Wednesday, February 1, 2012

Pitfalls in stroke

Perhaps my recollections of working in Major Med are different than your experiences. When I was up to my ears in patients, spread a mile wide and a centimeter deep, the last thing I wanted to hear was...
... unless it was immediately followed by "EMS is bringing in a stroke alert!" That meant that a neurology resident would be swooping in, protocols falling in place, and things happening without needing my constant input. 

Perhaps it was wrong of me, karmically improper, to wish that everyone would have their stroke during my shift. Where I work now, the whole stroke team is this guy:

As a result of my new status, I've had to learn a bit more about fibrinolysis of the acute CVA than I did previously. Mostly this has to do with finding exclusions to administering tPA. Clearly, using a check-list is essential here, as well as knowing the policies of your own institution. With that in mind, however, I want to emphasize some pitfalls in the process. For a quick reference to the tPA guidelines, refer to my old post.

Case 1:
EMS calls in with 55 year old female who has had left arm & leg paresis, with a witnessed onset 30 minutes ago. Vital signs and glucose are normal. When you press the paramedics for more information they say that her whole family witnessed the episode. She was seated at the dinner table, but had not yet started eating, when she stiffened up, and then shook for a minute. When EMS arrived she was talking, but she had the new left-sided deficits. 

In the ED, the CT shows no bleed, and so the pharmacy calls down and asks if they should send the tPA with a runner...

Contraindication: Patient had seizure at onset of stroke. 
Or rather, the patient had a seizure at the onset of a "potential stroke-mimic." The concern is that the apparent CVA is actually a Todd's paralysis (pdf), having nothing to do with an embolic or thrombotic etiology. This exclusion has been modified recently, allowing for lysis if imaging can demonstrate an acute vascular occlusion, but this is difficult to achieve in most institutions.

Case 2: 
A 60 year old man is brought in about 2 hours after witnessed onset of right facial droop and aphasia. His vital signs and glucose are normal, his NIHSS  works out to 10, and the CT scan shows no bleed. He take warfarin for paroxysmal AF, but (thankfully?) due to non-adherence, his INR is 1.1. At the 2.75 hours post-onset mark, tPA is started.
Fine - no exclusions.
However, say this same patient was brought in 15 minutes later, and as a result, we are standing over him with the tPA at 3.1 hours post-onset...


Contraindication: Oral anticoagulation therapy, even if INR within normal limits.
The ECASS III trial, which provided the evidence to extend the tPA-window to 4.5 hours after stroke onset, had different exclusion criteria than the NINDS-based 3 hour window. Specifically, if a patient will be getting tPA within the 3-hour window, their INR had to be less than 1.7. In the 3-4.5 hour window, though, any use of an oral anticoagulant is an exclusion. 

Case 3:
A 75 year-old woman is brought in with left arm paresis, as well as left side neglect. She was last seen normal 2 hours ago. She does not take warfarin, Other history include HTN and a prior CVA 5 years ago. The CT is read as negative for acute findings, and the labs return at 3.25 hours after the onset of symptoms. After discussion with the neurologist, you start the tPA at 3.5 hours.

Sounds good.

"I forgot the rest of her med list," says the patient's daughter, right before the nurse starts the tPA. Let's see; metoprolol, amlodipine, metformin, Ambien... Wait, what?

Exclusion: Combination of previous stroke and diabetes mellitus
Yeah, I am trying to emphasize that the 3-hour window is not the same as the 3-4.5 hour window. One of the trickier exclusion criteria is the combination of DM and prior CVA - this is why you have checklists! Now, I don't recall the reasoning behind this particular exclusion, but, in general, ECASS III was trying to enroll a healthier cohort than NINDS, and people with both those comorbidities are not likely very healthy.

Case 4:
A 65 year-old male with acute onset dysarthria and left-sided deficits presents within 3 hours, and has a negative CT. Unfortunately (?) he is excluded for receiving tPA for other reasons. The RN notes that his blood pressure is sticking around 200/100, and asks you what you're going to do about it - "They won't take 'em on the floor with that pressure!"
"Drink me."
You reach for the labetalol and say..

"Hey, did you read that recent study in Stroke? There was a great review of the paper on this cool EM blog... We better not give this labetalol just yet."
Me on my days off. I'm a player.
There are a lot of times when we don't want to treat the blood pressure, and this is the prime example. Per the AHA class 1 recommendation in the stroke guidelines:

It is generally agreed that patients with markedly elevated blood pressure may have their blood pressure lowered. A reasonable goal would be to lower blood pressure by {approx}15% during the first 24 hours after onset of stroke. The level of blood pressure that would mandate such treatment is not known, but consensus exists that medications should be withheld unless the systolic blood pressure is >220mm Hg or the mean blood pressure is >120 mm Hg (Class I,Level of Evidence C).

Okay, wait until the SBP is over 220 - seems crystal clear, right? Well, people have a hard time ignoring those pressures, and end up treating them. As this recent study shows, we treat when he don't have to, and then we over-treat, dropping the pressure too far, risking a watershed infarct

Okay, I think we get the idea! 


  1. right you are brooks walsh, right you are. best case scenario is that the neuro resident is already there when the pt arrives so you just had to look over there shoulder and watch the exam so you know what to right in the chart. and like you, i'm incurring lifetimes of karmic decimation by agreeing with you; sadly however, this is on the bland side of things to which i've become accustomed to as an EM resident. remember when you prayed that the trauma being announced on the trauma pager was 03330? the pedi modified...

    1. The perfect patient, probably, is one that requires no phone calls to be made by yourself. And an intubation.

  2. if by intubated, you're referring to the fact that they can no longer talk at you, then i agree.