical Care
Image Credit: iStock
