know when to extubate

Most patients who are receiving mechanical ventilation don’t need to be weaned—they need to be liberated. Remember that the ventilator doesn’t do anything in particular to cure the patient, which is why most studies on different modes of ventilation haven’t shown one to be superior to the others. It’s not about the ventilator or its settings. When the reason for intubation has begun to resolve, it’s time to start assessing readiness for extubation.


Every patient on a ventilator should be assessed for potential extubation on a daily basis. The following are reasons not to do a spontaneous breathing trial:

  • Hemodynamic instability (more than one vasopressor, or a high dose of a single pressor)

  • Coma or deep stupor—a patient doesn’t have to be wide awake to be extubated, but he should be able to cough and handle his secretions

  • An unstable airway

  • FiO2 > 50% or PEEP > 8 cm H2O needed to keep SpO2 ≥ 88%

  • Persistent ventilatory failure (high spinal cord lesion, neuromuscular disease, etc)

  • A planned procedure or trip to the OR that day


We do our spontaneous breathing trials using Pressure Support Ventilation. The respiratory therapists are empowered to do the screening (using the above criteria) and initiate the spontaneous breathing trial without a specific order. The usual procedure is to:

  1. Set the vent to PSV, with a PS of 7 and a CPAP of 5

  2. Hold sedation and tube feeding during the trial

  3. Allow the patient to breathe on this mode for 30-60 minutes

  4. If the f/VT ratio is ≤ 80, extubate the patient


Three factors have been identified as predicting extubation failure in patients who pass the spontaneous breathing trial. These are:

  • Cough velocity < 60 L/min

  • Inability to follow four simple commands

    • Open eyes

    • Follow an observer with eyes

    • Grasp fingers

    • Stick out tongue

  • Secretions > 2.5 mL/hour

If none of these are present, the likelihood of successful extubation is around 97%. If all three conditions are present, then the likelihood of successful extubation is very low (0% in the study, probably < 3% in real life). If one or two are present, you can still extubate but be on the lookout for signs of failure and the need for reintubation.


An ABG is not part of the usual procedure, but can be obtained if there’s concern about occult hypercapnia. Things like the maximal inspiratory pressure or forced vital capacity are not routinely measured, but can be if you feel it will help. The most important part of the spontaneous breathing trial is to make a decision after the 30-60 minutes on PSV! If the f/VT ratio is ≤ 80, the patient should be extubated most of the time. There is no need for prolonged PSV trials, and there is no benefit to “working him out”. Either extubate, or put the vent back to assist-control and try again tomorrow.


T-piece trials are another method of assessing readiness for extubation, and may be better than PSV in patients with systolic heart failure or neuromuscular disease. Still, make the decision to extubate or not after 30-60 minutes.