Q. What criteria do you use in clinical practice to determine if a patient is ready for extubation?

The first thing to assess is whether the underlying condition that led to intubation is resolving or has improved. We assess whether the patient is clinically stable; that they have a stable blood pressure and that their vital signs are stable, and we check that certain gas exchange parameters, such as PO₂ and PCO₂, are in an acceptable range. After we have done this in a preliminary way, we usually use an assessment called the spontaneous breathing trial, where the patient is allowed to breathe on their own without the ventilator. We have certain metrics that we use during that process, which identify whether the patient is ready to be liberated from the ventilator.

Terminology can be important. We talk about liberation rather than weaning these days, because liberation from the ventilator indicates that the person is being disconnected from the treatment device, whereas weaning suggests a process in moving the patient along. The current approach is to assess whether the patient is ready, and if they are, we remove them from the ventilator. There are many people who don’t require a process like this; there are people intubated for surgery or intubated temporarily for a procedure, who can be expeditiously extubated without any process. This process is more applicable to patients with a chronic medical illness, requiring mechanical ventilatory support.

So, we assess readiness to liberate by looking at clinical stability parameters, whether the cause of respiratory failure is improving, and then by using a spontaneous breathing trial as an assessment tool.

Q. How do you choose a type of spontaneous breathing trial to perform, and does this vary between different patient groups?

There is variance in how this is done. In the United States, we use continuous positive airway pressure (CPAP) or bilevel airway pressure (BiPAP) during the spontaneous breathing trial. The ventilator is set up so that the patient triggers the breaths spontaneously. During inspiration, the pressure in the airway in the tubing is augmented so that the patient can breathe more easily. Usually, we maintain a certain expiratory pressure as well.

This mechanism has several advantages; it supports the ability of the patient to perform the spontaneous breathing trial more easily and it also allows us to conserve the monitoring capabilities of the ventilator. We can switch the ventilator to a mode that allows this to happen and continue to use the monitoring capabilities of the ventilator during the spontaneous breathing trial. A potential disadvantage is that the patient may be more likely to be reintubated, if using support has made the process too easy and not representative of breathing off the ventilator. Patients who perform well on the test may not be able to sustain the same kind of respiratory cycle when they are disconnected from the ventilator, because they are no longer being supported.

In some places, particularly in Europe, they use the T-Piece trial, where the patient is disconnected from the ventilator and placed on supportive oxygen or some other supportive method to breathe without the pressure support from the machine. This type of spontaneous breathing trial has the advantage of detecting more patients who will be successfully liberated, without reintubation, but it makes it harder for them to perform the trial, potentially increasing the duration of mechanical ventilation. There are a couple of other methods of spontaneous breathing trial being explored; for example, there are innovative tools that are using high-flow nasal cannula or high-flow devices to support during a spontaneous breathing trial.

Q. How do you distinguish normal physiological changes during a spontaneous breathing trial from signs of genuine failure?

We expect that the patient is going to perhaps breathe more rapidly and more shallowly during a spontaneous breathing trial, and that may be normal. So, we use certain metrics that help us with that. The metric that has been used for 20 or 30 years is the frequency to tidal volume ratio, and that is the ratio of the respiratory rate to the tidal volume. The threshold value that we use is approximately 100. If the patient’s frequency to tidal volume ratio is over 100, we suspect that they’re not performing properly, and they won’t be able to pass.

There has been a lot of debate about the utility of this metric; it doesn’t precisely tell us which patients are going to be successful or not, and there is overlap between the groups who will fail and those who will succeed. Therefore, this metric should be considered in the context of the overall situation. Most clinicians now use the metric alongside other factors, including the patient’s ability to cough, their mental status, their overall clinical stability during the spontaneous breathing trial, their vital signs, and their gas exchange parameters. Today, most clinicians use a global assessment that incorporates multiple features.

To make the judgment of whether a patient can be safely disconnected or not requires the physician at the bedside, considering multiple factors in the context of the individual patient. We also know that despite our best efforts, there are certain patients who will fail and be re-intubated even after successfully passing their assessment. Most consider that a failure rate around 10–15% is acceptable, and the reason for that is because we want to be aggressive at getting people off the ventilator. We know that the longer a person is on the ventilator, the more their predicted mortality increases. So, we want to be aggressive, and we can accept a certain number of patients failing the assessment, but it is a difficult question that clinicians at the bedside deal with every day.

Q. When sedation is required, how do you prevent it from becoming a barrier to liberation?

When it comes to liberation from the ventilator, less is more. We know that patients do better when they have less sedation. Studies have shown that if you interrupt a patient’s sedation once a day, stopping it completely, they will be more likely to be liberated than those who have continuous sedation. We also know that patients who receive episodic sedation, when needed, do better coming off the ventilator sooner than those receiving continuous sedation. Protocolized sedation policies are also important; patients are more likely to be liberated quickly if we use a standardized protocol or a plan to extubate our patients. Those are really the hallmarks of the sedation programs.

Q. How do you identify and address reversible causes when liberation from mechanical ventilation is unsuccessful?

If a patient fails a spontaneous breathing trial, it’s important to identify any reversible causes. Firstly, it is important to look at the patient’s initial illness that led them to respiratory failure (e.g., pneumonia, heart failure), and identify whether it has resolved yet or if it needs more time. There are also other factors identified as very important.

Often patients have some degree of pulmonary oedema which impairs their ability to be disconnected. In this case, it is very important to be aggressive with diuresis and volume management in patients where you are considering disconnection. Assess patients to make sure that this does not become a reason for failure.

If a patient has rapid, shallow breathing, consider neuromuscular weakness as a cause. Measure respiratory muscle forces and maybe investigate the patient’s muscular strength using electromyography (EMG) techniques to learn if they have respiratory muscle weakness. It is also important to ensure the patient can trigger the ventilator, meaning they are able to initiate their own spontaneous breaths.

You should review all these possible causes to determine which is most likely affecting the patient and address it. We usually advocate doing daily spontaneous breathing trial assessments. In my ICU, we usually do a spontaneous breathing trial in the morning and if the patient fails, we plan interventions for the next 24 hours that are most likely to correct the reversible factors and try the spontaneous breathing trial again the following day.

Q. When is planned post-extubation respiratory support indicated, and how do you choose between non-invasive ventilation and high-flow nasal oxygen?

We certainly don’t want patients to be re-intubated after they have been liberated, and there are interventions that can prevent certain patients from being re-intubated. First of all, we know that post-extubation strategies are not necessary for everyone. There have been large studies, the biggest one being out of Vanderbilt University, demonstrating that not all patients in the ICU need support after extubation. There have also been a number of trials looking at patients who are most at risk for extubation failure, and in most studies it is patients who have chronic obstructive pulmonary disease, heart failure, or chronic respiratory acidosis with an elevated PCO₂. So, we target those groups for post-extubation strategies.

The first strategy used effectively was giving the patient non-invasive ventilation with a BiPAP mask after their extubation. The patient is placed on mask ventilation, and the ventilation is usually cycled between an expiratory and an inspiratory pressure. This is an effective way post-extubation of preventing re-intubation. Following the emergence of the high-flow nasal cannula, that has been looked at as well. What we know with the BiPAP mask strategy is, first of all, not all patients can use them: the patient has to have a level of consciousness that allows them to be able to take off the mask; if a person has injuries to the face or to the neck, they are not able to wear the mask; and there are some patients who are claustrophobic and will not tolerate the mask.

In comparing high-flow nasal cannula to non-invasive ventilation, it’s clear that the high-flow nasal cannula can be effective. Overall, looking at multiple studies, there have been some meta-analyses that suggest that mask ventilation is better than high-flow nasal cannula alone, but the difference is small, and some patients can use that. Other studies have shown us that a combination of the two methods might be the most effective. Patients don’t necessarily wear a mask 24-7, for example, they have to remove the mask to eat or drink and they might find the mask uncomfortable. A strategy that is especially effective is where the patient uses the non-invasive mask, particularly at night, and then during the daytime uses periods of high-flow instead of the mask.