I had the amazing pleasure of taking part in the online version of North York General’s Emergency Medicine Update conference this year. (There are more free webinars scheduled. Sign up here.) I gave a webinar entitled “Status Epilepticus: A less shaky approach”. The talk was recorded, and is now available online here:
The premise of this talk is that we need to be much more aggressive when managing status epilepticus. There are multiple supporting blog posts with more information and references:
- My general approach to status epilepticus
- Aggressive seizure algorithm: why I use propofol early
- A post about the Burman study.
- Our second line agents for status epilepticus all suck (The ESETT trial)
- Levetiracetam versus Phenytoin in Status Epilepticus (ConSEPT and EcLiPSE)
- Is there any reason to use intranasal midazolam for seizures?
There are some very brief answers to the questions that accompanied this talk:
What is your choice of traditional second line agent?
- It probably doesn’t matter. In every trial, they are all the same.
- Personally, I think phenytoin is still a great drug – and is my go to with a few caveats:
- If there is a chance that this is an overdose – a toxicologic seizure – I don’t want the sodium channel blocking properties on board.
- If the patient has significant cardiac disease, I shy away from phenytoin.
- And if the patient is really heavy – say over 100 kg – so the total dose of phenytoin is getting high. I think those patients are more likely to have side effects, so I tend to go with a second option.
- My second option these days is levetiracetam, not because it has great evidence, but because that is what the neurologist is going to prescribe anyway, and it does have fewer contraindications than everything else.
- When I use levetiracetam, I use the ESETT dose – which is 60 mg/kg up to 4.5 grams. So a lot bigger dose than you might be used to.
- The lorazepam dose in the VA cooperative study was 0.1 mg/kg. (They did give the lorazepam slowly though, at 2 mg/minute. I am not sure this is needed, but if you wanted to give your 8 mg loraz dose as 2 4 mg doses separated by 2 minutes, I think that would be OK).
- 8 vs 16 mg of midazolam – no difference in seizure cessation, but also no difference in respiratory depression. Mentioned in: https://emcrit.org/emcrit/status-epilepticus/
Role of ketamine?
- Lots of very smart people suggest it. Josh Farkas uses the same algorithm I do, except he gives ketamine and propofol together. That is reasonable. (Josh Farkas post).
- Theoretically, because it acts on the NMDA receptors, which aren’t touched by either benzos or propofol, it should be a really nice combination.
- There just isn’t great evidence – and I worry that it complicates things a bit – and I like my resuscitations to be as simple as possible.
- So right now I keep ketamine for my third line agent if propofol doesn’t work – which hasn’t happened yet.
- This is a very difficult area, with lots of strong opinions, but essentially no data.
- When to intubate:
- The Burman study seems to indicate that as long as we are more aggressive in breaking seizures, we might get away with intubating fewer patients.
- I currently intubate with RSI when pushing the propofol. One major adjustment to the RSI is I dont think these patients will be adequately pre-oxygenated, so I use BVM ventilations during the apneic period.
- RSI vs Sedative only
- Lots of people will argue for avoiding paralytics. The paralytic makes everything more complicated because we can’t tell when these patients are seizing.
- However, these are high risk intubations. There is no way to adequately pre-oxygenate, the patient will probably have significant metabolic acidosis and big loads of drugs on board – so the chance of a peri-intubation crash is really high. Therefore, I really want my first pass success to be very high, so I stick to a standard RSI (except I bag during the apneic period).
- But some very smart people will disagree with me. Scott Weingartt suggests a sedative only intubation in his podcast: https://emcrit.org/emcrit/status-epilepticus/ On the other hand, George Kovacs suggests the opposite: https://emergencymedicinecases.com/status-epilepticus/
- Which paralytic?
- There is no good answer.
- One you paralyze someone, the chance of non-convulsive status is very high, and it has really bad outcomes.
- Some say succinylcholine because it wears off faster (but once paralyzed, my guess is that non-convulsive status is still possible even as the paralytic wears off). The big problem with sux is that these patients, if they have been seizing for a while, often have contraindications – rhabdomyolysis, hyperkalemia. They are probably also more likely than our average patient to have an underlying neurologic disorder.
- The benefit of rocuronium is none of the contraindications apply. It will last a long time, but you could reverse it with sugammadex.
- Personally, I use sux for seizures less than 15 minutes, but if I am more worried about contra-indications, I use roc.
- I assume every paralyzed patient could be in non-convulsive status – so all these patients need to get somewhere with EEG capabilities fast.
Propofol infusion syndrome
- Propofol infusion syndrome is an important consideration. Our ICU / PICU colleagues will want to stop propofol infusions for many of these patients, and transition them to midazolam or phenobarbital, if ongoing treatment is necessary. However, a few hours of propofol in the emergency department is safe, and I think it is our best option.
- I think the benzo benzo propofol approach is far more successful than our old approach, so we might actually see nonconvulsive status less often, because we are breaking the seizures early enough.
- But you have to be very careful when you think you are successful – outcomes from subtle or nonconvulsive status are really bad.
- Look carefully at the eyes, hands and feet for any subtle movement.
- Check for tone – these patients should be really loose post benzos. Increased tone is an indications of possible non-convulsive status.
- Come back and do frequent repeat exams – if the patient is not waking up, consider status.
- If there is any doubt – get the patient an EEG. Realistically, that doesn’t alway happen. If you can’t get an EEG and there is doubt, treat empirically (intubate / propofol / midazolam infusion).
- Although EEG access is difficult, you wouldn’t look after arrhythmias in a setting without access to ECG. You need the right equipment to manage the patient.
- What do we do if the algorithm fails? Luckily, it fails far less frequently than our normal algorithms. Burman had a 99% success rate! However, we still need to know options 3 and 4.
- Get specialist help – ICU and neurology.
- Add ketamine.
- Re-bolus the propofol and set the drip to max.
- Add a midazolam infusion. If midaz infusion – high dose better than low dose (0.4mg/kg/hr had lower mortality and lower seizures than 0.2 mg/kg/hr). (Trinka 2015)
- If your anti-epileptic doses are limited by hypotension, use a vasopressor
- Add a second traditional anti-epileptic agent. So if you started with phenytoin, add levetiracetam, or vice versa. Just keep in mind you can’t give both phenytoin and valproate – they interact.
- Medical coma with phenobarb.
Comment on the evidence
- I think there is very strong evidence that the general anesthesia approach will stop the seizure and do so very quickly. The Burman trial is the best RCT evidence, but there is also observational evidence, which I cover in this post.
- However – we need to be cognizant that this is a surrogate outcome. What we really want to see is better long term outcomes (survival and neurologic function) and there is just no science looking at that. Almost all of our trials look at time to seizure cessation as a surrogate.
If both propofol and midazolam act on the GABA receptors, what is the advantage of moving to propofol?
- OK, I am going to need a neurophysiologist on this, but my understanding is that there are actually multiple subtypes of the GABA receptors. The GABAa receptors are sensitive to benzos, but they are swapped out of the membrane as the seizure goes on, and replaced by benzodiazepine insensitive GABA receptors – but these receptors are still responsive to propofol and barbiturates.