Introduction with a common clinical conundrum
A patient is admitted to the ICU following attempted suicide with lorazepam. She is intubated for airway protection. Over three days in the ICU she regains consciousness and begins to require propofol for sedation. Whenever propofol sedation is lifted she is extremely agitated, making it impossible to perform a spontaneous breathing trial. Her chest radiograph is clear and she is on 30% FiO2. What is the best approach to this patient?
- (a) Resume propofol, try again tomorrow to wake her up and perform a spontaneous breathing trial.
- (b) Start dexmedetomidine infusion, attempt a spontaneous breathing trial while on dexmedetomidine.
- (c) Give a dose of intravenous antipsychotic (e.g. haloperidol or olanzapine), then attempt a spontaneous breathing trial again.
- (d) Stop propofol, extubate her when she becomes is agitated, and then re-assess and treat agitation as needed.
Vicious cycle: Intubated for agitation, agitated from intubation
The usual approach to extubation is to decrease the patient's sedation, perform a spontaneous breathing trial, and then extubate the patient if they pass the spontaneous breathing trial. Patients with hyperactive delirium may not do well with this approach. When sedation is decreased they may become agitated and tachypneic, thereby “failing” their spontaneous breathing trial. Sometimes, extubation may be avoided solely out of fear that the patient might become dangerously agitated afterwards.
Approaches to break the cycle
#1: Cowboy extubation (a.k.a., trial of extubation)
One approach is to stop the sedation and simply extubate the patient when they become agitated (without making any attempt at a spontaneous breathing trial). This aggressive “cowboy” approach does involve a risk that the patient could be agitated afterwards, which might be difficult to control without deep sedatives and an endotracheal tube.
In practice, I've found this approach to be simple and effective in selected patients (1). Although patients may look very agitated on the ventilator, removal of the endotracheal tube and restraints often improves comfort dramatically. Some patients will remain agitated following extubation, requiring titrated medication (e.g. IV antipsychotics, dexmedetomidine). However, such agitation can usually be managed without reintubation. Overall, the much-feared risk of uncontrollable post-extubation agitation doesn't seem to be an issue:
The main drawback of cowboy extubation is that it skips the spontaneous breathing trial. This could be dangerous in cases where the patient isn't ready for extubation due to respiratory insufficiency. If there is doubt about whether the patient is strong enough to breathe off the ventilator, cowboy extubation is contraindicated.
#2: Dexmedetomidine-facilitated extubation
Another approach to breaking this cycle is transitioning to dexmedetomidine as a sedative. Dexmedetomidine doesn't suppress respiration, so a spontaneous breathing trial and extubation may all be performed without stopping the dexmedetomidine (2). Dexmedetomidine thus creates a sedative “bridge” from intubation to extubation, avoiding agitation throughout the entire process (3). Once the patient is extubated, they are usually more comfortable and the dexmedetomidine can often be weaned off rapidly.
This generally is safe and effective. The main disadvantage is that it delays and complicates extubation, compared to simply removing the endotracheal tube. I was more enthusiastic about this a few years ago, but over time it has grown tedious. Most often a patient would fail a spontaneous breathing trial in the morning due to agitation, so we would start dexmedetomidine and extubate the patient in the afternoon, and then stop the dexmedetomidine hours later. Patients at low risk of respiratory insufficiency invariably passed their spontaneous breathing trial, implying that they would have done fine with a cowboy extubation.
Algorithm for breaking the intubation-agitation cycle
My general approach:
This was a RCT involving 74 intubated patients in whom the only barrier to extubation was agitation (“their degree of agitation was so severe as to make lessening their sedation and extubation unsafe”). Patients were randomized to receive dexmedetomidine vs. a placebo infusion (in addition to usual care, which most often involved sedation with propofol and an opioid).
A dexmedetomidine-facilitated extubation strategy was successful in increasing the ventilator-free time by 17 hours (the primary outcome). Among the secondary outcomes, dexmedetomidine caused a parallel reduction in time on the ventilator, and also reduced the duration of delirium:
For full details of the study please see The Bottom Line blog here.
Was ongoing intubation necessary?
It is debatable whether these patients actually required ongoing intubation. My usual approach to the sort of patients described in this study would be to perform immediate extubation (red box below). Sedation is decreased and the endotracheal tube is removed as soon as the patient becomes mildly agitated. This rarely leads to severe agitation, but rather the agitation usually improves as soon as the patient is extubated.
The extubation-failure rate of patients in this study was 1/71 (1.4%), which is very low. Meanwhile, nine patients (13%) underwent tracheostomy without ever receiving a trial of extubation. These data suggest an overly conservative approach to extubation, with patients spending excessive time on the ventilator.
Was dexmedetomidine necessary?
The timing of events suggests that the dexmedetomidine was not actually needed (table below). The day after study entry, most patients receiving dexmedetomidine were felt by their nurse to be ready for extubation. Extubation was performed shortly thereafter, and in most cases the dexmedetomidine was weaned off within a couple hours after extubation. Rapid discontinuation of dexmedetomidine following extubation implies that the cause of these patients' agitation was being intubated. Thus, extubation alone would have been sufficient to resolve the patient's agitation (regardless of the choice of sedative).
Timing of extubation was subjective
The primary flaw of this study was the influence of subjective assessments upon ventilator-free time (the primary outcome). The timing of extubation was “determined by senior ICU physicians, taking into account the assessments of bedside nurses.” There was no objective protocol for determining when patients were extubated. This makes the timing of extubation subjective, depending on how the nurses thought the patients looked. All other outcomes in the study (duration of ventilation, time to resolution of delirium, etc.), may have been driven by this subjective determination.
Overall it seems that these patients may not have required ongoing intubation at all. It is possible that their agitation was driven by stimulation from the endotracheal tube, for which the best remedy is extubation (not sedation). Most patients likely could have been successfully extubated whenever the ICU team decided to. Patients who received dexmedetomidine probably looked prettier on the ventilator, causing everyone to feel more comfortable extubating them sooner. Thus, the dexmedetomidine may have functioned as an anxiolytic for the ICU team, whereas it was not actually needed by the patient.
- Patients with agitated delirium may be difficult to extubate, leading to a vicious cycle of persistent intubation and delirium shown below.
- Dexmedetomidine may avoid agitation during the extubation process, because it may be continued throughout the entire spontaneous breathing trial and extubation procedure.
- Within the DahLIA study, dexmedetomidine reduced the duration of mechanical ventilation in patients when agitation was the only barrier to extubation.
- It is unclear whether patients in the DahLIA study truly required dexmedetomidine, or whether they could have been extubated immediately (with improvement in their agitation following removal of irritation from the endotracheal tube).
- The Bottom Line blog description of the DahLIA study. Lots of details were left out above, because The Bottom Line has it covered.
- Dexmedetomidine to facilitate noninvasive ventilation (PulmCrit)
- Delayed Sequence Intubation (EMCrit). In many ways, DSI is the reverse of dexmedetomidine-facilitated extubation.
- (1) This decision must be made on a patient-by-patient basis. In general, a cowboy extubation may be a viable approach if following criteria are met: (1) there is little concern that the patient will be strong enough to breathe on their own (2) there are no major concerns about re-intubating the patient if necessary (3) there is no pending procedure that would require the patient to lie still (e.g. MRI).
- (2) In contrast, propofol depresses the respiratory drive, so propofol could interfere with a spontaneous breathing trial.
- (3) This is basically the reverse of delayed sequence intubation. Delayed sequence intubation refers to sedation with an agent that doesn't suppress respiration (e.g. ketamine or dexmedetomidine) to facilitate preoxygenation and preparation for intubation. Dexmedetomidine-facilitate extubation refers to using dexmedetomidine to facilitate a spontaneous breathing trial and preparation for extubation.
Image credits: Cowboy image.