Wednesday, June 4, 2014

Rapid Sequence Termination (RST) of status epilepticus

Co-authored with Ryan Clouser (@neurocritguy).  


Status Epilepticus is our favorite neurologic emergency.   If managed correctly, patients will often have excellent neurologic outcomes and short ICU stays.   Incorrect management increases risk of brain damage, epilepsy, aspiration, and death.   Treatment really matters.     

The traditional approach to status epilepticus is based on targeting seizure control within 60 minutes of onset (Brophy 2012).   However, complications including neuronal loss, aspiration, hyperthermia, and myocardial ischemia may occur earlier than 60 minutes.   Rapid sequence termination (RST) is a streamlined approach to terminate status epilepticus in under 30 minutes.  

How long is it safe to leave a patient in generalized convulsive status epilepticus?

This is unclear.   Neuronal loss may begin after 30 minutes.   The physiology of status epilepticus involves an initial sympathetic surge with brain hyperperfusion lasting 30 minutes, followed by a drop in blood pressure.   After 30 minutes there is cerebral metabolic decompensation with impaired cerebral perfusion and a reduction in brain oxygen and glucose levels, causing brain injury (Chapman 2001).   Early intervention is also important because over time the seizure becomes more refractory to treatment, and the risk of recurrence increases.  

From a non-neurologic standpoint, a patient in generalized convulsive status epilepticus is extremely unstable with an unprotected airway and severe hemodynamic stress.   Over time there is increasing risk of hyperthermia, rhabdomyolysis, hyperkalemia, lactic acidosis, hypoglycemia, aspiration, myocardial ischemia, arrhythmia, and cardiac arrest.

A prospective randomized controlled trial showed that pre-hospital benzodiazepine administration improved seizure control and reduced mortality from 15% to 6% (1).   Although there are many studies correlating seizure duration with poor outcomes, this study proves that early intervention saves lives.  

We have the ability to control the seizure early, although this may require intubation.   The real issue, then, is weighing the benefit of early seizure control vs. the risks of intubating a patient may have avoided intubation.   With modern-day critical care, the risks of intubation are low.   As long as the seizure is controlled promptly and the patient has no complications, such patients are generally easy to extubate the next day.

Facing an unstable patient with ongoing neurologic and physiologic stress, it may be safest to pursue early seizure control, ideally in under 30 minutes.   We would prefer to err on the side of intubating a few extra patients, rather than have patients slip into refractory status epilepticus, aspirate, or suffer cardiac complications.

Achieving seizure control in under 30 minutes requires a new logistic approach.

The traditional approach to status epilepticus is based on trialing a series of anti-epileptic drugs prior to intubation and sedation.   Although this looks nice on paper, it’s difficult to achieve rapidly.   Consider the time required to order an anti-epileptic, receive it from pharmacy, infuse it into the patient (and, in the case of fosphenytoin, metabolize the prodrug into active drug) – this is difficult to achieve within 20-25 minutes after seizure onset.   And of course, there is no guarantee the anti-epileptic will be effective: second- and third-line antiepileptic agents often fail to control the seizure.

The most reliable way to achieve seizure control within 20-30 minutes is based on benzodiazepines, propofol, and ketamine – drugs which are immediately available in the ED/ICU and may be rapidly bolused.   (On the hospital ward an emergent call to anesthesia will bring an anesthesiologist running to the bedside with these medications and airway equipment – Perfect!).  

Rapid Sequence Termination (RST)

In order to achieve seizure control quickly, the first-line antiepileptic is a benzodiazepine and the second-line antiepileptic is propofol.   Propofol is a powerful antiepileptic agent, and when bolused early in the course of status epilepticus is extremely effective.   This is followed immediately by a propofol infusion to prevent seizure recurrence.   As status epilepticus evolves, the GABA receptors which propofol acts on decrease in number and the seizure may become less responsive to propofol.  

If possible, an anti-epileptic agent (i.e. phenytoin, levtiracetam, valproic acid) should be given at the five-minute timepoint.   After the patient has seized for five minutes they meet the current definition of status epilepticus and require a maintenance anti-epileptic agent (Brophy 2012).   There is little to be gained by delaying anti-epileptic therapy because it will be required regardless of whether the seizure responds to benzodiazepine.   However, in reality the anti-epileptic agent will generally arrive somewhere between 15-25 minutes after the patient started seizing.   As discussed above, to logistically achieve seizure control within 20-30 minutes the algorithm must proceed without delaying to await the anti-epileptic agent.    The RST algorithm is designed to work with or without the anti-epileptic agent, and in many cases the seizure will be controlled before the anti-epileptic is actually given.

Ketamine blocks NMDA receptors, giving it anti-epileptic and neuroprotective properties.   Combining ketamine with propofol for rapid-sequence intubation (“ketofol”) may improve the blood pressure and seizure control.   As status epilepticus evolves, the number of NMDA receptors increases, making ketamine theoretically more effective.   Combining propofol with ketamine may provide synergistic activity and improved efficacy throughout the evolution of status epilepticus (i.e., early on the propofol may be more effective than the ketamine, whereas later on the ketamine may do more of the work).   Unfortunately, there is no evidence regarding the use of ketamine this early in the course of status epilepticus (evidence pertains solely to refractory status epilepticus Dorandeu 2013).   Extrapolating available data, the benefits of ketamine likely outweigh potential risks, so administering ketamine in combination with propofol during intubation is reasonable.  

Propofol is generally not a great sedative for rapid sequence intubation of critically ill patients given its tendency to cause hypotension (vis a vis the Propofol Assasins).   However, this is one situation where the benefits of its anti-epileptic effect will usually outweigh the risks of hypotension.   Combining propofol with ketamine and possibly a vasopressor may be considered to avoid hypotension.

Although usually Rocuronium is the preferred paralytic, in status epilepticus succinylcholine may be preferable to allow determination of whether the seizure has been controlled.    If Rocuronium is used, there is a risk that the patient may have ongoing seizure activity which is not observable, but which is nonetheless causing brain damage.   Hyperkalemia secondary to rhabdomyolysis takes time to develop, so status epilepticus of short duration (<15-20 min) itself is not a contraindication to succinylcholine.   For a patient who presents to the ED with seizure of unknown duration, Rocuronium is safer.   Alternatively, this may be one situation in which sedative-only intubation may be a reasonable approach, as high-dose propofol will typically provide good intubation conditions provided that it breaks the seizure [more discussion about this below - see Addendum #2].


The duration of time which it is safe to leave a patient in generalized convulsive status epilepticus is unknown.   Until this is clarified, it may be safer to err on the side of caution and avoid periods of status epilepticus longer than 30 minutes.   

Rapid sequence termination is a simple approach to achieve seizure control very rapidly.   This is a true emergency, and waiting for drugs to arrive from the pharmacy or praying that the third-line antiepileptic works may not be the best approach.   Instead, RST allows you to take control of the situation and break the seizure quickly.   This also reduces the time delay to neuroimaging, lumbar puncture, EEG, and other components of patient care which will be ignored during the seizure itself.  

Neurologists and ED/ICU physicians have different perspectives on intubation.   From a neurology standpoint, intubation may be viewed as a failure of antiepileptic treatment to work.   From an ED/ICU standpoint, intubation is more likely to be viewed as an essential step to stabilize the patient and facilitate definitive therapy.

Unfortunately, status epilepticus is very difficult to study due to its emergent and infrequent nature.   There are only a handful of randomized controlled trials available, and many important questions remain unanswered.   Although we believe that RST is a reasonable approach, in the absence of data there are a broad range of approaches which are equally reasonable.   We are not the first clinicians to propose this streamlined approach of intubation and propofol infusion for patients failing to respond to benzodiazepine (Marik 2004).   Ultimately it is up to the bedside clinician to determine which strategy makes the most sense for any given patient given the circumstances and available resources.  

ADDENDUM #1 August 2014: Expert opinion

A while after this post was published, I heard a great talk by renouned neurologist Thomas Bleck on status epilepticus (see: FreeEmergency Medicine Talks).   Dr. Bleck advocates for going directly to intubation for a patient in status epilepticus who fails to respond benzodiazepines.   So for those of you who thought this was insanely agressive, well... maybe it is, but Dr. Bleck agrees with it.

ADDENDUM #2 December 2014: Discussion about how to intubate the seizing airway.

Recently there was an interesting discussion about this on twitter (selected tweets from this below):

Unfortunately I was in the midst of two weeks of 12-hour nightshifts, so I missed the discussion.   Now that I'm somewhat awake, here are some thoughts on this: 

(1) Sedation-only intubation with propofol is not necessarily pure evil

Sedative-only vs. paralysis is an interesting debate (if anyone hasn’t seen the throwdown between Paul Mayo and Scott Weingart, its here).  This is my perspective.  As a fellow I was trained in sedative-only intubation.  Subsequently I’ve transitioned to almost exclusively using paralytics.  Having experienced both approaches, I feel that adding paralytics is a vast improvement with regards to improved first-pass sucess, reduced vomiting, reduced dental trauma, improved bag-mask ventilation, and overall procedural safety.  I’ve posted previously about using rocuronium to rapidly achieve respiratory and hemodyanamic control of the patient.

That said, there may be an occasional role for sedation-only intubation, primarily for status epilepticus.  A bolus of propofol ~1.5 mg/kg will suppress vomiting and cause some muscle relaxation, usually producing good intubating conditions.  There are two major drawbacks to this: hemodynamic deterioration and a limited time window for intubation (if you don’t get the airway on the first or second pass, the propofol wears off).  Propofol can definitely kill patients with minimal cardiac reserve, as explored in this magnificent rant by Cliff Reid on the propofol assassins.

Thus for a patient who is seizing, has contraindications to succinylcholine, has a good blood pressure, and normal-appearing airway anatomy propofol 1.5 mg/kg may be a reasonable approach.  Certainly there has to be a plan-B, such as pushing rocuronium and proceeding down a RSI-pathway.  This set of consequences is rare.  I’ve done it a couple of times, every time the patient has stopped seizing and relaxed nicely allowing successful intubation.  There is some elegance here in controlling the airway and seizure simultaneously.  A propofol infusion should be started immediately to maintain seizure control.  

(2) Dosing rocuronium: How low can you go?

Low-dose rocuronium for intubation raises some concerns.   First, slowly paralyzing the patient may create a period of hypoventilation prior to laryngoscopy, setting the stage for hypercapnia and hypoxemia.   Second, there is some evidence that low-dose rocuronium (e.g. 0.25-0.4 mg/kg) may take longer to paralyze the larynx compared to other skeletal muscles, which theoretically could lead to a situation where the patient is unable to breathe but the trachea cannot be intubated (Wright 1994, El-Orbany 2009).

However, for a patient with status epilepticus, it could make sense to use a lower dose of rocuronium than you would use, for example, on a shocked patient.   For deeply shocked patients who need to be paralyzed rapidly, I subscribe to Scott Weingart’s concept of pushing the dose of rocuronium high (e.g. up to 1.4 mg/kg).   For a seizing patient, it might make sense to use a lower dose of rocuronium (i.e. 0.6 mg/kg), accepting that it will take the patient a bit longer to reach paralysis.

I don't have any experience with low-dose rocuronium, and would be interested to hear from anyone who has used this.

How do you manage the seizing airway?   The comments section below has been opened to allow immediate posting of your thoughts, to facilitate discussion (similar to Google-plus).   


(1) Alldredge 2001 performed a prospective randomized controlled trial wherein pre-hospital patients were treated with placebo vs. lorazepam vs. diazepam.   Mortality rates were 16% (placebo), 8% (lorazepam), and 5% (diazepam).   The study compared the three different mortality rates and found a nearly significant difference (p = 0.08).   However, if you combine the lorazepam and diazepam groups and then perform a comparison of placebo (11/71 = 15% mortality) versus “any benzodiazepine” (8/134 = 6% mortality), with a Fisher exact test this yields a p-value of 0.04.   This is admittedly a retrospective analysis, but I do think there’s a signal here that early benzodiazepines reduce mortality.

(2) Image credits:


  1. Trying to wrap my brain around this... Not sure what the big deal is to paralyze? You have already exhausted first line and second line meds at least. With a paralytic you now have just the brain seizing(still bad), but at least the body is not seizing causing further damage and increased temp, etc. Seems actually more safe to give a full RSI dose of Roc especially in transport, knowing they are still seizing in the brain, so you can safely intubate them under more optimal conditions. It seems better to go with Roc since the bodies hyper-metabolic activity will eat the Succs up in a few minutes possibly causing to redoes the Succs. Full dose Roc seems the lesser of the evils with the evils being: 1. Potential non-optimal intubating conditions using sedative only. 2. Longer paralysis(since redoseing succs could cause Hyper K+). 3. Just brain and not body and brain seizing. Plus if the ER or ICU is a decent one at all they will have the PT on an EEG monitor anyways. Propofol just seems too risky especially if PT loses autonomic brain function in regards to hemodynamics. Also I am a flight paramedic and so I don't carry or continue Propofol for PH or Pre-Tertiary Care Center PTs. I just like to keep things as simple as possible and not have to chase things around in our physiologic stressed filled environment. Please let me know what you think of my thoughts on this. Thanks!

    1. Obtaining continuous EEG monitoring is easier said than done. Logistically this may take several hours (especially on nights/weekends), during which time brain damage could occur.

      I agree that controlling the airway is the most important priority. However, intentionally controlling status epilepticus with a paralytic alone is generally a mistake, as this will leave the patient with ongoing seizure activity which poses a risk of permenant brain damage. Thus, I would disagree that a long-acting paralytic is beneficial, as paralysis may mask seizure activity and thereby provide a false sense of security. Although airway managment is the first priority, all possible efforts should be made to break the seizure.

      Your treatment options may be very limited as a flight paramedic. Thus, these general considerations would have to be adapted to your specific situation and the medications availible to you (for example, benzodiazepines and ketamine could be used to break the seizure).

    2. To clarify that last point: I'm not suggesting a sedative-only intubation using benzodiazepines or ketamine (unlike propofol, these drugs are less likely to produce good intubating conditions). But rather you could intubate with succinylcholine or a lowish dose of rocuronium and, if there was ongoing seizure activity, try to control the seizure with benzodiazepines and ketamine.

  2. Thanks Josh for taking the time to address my thought process. And just so I'm clear I would never treat SE with just NMBAs alone. Sorry I should have clarified. I would obviously be treating with all the agents I had or 3rd line AE drugs from facilities or clinics and maintain hemodynamics. Thanks again... I really dig your work! Keep up the strong effort!

  3. Difficult to be too dogmatic with the level of evidence on the topic.
    I would just like to add that intubation is still a high risk procedure and NMBAs greatly facilitate a safe procedure. In that respect, in the rare case that suxamethonium is contraindicated and you're worried about masking ongoing seizures you could either use proper dose roc and proceed algorithm as if continuing seizures, or, consider using sugammadex to immediately reverse the roc.

    That puts you in a place where a second dose of NMBA becomes more difficult (although there's cis(atracurium) and others) should you lose the tube, but that's probably not your biggest concern in these cases.

  4. Is the listed RST a protocol currently utilized in your ICU? Our department (same hospital as you) recently adopted a different protocol that utilizes Diazepam 5-10mg as the first line agent and does not include Ketamine at all. This seems to defy all the recent literature I have read. What are your thoughts...