New to the NeuroICU? Rotating on a Stroke Consult Service? Caring for the undifferentiated neurocritically ill patients as they roll in from the ambulance bay?
31 pearls that will have you looking like you've been doing this for years.
If you prefer a hand-written version, these were initially posted on Twitter over the month of July for all new learners and can be found by searching #NeuroPostItPearls there.
#1 Not every “spell” or shake is a seizure. Avoid premature closure for a transient event.
✨Convulsive Syncope (PE? Arrythmia? Critical valvular pathology?)
✨Psychogenic Non-Electrographic Seizures (PNES)
✨Movement disorders: Tremors, rigors, dystonias, dyskinesias
✨Intoxication / withdrawal
#4: Practice deliberate language for goals of care meetings. We may liberate from machines, but we NEVER “withdraw care.”
I have learned so much about deliberate language for goals of care conversations from the tweets of Shunichi Nakagawa, MD.
# 7: 10 of 10 healthcare workers agree that it's really noxious to get the COVID nasal swab. Similarly, a Q-tip nasal tickle is a STRONG (but humane appearing) central noxious stimuli. Wake up, obtunded patient!
Much nicer than pinching, nipple twisting (god, please no!), trap squeezing, etc.
# 8: Deliberately practice thinking like a neurologist. Which is basically localization and tempo. This #PostItNeuroPearl from Dr. Lyell Jones (@LyellJ) a phenomenal neurologist and educator:
🦪 This pearl was fully explored in a case from the book “How to Think Like A Neurologist.”
✨You can check out the #tweetorial here 👉 (The Man Who Started to Drool).
# 10: IMPORTANT: when measuring opening pressure on an LP manometer you are measuring in cmH2O. When guidelines talk about thresholds for intracranial hypertension, they mmHg. THESE ARE NOT THE SAME.
Read the Tweetorial Here!
#15: The pupillary muscles are smooth muscles and thus not impacted by neuromuscular blockade. For everything else you'll need to wait ~10-15 mins for succinylcholine, 1 hour+ for rocuronium.
- The effect of neuromuscular blocking agents can be tracked with the train of fours (TOF). <75% of receptors are blocked when 4/4 twitches return. If the exam is needed more urgently and roc was used, consider reversal with sugammadex.
- Trial Data looking at pupils in paralyzed patients: here and here
#16: De”Cor”ticate posturing: Arms in towards the “Core”
- Alternatively, some noted that decerebrate has more Es for Extension
- Images from The Acute Neurology Survival Guide
#26 Eye movements really confuse me. Maybe you're an eye-movemet savant, but if not, here's a tip: take the time to video a patient's eye movements.
You can even do it in “slow-mo.” Then, you can watch it over and over and compare it to diagrams and ask friend. Eye movements are so localizing, it's worth taking the time to get it right.
Diagrams from The Acute Neurology Survival Guide
# 28 Use the right dizzy exam. HINTS is for a patient with persistent vertigo and observable nystagmus. Dix-Hallpike is for the patient with episodic dizziness.
Awesome videos and teaching about HINTS and posterior circulation strokes here
Stroke (ischemic / hemorrhagic) Pearls
#2 Keep a folder in your phone's “notes” of these 4 things:
- NIHSS score card / print out of the pictures
- How to calculate the modified Rankin Score
- ASPECT scoring guide
- tPA / TNK exclusion criteria and Consent Wording
#9 About to order dual anti-platelet therapy? Ask yourself … is this person going to need a trach and/or PEG?
- Some surgeons are comfortable with doing the procedures regardless… but not all, and then it becomes a real hassle. So if the DAPT is for a non-urgent indication, get the procedures first. If the DAPT is crucial (such as for fresh stent), cangrelor is a good option, if on the formulary. (v pricey, but the re-occlusion of a stent is obviously worse, published experience in the cardiac fresh-stent population here.)
#12: Steps for a tPA/TNK related bleed:
- If tPA is still infusing, stop it!!
- Send a stat fibrinogen.
- DO NOT WAIT for the level to come back. Give 10U of cyroprecipitate and 1g TXA/5g Amicar (guidelines are +/- on TXA/5g Amicar), but mechanistically it makes sense to counter-act what thrombolytics are doing).
- If the fibrinogen is <150, give an additional 10U and recheck. Keep repleting until the patient is out of iatrogenic DIC.
Review more at the 2108 AHA/ASA guidelines for AIS.
# 13: When establishing “Last seen Well” ask SPECIFICs – “when did you last see the person doing [insert deficit] normally?”
- For example: “They got up to use the bathroom @ 2 AM” doesn't establish that the patients whose only deficit is aphasia was talking at that time.
# 14: Stroke Syndrome + <4.5 hours LSW = BP <185/110 NOW.
A hypertensive patient arrives within the tPA/TNK window (up to 4.5 hours) and has a clear stroke syndrome, it's either a thrombolytic-eligible ischemic stroke or an ICH … either way, the goal BP is going to be <185/110… Just go ahead and start an antihypertensive (ideally, something titratable)
#20 K-Centra contains Heparin. It should not be given to a patient with a history of heparin induced thrombocytopenia.
From the package insert, more information about K centra here
#24 Warfarin associated ICH? Give 1500 IU 4F-PCC as soon as it's a confirmed bleed.
The @StrokeAHA_ASA guidelines for ICH recommend INR-based dosing for reversal of vitamin K. This requires waiting to know the INR though. Unless you really doubt the patient was taking the med, draw the INR, and give a standardized dose (1500 IU) up front.
#Timeisbrain and reversing quickly is so important. Even if the INR comes back as 1.3, the 1500 IU load is within (the class 2b) supported 10-20 IU/kg range for most normal weight patients.
If the INR comes back as 3, then your total dose may end up being more like ~3000 IU and they’d need a second 1500 IU, but great that they’ve already gotten some of the reversal! The only draw back is that you must remember to follow up the INR and finish the treatment.
There is some evidence that fixed dose strategies are faster. And consistent quick action is crucial. But, giving a follow up dose adheres to the weight-based dosed per the guideline, best of both strategies
#25: During an Acute Stroke Code, keep the family on the phone!
Nothing is worse than wanting to get necessary information and not being able to reach the family. Once you've got them, keep them on the phone. Information you may need includes:
- Last seen well time
- Medication list, particularly anticoagulants
- Baseline modified Rankin Score
- tPA/TNK consent and exclusion questions
- Mechanical thrombectomy consent
# 27 Think VIVID for each acute stroke consult.
V – Venous
I – Infection
V – Vasculitis
I – Inflammatory
D – Dissection
These are some atypical etiologies of stroke. If you don't remember to check for them or consider them, you will miss them! Thank you @drokane for this!
# 29 Like Stroke, groin complications after mechanical thrombectomy may either be hemorrhagic or ischemic!
All types of closures can result in pseudoaneurysm, RP hematoma, and limb ischemia — but some have a higher risk than others
- Perclose is a suture based closure, the thing to watch for are pseudoaneurysms. This is generally considered the safest closure by vascular surgery
- Angioseal is a plug-based closure. If the plug is displaced or deployed erroneously, it can cause critical limb ischemia.
- Neha and I plan to do a whole post dedicated to the complications after mechanical thrombectomy and other angio-based procedures. Stay tuned.
# 30 The ICH score should not be used for acute neurologic prognostication. Resuscitate first (unless clearly outside of GOCs) then prognosticate.
Whole post dedicated to this! Read it here. This is emphasized in a recently published analysis of patients enrolled in CLEAR-III. More than 40% of patients with a poor outcome at 30 days recovered to an ok outcome (mRS<=3) by 1 year. We cannot lose faith on these patients!
Read the article here.
#3 When reviewing an MRI, before calling anything “enhancing” make sure it is not intrinsically T1 Bright
👈It would be tempting to say this is a “rim-enhancing lesion” on the T1 Post-gadolinium sequence.
Do not be fooled though! What is bright on the post-gad sequence is just intrinsically T1 bright (here on the T1 no gad). 👇
This finding was caused by subacute blood in the cavity (this was a resection-confirmed cav-mal). All about this case here.
#11 Windowing the non-con Head CT can make all the difference in detecting acute stroke!
This is particularly important when calculating the ASPECT score. The literature describes windowing to Level 40 HU x Width 40 HU or Level 32 x width 8 HU (PMID 10540655). However, I personally think the blurring of the gray-white junction is easiest to see at a ~30×30 window.
The left is the pre-set brain window (W 100, L 30). Something kinda looks off in the R frontal lobe… but its hard to see.
Change the windowing (30×30 on the right) and 💥 bam! 💥 that acute stroke really pops out. This is the windowing you should grade ASPECT scores in. You can see what the W/L are set as by looking at the bottom righthand of the screen.
#17 Not all microbleeds mean cerebral amyloid angiopathy.
Boston Criteria 2.0 was released and emphasizes that only lobar micro bleeds in symptomatic patients can be used as supporting evidence for CAA.
# 31 Beware of the Ghost Core!
CT perfusion may overestimate the “core tissue” (what is usually purple in the rapid maps, rCBF<30%). Especially in the early (<6 hours since LSW) window.
The limitations of perfusion imaging are really well described in episode #160 of the BrainWaves Podcast (by @JimSiegler).
CTP perfusion scans define the “core” (purple colored) by the area in the relative cerebral blood flow is <30%. That' is not enough blood flow and thus, these is the area that is “destined to infarct.” As time goes on, more of what is “destined to infarct” has in fact infarcted. (There is an example of perfusion imaging in AIS part 2)
However, in the early time window, these hypoperfused areas may not have undergone permanent ischemic damage. Thus the “core” is a “ghost” (ie not real).
In these early window cases, it can be easier to just use the traditional ASPECT score to determine candidacy for EVT. More about the ASPECT score in AIS part 1.
#18 About to give a Carbapenem? Better make sure the patient is not on Valproic Acid!
- Carbapenems increase VPA glucuronidation
- This causes a significant (60-90%) drop of VPA within 24-72 hours of a carpenem administration. And it's gonna take 1-4 weeks before the levels may recover…. oops
- this is great for VPA overdose emcrit.org/ibcc/vpa/
- Not so good in this #tweetorial that highlights how this can rapidly cause issues:
#21 Remember status dosing by the pheny-twenty. Fosphenytoin is a load of 20 PE/kg (up to 1500mg), the other doses go up by 20. 40mg/kg valproic acid (up to 3000mg) and 60mg/kg levetiracetam (up to 4500mg).
- Underdosing Keppra seems a popular thing to do. Don't do this! If the patient is in status or even if you think the patient might be in status, give the supersized dose.
#22 Tube Feeding affects the bioavailability of oral phenytoin. Hold tube feeds 2 hours before and after the dose.
- This is mostly meant to emphasize that it's a terrible idea to discharge a patient to a nursing home (or even home!) on tube feeds and oral phenytoin. It's setting people up for failure.
#5 For the #ncc fellows– you are a “sometimes intubator” which means it is crucial that you have a really good plan. Have a checklist and use it every single time. Mine is SOAP TIME.
The whole EMCrit site is filled with tips on intubation [obvi]!
#6 Use a Checklists for Neurologic Emergencies!
These are what we give fellows to use for triage, it's basically the same for doing a consult or admission or ED management.
#19 For most EVDs you must clamp the EVD to measure ICP.
Unless the patient is on a clamp trial (in which case they are clamped as long as tolerated), most protocols will have the RN clamp the EVD and chart the ICP once per hour. Which means the number displayed on the monitor is inaccurate about 55 mins of every hour. More about EVDs in our first #TeamNeuroEMCrit post: here!
# 25 “Overbreathing” the vent does not always disqualify brain death testing.
Sometimes “breaths” are auto-triggered and it's critical to tease this out: here's how.
# extra (I counted 25 twice): Guillain Barre Patients – waiting for hypoxia or hypercarbia is waiting TOO long. The 20/30/40 can be helpful.
When patients can make a good seal for PFTs the following are signs of impending respiratory failure:
- Vital capacity <20 mL/kg
- Max inspiratory pressure < -30 cm H2o
- Max expiratory pressure < 40 cmH2O
Do not leave these people on the floor; they need an ICU.
A rough surrogate for VC of 20 ml/kg is having the patient count to 20 in one breath. As Josh pointed out in this recent post on neuromuscular emergencies this is not all that accurate or reliable, but I think a lot of it depends on the patient's baseline. If they could do the counting test on admission and now they can't?? That's bad news.
Every now and then I'm on a triage call where a patient is per-herniation and the hypertonic saline or mannitol is going to “come from pharmacy” … these are not readily available and thus are no help. You know what is readily available and in code carts everywhere? Hypertonic bicarb. You know what's really salty? Hypertonic bicarb. In fact, two amps of this will get you about the same amount of salt as the standard 250cc of 3%. Check out Neha's awesome post about everything you need to know about hyperosmolar therapy!