Upstairs Care Downstairs: Delayed Sequence Intubation (DSI)
By Scott Weingart, MD
You have a 50-year-old male with bad bilateral pneumonia; BP 108/70, HR 96, RR 28. He is delirious, agitated, and looks sick, sick, sick! Saturation is 70% on a nasal cannula; when you try to place the patient on a non-rebreather (NRB), he just swats your hand away and rips off the mask.
It is obvious to everyone in the room that this patient needs intubation -- but the question is, how are you going to do it?
Your first impulse may be to perform rapid sequence intubation (RSI), maybe with some bagging during the paralysis period. This is essentially a gamble.
If you have first-pass success with RSI, you (and your patient) may just luck out, allowing you to get the tube in and start ventilation before critical desaturation and the resultant hemodynamic instability.
However, the odds are against you: Bagging during RSI predisposes to aspiration; conventional bag valve mask (BVM) without a positive end expiratory pressure (PEEP) valve is unlikely to raise the saturation in this shunted patient; and if there is any difficulty in first-pass tube placement, your patient will be in a very bad place.
A Better Way
Sometimes patients such as this one who desperately require preoxygenation will impede its provision.
Hypoxia and hypercapnia can lead to a state of delirium, causing these patients to rip off their NRB or noninvasive ventilation (NIV) masks.
This delirium, combined with the low oxygen desaturation on the monitor, often leads to precipitous attempts at intubation without adequate preoxygenation.
Standard RSI consists of the simultaneous administration of a sedative and a paralytic agent and the provision of no ventilations until after endotracheal intubation.1 This sequence can be broken to allow for adequate preoxygenation without risking gastric insufflation or aspiration; we call this method "delayed sequence intubation" (DSI).
DSI consists of the administration of specific sedative agents, which do not blunt spontaneous ventilations or airway reflexes, followed by a period of preoxygenation before the administration of a paralytic agent.2
Another way of thinking about the DSI method is to view it as a procedural sedation--with the procedure in this case being effective preoxygenation.
After the completion of this procedure, the patient can be paralyzed and intubated. Just as in a procedural sedation, we want patients to be calm but still spontaneously breathing and protecting their airway.
The ideal agent for this use is ketamine. This medication will not blunt patient respirations or airway reflexes. In addition, ketamine provides a dissociative state, allowing the application of preoxygenation.3
A ketamine dose of 1.0-1.5 mg/kg administered by slow intravenous push will produce a calmed patient within approximately 45 seconds.
Preoxygenation can then proceed in a safe, controlled fashion. This can be accomplished with an NRB, or preferably in a patient exhibiting shunt, by the use of a noninvasive mask hooked up to ventilator with a continuous positive airway pressure (CPAP) setting of 5-15 cm H2O.
After a saturation of greater than 95% is achieved, the patient is allowed to breathe the high FiO2 (fraction of inspired oxygen) oxygen for an additional 2-3 minutes in order to achieve adequate denitrogenation. A paralytic is then administered, and following the 45- to 60-second apneic period, the patient can be intubated.
In patients who have high blood pressure or tachycardia, the sympathomimetic effects of ketamine may be undesirable.
While these effects can be blunted with small doses of benzodiazepine and labetalol,3 a preferable sedation agent is available for these hypertensive or tachycardic patients. Dexmedetomidine is an alpha-2 agonist that provides sedation without any blunting of the respiratory drive or airway reflexes.4,5
A dexmedetomidine dose of 1 mcg/kg administered to the patient over a period of 10 minutes will lead to a sedated patient who will accept preoxgenation after 3-5 minutes in a majority of cases.
Another advantage of this medication is that, after the sedative agent is administered and the patient is placed on noninvasive ventilation, frequently the respiratory parameters improve so dramatically that intubation can be avoided.
In these cases, we then allow the sedative to wear off, and we reassess the patient's mental status and work of breathing.
If we deem that intubation is still necessary at this point, we can proceed with standard RSI by administering a conventional sedation agent (such as etomidate or additional ketamine) in combination with a paralytic, because the patient has already been appropriately preoxygenated.
A video demonstrating these concepts is available online at http://blog.emcrit.org/misc/preox.
- Walls RM, Murphy MF. Manual of emergency airway management, 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2008.
- Weingart SD. Preoxygenation, reoxygenation, and delayed sequence intubation in the emergency department. J. Emerg. Med. 2010 Apr 7. [Epub ahead of print].
- Aroni F, Iacovidou N, Dontas I, Pourzitaki C, Xanthos T. Pharmacological aspects and potential new clinical applications of ketamine: reevaluation of an old drug. J. Clin. Pharmacol. 2009;49:957-64.
- Carollo DS, Nossaman BD, Ramadhyani U. Dexmedetomidine: a review of clinical applications. Curr. Opin. Anaesthesiol. 2008;21:457-61.
- Abdelmalak B, Makary L, Hoban J, Doyle DJ. Dexmedetomidine as sole sedative for awake intubation in management of the critical airway. J. Clin. Anesth. 2007;19:370-3.
Dr. Weingart s an emergency intensivist from Mount Sinai School of Medicine and Elmhurst Hospital Center in New York. You can find Dr. Weingart's ED critical care blog as well as his podcast at http://blog.emcrit.org.