Management of Status Epilepticus in ICU

Status epilepticus has been asked about in Question 16 from the second paper of 2014, where it was presented wrapped in a case of subarachnoid haemorrhage. Specifically, the college wanted a demonstration of an escalating "ladder of management", as well as drug doses, strategies for weaning sedation, and some opinion about progosis (which probably has more to do with the subarachnoid haemorrhage than with the seizures).

Other questions on this topic have included Question 22 from the second paper of 2005. Peripherally, revising exam candidates may also find some interest in Question 17 from the first paper of 2022, which asked about the pharmacology of antiepileptics - something that probably belongs in the anticonvulsant section of the First Part exam syllabus.

Oh's Manual explores this issue in Chapter 49 (pp. 549, "Disorders  of  consciousness")  by Balasubramanian  Venkatesh, and Chapter 50 (pp. 560, " Status  epilepticus")  by Helen  I  Opdam.

Good (free) published resources include the following articles:

The time-poor exam candidate may safely limit themselves to this LITFL page on status epilepticus.


Status epilepticus is variably defined as

  • 5 minutes or more of continuous seizure activity, or two seizures with no intervening recovery of consciousness. (Oh's Manual; also 2012 Guidelines)
  • A continuous state of seizures, or multiple seizures, without return to baseline, resulting in observable or even subjectively perceived sensory, motor, and/or cognitive dysfunction for at least 30 minutes (Question 16 from the second paper of 2014)

“Refractory” status epilepticus is defined as any sort of seizure activity which fails to respond to the usual bolus dose of benzodiazepines and first-line antiepileptics.

"Super-refractory" status epilepticus is defined as any seizure activity seen on EEG which continues despite general anaesthesia, i.e. sedation deep enough to permit major surgery.

Status epilepticus is further divided into “convulsive” and “non-convulsive" categories. Predictably, the distinction rests on the motor manifestations. Non-convulsive status epilepticus is a weird animal – these people are typically described as confused; consciousness is not completely lost as in a grand mal tonic-clonic seizure – but rather, it is altered, to the extent that the patient no longer “acts normal” in some (potentially very subtle) way. This bizarre disorder accounts for a quarter of the cases; the rest display bog-standard convulsive features.

Management of status epilepticus in summary

The management strategies can be compiled into a table, which includes the routine and omits the wackiest of therapies.

Management of Status Epilepticus 
From the 2012 "Guidelines for the evaluation and management of status epilepticus"
as well as the review of "Super-refractory status epilepticus"  (2011)

First line therapy

  • Benzodiazepines: boluses every 2-5 minutes (lorazepam apparently superior)
  • Earlier is better (late benzodiazepine therapy is less effective)
  • Phenytoin: 20mg/kg loading dose

Second line therapy

  • Midazolam infusion
  • More phenytoin (O'hs Manual recommends up to 30mg/kg total dose)
  • Sodium valproate, 30mg/kg (Oh's Manual recommends a dose range of 10-40mg/kg)
  • Phenobarbital/thiopentone, and levetiracetam

Refractory status

  • Propofol infusion, or midazolam infusion, or thiopentone infusion.
    • No real way to discriminate between them all in terms of efficacy
  • Continuous EEG monitoring
  • Probably no benefit from adding any more traditional antiepileptic drugs once burst suppression is achieved
  • Once the seizures resolve, it is recommended you wait for 12 hours before weaning the infusion of anaesthetic drugs.

Management of status epilepticus in detail

A recently (2012) published set of guidelines light the way in an otherwise murky fog of evidence-free opinion. As a way of guiding decisionmaking, these people offer a table, which separates the management strategy for status epilepticus into first, second, third and fourth lines of pharmacotherapy.

Other good (free) published resources include the following articles:

Overall, the objective is to get the seizures to stop - preferably within the first hour of presentation. There does not seem to be any benefit in distingusihing between convulsive and non-convulsive status epilepticus for the purposes of pharmacotherapy, as the management is essentially the same.

First line agents

  • Benzodiazepines: boluses every 2-5 minutes
  • Phenytoin: 20mg/kg loading dose
    • Phenytoin on its own is useless. Or rather, it is inferior to benzodiazepines as a solitary agent. Always, both must be used simultaneously.

Second line agents

  • Midazolam infusion
  • Phenytoin (well, rather, the American study recommends fosphenytoin)
  • Phenobarbital and levetiracetam are also in this second line of attack

Third line agents: for refractory status epilepticus

  • Propofol infusion, or midazolam infusion, or thiopentone infusion. 
    Which is best? Prabhakar et al (2017) made a valiant attempt to Cochrane the evidence, and performed a heroic meta-analysis of one fucking study which included only 24 patients. Surprisingly, it was inconclusive. All that can be said of the aforementioned agents is that propofol is likely to wear off faster and offer a better chance of early extubation.
  • At this stage, continuous EEG monitoring becomes mandatory
    • It does not matter which "general anaesthetic agent" you start an infusion of in refractory status epilepticus. However, one would be well advised to mention in their CICM fellowship answer that the patient has been intubated a long time ago, and that continuous EEG monitoring is now in progress.
  • The role of traditional antiepileptic drugs is also exhausted at this stage, as there will probably be no benefit from adding them into a situation where a constantly observed burst suppression is already achieved by high dose anaesthetic infusion.

Fourth line agents: for these, there is little evidence.

Fifth line therapies:

Hypothermia was recently explored in the HYBERNATUS trial (Legriel et al, 2016) mentioned in the college answer to Question 9 from the first paper of 2015.  The authors randomised 270 patients to either have conventional therapy alone, or to also undergo hypothermia down to 32°C. Strangely, in this study the intervention was not a "fifth-line" therapy - to qualify for cooling, all you had to do was rock up to the hospital and have a fit for longer than five minutes (i.e. satisfy the major criteria for SE which are mentioned in the grey box above). Then, half of you got some variety of anaesthesia, and the other half got their anaesthesia and cooling. 

The primary outcome measures reflect the opinion of the authors that hypothermia should have some sort of a direct neuroprotective effect, something which the other therapies cannot boast. They were looking mainly at functional outcome at day 90. Unsurprisingly, no significant improvement was seen in any of the primary or secondary outcomes.  but the authors did find that of the cooled patients fewer ended up still in status on the second day of the study, i.e. after 24 hours of cooling the patients were half as likely to still have ongoing epileptiform activity on EEG (which describes "refractory" or "super-refractory" status epilepticus). Another unpleasant but expected revelation was the finding that adverse events were more frequent among the cooled patients.

Another criticism of this study is a non-standard use of hypothermia. The authors only decided to use a 24-hour cooling session, which is weird. What if the patient remains in status after that? Will you de-escalate hypothermia in the face of ongoing EEG evidence of status?  Odd. Moreover, the target temperature took too long to achieve.

The study was interesting, but is unlikely to change anybody's mind about hypothermia in this context. Rather, it confirms what most people already knew - that hypothermia will have no benefit unless it is used properly, and in carefully selected patients. Most of the community will continue using hypothermia for refractory and super-refractory status epilepticus, i.e. SE which has not responded to conventional general anaesthesia. In such cases hypothermia is sustained while EEG features of SE persist, occasionally test-rewarming under continuous  EEG monitoring to see whether it is safe for the patient to return to a more normal temperature.

Management of non-convulsive status epilepticus

The pharmacotherapy fo non-convulsive status epilepticus is not as well researched. Again, this is a topic which is explored in a separate summary article. In general, the treatment should be directed at the cause, i.e. at the management of the septic episode for example.

Because the level of awareness is distorted and its recovery marks a response to treatment, non-sedating anticonvulsant agents are preferred. However, benzodiazepines are still the first line, and Oh's Manual mentions sodium valproate and levitiracetam as options.

Pharmacological caveats in the management of status epilepticus

Certain features of status epilepticus influence the pharmacological management; for instance:

  • The GABA receptors diminish in number during the course of an episode, and thus it is vitally important to give the benzodiazepines early, while they still have some effect.
  • There is probably no difference between various benzodiazepine agents, but if you had to pick a single agent, you would pick lorazepam because it is longer acting and therefore associated with a decreased risk of seizure recurrence.
  • On one hand, one does not want to suffer the terrible metabolic consequences of sustained whole-body muscle spasm. On the other hand, paralysis masks seizures. On the balance of things, it is important to know whether the patient is still fitting. Try to avoid paralysis. If paralysisis in use, you should also be using continuous EEG monitoring.
  • Once the seizures resolve, it is recommended you wait for 12 hours before weaning the infusion of heavy anaesthetic drugs.

Which antiepileptic drug, and why does it matter?

Apart from the benzodiazepines, the pharmacological arsenal contains several munitions with the properties of which the savvy exam candiate should have at least a passing familiarity.

Rather than go into extensive detail, I will attempt to mention a few significant pharmacological features, and point out advantages and disadvantages.

Phenytoin: 20mg/kg. Depends on hepatic metabolism.

Sodium Valproate: 10mg/kg usually; but UpToDate recommends a loading dose of up to 30mg/kg.  Depends on hepatic metabolism.

Levitiracetam: 1000 to 3000mg; does not rely on any specific organ for clearance, and unlike the other seconds line agents this one does not compromise one's airway with sedation. In fact UpToDate recommend 60 mg/kg up to a maximum of 4500 mg.

Barbiturates (specifically, thiopentone) -After intubationg the patient with 3-5mg/kg, one would continue with 1mg/kg boluses until the seizures are controlled, and then start an infusion to defeat the constant redistribution of this drug into the fat compartment. Once that compartment is saturated, the rate of infusion can slow down. Then, its just a matter of waiting for several weeks for the patient to wake up.

Propofol has the advantage of a very brief duration of activity. In order to accomplish EEG suppression, one may need to use it in quantities well in excess of what we are normally accustomed to using in the ICU. In general it is thought that for most people a dose rate of around 4mg/kg for longer than 48 hours is enough to precipitate the catastrophic propofol infusion syndrome; however this is extrapolated from paediatric data. Oh's Manual recommends 5mg/kg.

Outcomes for status epilepticus

The following table is an abridgement of the far better table available in the 2012 Guidelines for Management of Status Epilepticus.

Status Epilepticus


  • At hospital discharge: 9–21 %
  • At 30 days: 19–27 % 
  • At 90 days: 19 %
  • Older age
  • Unconsciousness
  • Duration of seizures,
  • Focal neurological signs
  • Medical complications

Non-convulsive status


  • At hospital discharge: 18-52%
  • At 30 days: 65%

Time-critical diagnosis:

  • Diagnosis within 30 min of seizure onset:
    mortality 36 %
  • Diagnosed 24 hr after seizure onset:
    mortality 75 %

Features associated with poor outcome:

  • Severe mental status impairment
  • Unknown cause
  • Longer seizure duration
    • Less than 10hrs:
      10% mortality
    • More than 20hrs:
      85% mortality

Refractory status


  • At hospital discharge: 23–61 % 
  • At 3 months: 39 %

Features associated with poor outcome:

  • Older age (e.g., >50 years)
  • long seizure duration
  • high APACHE-2 scale scores


Oh's Intensive Care manual:

Chapter 49   (pp. 549) Disorders  of  consciousness  by Balasubramanian  Venkatesh

Chapter   50   (pp. 560) Status  epilepticus  by Helen  I  Opdam

Meldrum, Brian S., Roger A. Vigouroux, and James B. Brierley. "Systemic factors and epileptic brain damage: prolonged seizures in paralyzed, artificially ventilated baboons." Archives of Neurology 29.2 (1973): 82-87.

Olney, J. W., R. C. Collins, and R. S. Sloviter. "Excitotoxic mechanisms of epileptic brain damage." Advances in neurology 44 (1985): 857-877.

Beal, M. Flint. "Mechanisms of excitotoxicity in neurologic diseases." The FASEB journal 6.15 (1992): 3338-3344.

Yeh, Huei-Ming, et al. "Convulsions and refractory ventricular fibrillation after intrathecal injection of a massive dose of tranexamic acid." Anesthesiology 98.1 (2003): 270-272.

Murkin, John M., et al. "High-dose tranexamic acid is associated with nonischemic clinical seizures in cardiac surgical patients." Anesthesia & Analgesia 110.2 (2010): 350-353.

Tan, R. Y. L., A. Neligan, and S. D. Shorvon. "The uncommon causes of status epilepticus: a systematic review." Epilepsy research 91.2 (2010): 111-122.

Tibussek, Daniel, et al. "Status epilepticus due to attempted suicide with isoniazid." European journal of pediatrics 165.2 (2006): 136-137.

Wyderski, Richard J., W. Grant Starrett, and Alaa Abou-Saif. "Fatal status epilepticus associated with olanzapine therapy." Annals of Pharmacotherapy33.7-8 (1999): 787-789.

Martínez-Rodríguez, Jose E., et al. "Nonconvulsive status epilepticus associated with cephalosporins in patients with renal failure." The American journal of medicine 111.2 (2001): 115-119.

De Sarro, A., et al. "Relationship between structure and convulsant properties of some beta-lactam antibiotics following intracerebroventricular microinjection in rats." Antimicrobial agents and chemotherapy 39.1 (1995): 232-237.

Johnson, Herbert C., and A. Earl Walker. "Intraventricular penicillin: a note of warning." Journal of the American Medical Association 127.4 (1945): 217-219.

Johnson, Nicholas, et al. "Anti-NMDA receptor encephalitis causing prolonged nonconvulsive status epilepticus." Neurology 75.16 (2010): 1480-1482.

Barry, Elizabeth, and W. Allen Hauser. "Pleocytosis after status epilepticus."Archives of neurology 51.2 (1994): 190.

Singhal, P. C., K. S. Chugh, and D. R. Gulati. "Myoglobinuria and renal failure after status epilepticus." Neurology 28.2 (1978): 200-200.

Collins, W. C., O. Lanigan, and N. Callaghan. "Plasma prolactin concentrations following epileptic and pseudoseizures." Journal of Neurology, Neurosurgery & Psychiatry 46.6 (1983): 505-508.

Towne, A. R., et al. "Prevalence of nonconvulsive status epilepticus in comatose patients." Neurology 54.2 (2000): 340-340.

Brophy, Gretchen M., et al. "Guidelines for the evaluation and management of status epilepticus." Neurocritical care 17.1 (2012): 3-23.

Ma, Xiaoping, et al. "Neurosurgical treatment of medically intractable status epilepticus."  Epilepsy research 46.1 (2001): 33-38.

Sass, K. J., et al. "Corpus callosotomy for epilepsy. II. Neurologic and neuropsychological outcome." Neurology 38.1 (1988): 24-24.

Chen, James WY, and Claude G. Wasterlain. "Status epilepticus: pathophysiology and management in adults." The Lancet Neurology 5.3 (2006): 246-256.

Treiman, David M., et al. "A comparison of four treatments for generalized convulsive status epilepticus." New England Journal of Medicine 339.12 (1998): 792-798.

Meierkord, Hartmut, and Martin Holtkamp. "Non-convulsive status epilepticus in adults: clinical forms and treatment." The Lancet Neurology 6.4 (2007): 329-339.

Borris, Douglas J., Edward H. Bertram, and Jaideep Kapur. "Ketamine controls prolonged status epilepticus." Epilepsy research 42.2 (2000): 117-122.

Slovis, Corey M. "Lidocaine in the treatment of status epilepticus." Academic emergency medicine 4.9 (1997): 918-925.

Storchheim, Frederic. "Status epilepticus treated by magnesium sulphate, injected intravenously." Journal of the American Medical Association 101.17 (1933): 1313-1314.

Kluger, G., et al. "Pyridoxine-dependent epilepsy: normal outcome in a patient with late diagnosis after prolonged status epilepticus causing cortical blindness." Neuropediatrics 39.5 (2008): 276-279.

Mirsattari, Seyed M., Michael D. Sharpe, and G. Bryan Young. "Treatment of refractory status epilepticus with inhalational anesthetic agents isoflurane and desflurane." Archives of neurology 61.8 (2004): 1254-1259.

Hamani, Clement, et al. "Deep brain stimulation of the anterior nucleus of the thalamus: effects of electrical stimulation on pilocarpine-induced seizures and status epilepticus." Epilepsy research 78.2 (2008): 117-123.

Cervenka, Mackenzie C., et al. "The ketogenic diet for medically and surgically refractory status epilepticus in the neurocritical care unit." Neurocritical care15.3 (2011): 519-524.

Corry, Jesse J., et al. "Hypothermia for refractory status epilepticus."Neurocritical care 9.2 (2008): 189-197.