Schrödinger’s Fence

…or, where we currently sit on the matter of thrombolysis in Acute Stroke.

An Opinion Piece 

Firstly, most sincere and heartfelt apologies to you all, good readers, regarding the use of the Schrödinger’s Cat analogy here. As you are all well aware, Erwin Schrödinger (Austrian Physicist, in 1935) proposed this rather absurd construct to explain the contradictions of quantum physics, whereby the cat, in the box, was either alive or dead, because of the random nature of the decay of a radioactive atom, allowing the cat, until the box is opened, to be both alive AND dead.

How has this any relevance to the current state of play of thrombolysis for acute strokes in the ED?? Possibly in no way whatsoever, but I have been struggling to come to terms with understanding our current position, and more importantly, the implication it has for my practice, my specialty and the way I teach my junior staff.

This juncture seems to be one of ‘fence sitting’ and by my reckoning this position is either one of great sagacity/wisdom, or great cowardice/avoidance.  Could it be both simultaneously?

The Evidence

As proponents and partakers of the FOAMed paradigm, I have no doubt that you have all had the opportunity to digest the opinions of the Titans regarding the journey that the use of thrombolytics has taken in the therapy of acute CVAs. If not, drop everything, don’t even breathe, until you’ve listened to the David Newman and Ashley Shreves podcast on SMART EM, and have read Andy Neill’s multi-part summary 

The other must reads are:

  • Daniel M Fatovich, Stephen P MacDonald, Simon G Brown: Thrombolysis in acute ischaemic stroke The Lancet, Volume 380, Issue 9847, Page 1053, 22 September 2012 [The Lancet]
  • Daniel M Fatovich: Believing is seeing: Stroke thrombolysis remains unproven after the third international stroke trial (IST-3) Emergency Medicine Australasia (2012) 24, 477–479 [EMA Full Text]
  • Ryan Radecki: The Third International Stroke Trial: IST-3 [Emergency Medicine Literature of Note] (plus the fabulous comment from Greg Press)
  • Domnhall Brannigan: Stroke thrombolysis and IST-3 – is it another false dawn? [The Underneaths of EM]
  • Jerome R Hoffman, Richelle J Cooper: How is more negative evidence being used to support claims of benefit: The curious case of the third international stroke trial (IST-3) Emergency Medicine Australasia (2012) 24, 473–476 [EMA Free Full Text]

The conclusions are reasonably uniform throughout all of these pieces.  Almost all of them carefully dissect the data and conclusions drawn from the totality of trials investigating the utility of thrombolysis in acute stroke, and feel that the summaries and recommendations by the authors do not stand up to the highest level of scientific scrutiny (particularly in regard to the most recent, and largest trial, IST-3)

For reference, the IST-3 paper is included here, the comment published in the same Lancet journal edition and the current Cochrane Review on the subject.

  • The benefits and harms of intravenous thrombolysis with recombinant tissue plasminogen activator within 6 h of acute ischaemic stroke (the third international stroke trial [IST-3]): a randomised controlled trial. The Lancet, Volume 379, Issue 9834, Pages 2352 – 2363, 23 June 2012 [Lancet Reference]
  • Didier Leys, Charlotte Cordonnier: rt-PA for ischaemic stroke: what will the next question be? The Lancet, Volume 379, Issue 9834, Pages 2320 – 2321, 23 June 2012 [Lancet Comment]
  • Wardlaw JM, Murray V, Berge E, del Zoppo GJ Thrombolysis for acute ischaemic stroke [Cochrane review]

Jerry Hoffman talks about IST-3 in the August 2012 edition of Emergency Medicine Abstracts (start at 39 min 20 seconds):

To hear more of this fascinating discussion, we greatly look forward to the presentations at SMACC 2013 by Domnhall Brannigan, and by David Newman at the ACEM winter symposium in Broome, in June 2013.

BUT where are we left currently??

On the damned fence!

Now, I have no role to play in trying to explain the nuances of the literature; the pros and the cons.  I am possibly statistically disabled.  I have a major issue with trying to understand the numbers wielded with such finesse by others, (and yes, let’s admit it, with direction sense, and parking.  I may have had a stroke in the mathematical analysis part of my brain – where was that tPA when I needed it, huh!?) So I am intensely grateful to those that can, and choose to share it. With these at my side, I have read the papers, and have come to a conclusion, which I know is shared by many others in the critical care world.

Currently there is no credible evidence that thrombolysis is clearly beneficial in acute stroke.  There is no mortality benefit (in fact, there is an overall increase in mortality with thrombolysis, primarily early on), and there is inconsistent evidence for an improved functional outcome across all groups (when considering the enormous heterogeneity of the reporting in the 12 major trials).  All of the reviewers mentioned above, however, have recognized that within the data, there is evidence that there must be some groups of stroke patients who have benefited from thrombolysis, but to date it is NOT CLEAR WHO IT IS (because of the multiple confounders with time to therapy, age, severity of stroke, amongst others).  Thus, we cannot yet definitively say that we are either harming or helping any subset of patients, or more importantly, the individual patient as they come through our ED door with an acute stroke.

This would all be fine and dandy – there are many therapies that yet ‘show promise’ but are not yet accepted, and are awaiting definitive proof (or, the absence of nullification).

But the problem here is that many of us work in departments where the neurology team FIRMLY believes that acute strokes ought to be thrombolysed.

Regarding the fence sitting – if you are one of the camp that believes that thrombolysis has little mortality and morbidity benefit to the patient in front of you, then you may think you are NOT a fence sitter, however if you then allow the neurology team to administer the therapy to the patient, because of the oft-spouted line that the Stroke Team will bear responsibility for the patient, long after your measly 4 hours is up, then perhaps you do still sit on that fence, more in action, rather than thought.

I bring this little personalized rant up, only in that I found myself using that line whilst teaching registrars, and then examining the principle behind it, trying to weigh up whether this was a cop out, or it was a wise path to take.  Hence the dichotomous title.

As a final little debate, I’d like to share a few of my internal conflicts about this situation.

FORthis being a position of wisdom, sense and perspicacity

  • There may not be any clear evidence for or against, therefore coming out punching on one side or the other may prove to be utterly incorrect come the next major definitive trial and a reversal (insert sense of optimism here)
  • It is not appropriate to fight this out over individual patients – it is tough enough to try and practice beneficence, without showing the patient that this may be in doubt
  • The neurologists are a smart bunch – it is presumed that they feel the data ought to be interpreted in the best interests of the individual and the population

FORthis being a position of pusillanimity and possibly cowardice

  • If you strongly believe that harm can be done to your patient, would this not be the time to intervene, or perhaps you may be less likely to identify those patients who the Stroke Team may consider for lysis?
  • By being complicit in a system that prioritises these patients for acute thrombolysis, are you not possibly diverting resources away from other patients, in the community, in the pre-hospital setting and the Emergency Department?

OR – could we completely destroy the binary/dualism theme and introduce another variable here?

  • Should we ‘hold fast’ and hope that, like has happened with reperfusion in acute myocardial infarction, technology and new, possibly physical (as opposed to pharmacological), interventions will soon burst onto the evidence-based stage?
  • CT perfusion scans may be a far greater diagnostic and stratifying tool than we have presently.  In combination with CT angiograms they have the ability to quantify not only cerebral perfusion, but cerebral blood volume, thus confirming salvagability of cerebral parenchyma REGARDLESS of time.
  • Interventional radiological embolectomy/direct catheter thrombolysis techniques – are they the new PCI?
  • Again – both of these topics are going to be debated in the first 6 months of next year, at the places to be (Sydney, and Broome)

A meagre mention should always be made regarding stroke units – the NNT for stroke units compared with general ward care for patients who were living at home at the end of 5 years? It would be a wonderful thing to see a drug have that kind of power. Stroke unit beds are monumentally short in Australia. Does seem rather a shame in the face of this evidence.

It seems a difficult jungle to navigate at this time.

Please add comments, join the discussion (although quantum mechanics purists, consider your retribution comments pre-empted)

I am seriously looking forward to conversations at the upcoming extraordinary conferences

ACEM 2013
SMACC 2013

And my final word? 

Wouldn’t it be a fine thing to tally up all the money spent on these trials, and compare that to a known stroke intervention – primary prevention?

Print Friendly


  1. Ian Rogers says

    Michelle you have so neatly summarised many of the issues around stroke thrombolysis.

    To make you feel better about calling that stoke team just remember that stroke neurologists are the gatekeepers to stroke units. Getting more of your patients into stroke units is getting them to the best evidence based therapy in stroke. The group benefit from access to stroke units more than outweighs the possible harm from thrombolysis. So dont feel guilty about it. You are actually manipulating the system as all good emergency physicians do to get the best care for as many patients as possible.

  2. says

    Hey Michelle,
    Just to be clear:
    If my cat is injected with a substance that (randomly) may or may not be tPA then I should put it in a box and stroke it for between 3 and 4.5 hours, or until Neurology arrive and push me off the fence?

    Makes sense now
    et al

  3. Manrique Umana says

    Amazing compilation Michelle!
    It puts together many different analysis and opinions from experts on data we’ve been reading and hearing so much about.
    I Recently sat with our Neuro Dept. staff and when I brought this up they were almost offended to say the least. So now I just sent them your post by email (hope you don’t mind) to see what they have to say.

  4. Mya Cubitt says

    And the scariest bit for me -- after you have decided to ‘sit on the fence’ and feed your patient to the stroke call system, you sit in to watch the process. Have you ever heard a neuro reg do consent for this stuff? Doesn’t sound like its explaining any evidence that I’ve read. They sell it like its a no brainer, and worse, they truelly believe that it is. I feel like a sell out every time I call a stroke call.

  5. says

    Just a fantastic bit of writing @Eleytherius. I am really impressed and intend on wide distribution to our group here.

    I have to agree with Ian -- the greatest good the the most people is from Stroke Reheb Units but we can also use discretion around our referrals.
    And I have had similar experiences to Mya regarding the consent process for stroke lysis.

  6. Kesh Khullar says

    We discussed the results of IST-3 at our Wednesday registrar teaching sessions and all came to the same conclusion that there were no benefits to t-PA, and the harm in the first 7 days weren’t necessarily spelt out to patients during the consent phase.

    In trying to ascertain how we could get the message across to neurologists regarding the data and results (and in fact the lancet paper clearly states in their own conclusion that the primary outcome showed no statistically significant benefit!), we were informed that stroke thrombolysis is now “Standard Practice” -- meaning in ED, even if we felt it was inappropriate to lyse these patients, it didn’t matter because hospital policy, government policy, health practices and “standard practice” dictate that we are obliged to inform the neurologist team of a potentially thrombolysable stroke, and then leave the patient in their hands to arrange CT scanning, consenting followed by administration of t-PA, in our Resus area mind, so that we are left to deal with the complications when they occur. Although considered a “neurology” patient, when faeces hit the fan, no doubt we are called upon to clear up the mess.

    It seems that unless more trials are done to confirm or refute the advantages of t-PA, or a big centre takes the lead in saying, “sorry but we’ve looked at the data and are not going to let you guys cause potential harm to our patients” so that the policy makers listen to our grievances, we are all left sitting on this proverbial fence -- and even if more trials are done, in the interim, how much harm are we still enforcing upon our patients?

    It baffles me that this “standard practice” line is dictating our management of patients, but unless policy changes and the bigwigs sitting upstairs read the results of this paper, and then decide that maybe a rethink is needed, nothing will change.

    Incidentally, I too am statistically inept, and still managed to understand the results of IST-3, so I’m sure the policy makers can read and understand what almost every emergency physician I’ve come across both in person and in the social media world, is talking about.

  7. Minh Le Cong says

    hey thanks Michelle and the comments here!
    I find it odd that earlier this year, there was a debate on various FOAMed and web based EM sites on the merit of giving lysis to submassive PE.
    Most EMCC docs argued that you consent the patient but err on the side of recommending lysis to minimise the risk of permanent pulmonary HT or the pulmonary cripple syndrome.
    despite no proven stat sig mortality benefit…

    Now with this stroke lysis debate, it seems we are arguing that lysis is too risky and there is no proven mortality benefit so we should avoid it? although there is some evidence of improvement in functioning.

    in other words, I cannot blame the neurologists for arguing that any improvement in functioning is worth the small risk involved in lysis. and when we recommend lysis in submassive PE despite the advice not to by pulmonologists, I dont think our logic is any better.

    so for those who point the finger at the neurologists, just remember what we might think in consenting a patient for lysis in submassive PE and if it is any different logic.

    In the far north here, stroke lysis is not routine or standard of care, basically because we have not had a neurology service until recently. The ED docs have resisted a proposal by the neurology service to start an acute lysis protocol, based on the lack of convincing evidence. I have heard Newmans podcast and tried to read the literature and must admit, there is nothing convincing but some suggestion of improved functioning in some patients. I have only ever seen two patients given lysis, both for massive CVA, with no success and in fact both patients died eventually but it was not due to a ICH from the lysis.

    I personally would not consent a patient to have lysis for CVA, unless it was a massive CVA and it was only chance of any meaningful recovery.

    But I do not condemn others for having a different opinion and consenting patients for lysis. I dont think the evidence is clear that you are definitely causing sig harm by stroke lysis as long as you keep to the trial criteria in the studies.

    if the patient or NOK wish to accept an elevated risk for first week or so, to try for a chance of better outcome eventually, twho are we to judge that value for their life?

    Newman stated it very well in his podcast. I am not sure how many have listened to the entire session as it is a marathon but he said something very wise. He said if he is asked his opinion he will give it and advise against lysis but he does not openly push this on patients as he feels it is unfair to expose them to medical confusion, when the stroke doctor consents them for lysis.

    I agree. we might have our professional differences of opinion, but this should not affect patient care in a time of stress and uncertainty, unless a clearly dangerous thing is going to happen.

    This is not sitting on the fence. this is respecting the fact we do not have all the answers now, yet we still need to work together for the patient who is likely trusting us all to look after them to the best possible knowledge but also to make some challenging decisions when the evidence is just not there.

    Of course we will get it wrong from time to time, but thats ok. thats medicine

    • says

      Minh, you make, as always, excellent points.
      Just a couple of responses:
      1. I certainly do not condemn the neurologists. I strongly believe that they, as we do , want the very best outcomes for their individual patients, and the population. What I find interesting about this debate is that how a very well planned and executed trial, could end up with interpretations that swing so wildly, between favourable, and not, depending upon who is authoring the conclusions. And yes, you are right, that is medicine. This is an opinion piece only, and I really do not know in which mortal plane the cat is. I agree that David and Ashley put it far better than this measly piece -- and I truly believe their conclusions, which is that thrombolysis really should only be given in the context of a clinical trial, not as a fait accompli, a given, a drug of proven benefit, and in a situation whereby resources are channelled towards getting all of these patients in, lights and sirens, with large pervasive advertising campaigns. I will be happy, nay grateful, to be wrong in 5 years time, when we prove that thrombolysis, or another clot removal technique, truly benefits a correctly identified subgroup of patients without causing the balance of harm.
      2. I am a little wary of the comparisons between the utility of thrombolysis in different pathophysiological states; with submassive PE, the intracranial haemorrhage rate is lower -- ~ 2%, rather than 6-10%, so maybe the balance of NNT is much better at the severe end of the submassive PE spectrum, significantly less than with strokes (and Cliff Reid has an excellent piece on thombolytics straddling the point of equipoise in submassive PE). But I do agree, that mortality is not necessarily the outcome we should be chasing. Having said that, if there is rise in mortality (in the case of strokes, early, but then equalising late), then surely the burden of proving improved outcome should be even higher.

      But I do so agree with you about the difficulties in this situation. It can only be healthy to engage in debate about these issues. Medicine has had a long and occasionally ignoble history of coming out firing in an area of therapeutics, only having to retract that position in a rather titanic reversal.

      The point of this piece was really a personal exploration of the challenges when it comes to how to interpret evidence. It was not meant to change anybody’s practice, or persuade anybody of anything, but simply to continue a debate which I believe we all think is important -- how do we decide the best therapy for the acute stroke patient -- based on the current evidence.

      • Anand Senthi says

        agree Minh, the patient is in a difficult and vulnerable situation and don’t want to unnecessarily confuse them, and I take note of Newman’s comments on the subject. On the other hand I do feel obliged to ensure they are making a truly informed decision about something so important. If the neurology reg is providing a biased opinion and he/she hasn’t even read the evidence, I would feel duty bound to inform the patient of the opposing view so they should make an informed decision.
        There are lots of points in Emergency Medicine investigation and treatment where the decision regarding whether to proceed is not clear. I manage those similarly by providing the best evidence in the most usable and understandable format possible and let the patient be involved in the decision. I see no reason why this situation should be any different.

  8. Kesh Khullar says

    Just wanting to clarify, as Michelle suggests, I think we all would love better evidence to show improved outcomes in CVA’s -- and I too have no beef with neurologists -- merely the acceptance of a therapy as standard practice -- as it is here in Vic (as per my consultants when we had this debate) -- as opposed to a trial therapy -- I think if patients or their NOK’s were made aware that evidence is still needed to show an overwhelming improvement in outcome vs the risks, and that this was still in trial stages -- then I agree, if consent is obtained in light of this, patients should be given therapy -- its the consent process and the fact that the trials themselves are still failing to show an overall greater risk/benefit improvement in patients, yet setting it out as “standard therapy,” that perturbs me.

    Another interesting point if I may, is that when you talk to some neurology registrars, many haven’t even read the results of IST-3 and seem to be somewhat indoctrinated into the belief that t-PA is some sort of wonder neurologists drug -- which I gather comes from the top down approach of persuasion by the drug companies to their seniors. As a junior registrar, although I miss all the free pens and delicious lunches that my fellow registrar colleagues get at their weekly grand rounds meetings, I’m kind of glad that our ED sessions aren’t funded -- guess I’ll just have to keep on begging for pens off my nursing colleagues when mine runs out!

  9. says

    Thanks Michelle et al

    I too am conflicted on this….much like the evidence.

    I remember the 1st patient that I was involved in giving thrombolysis to -- about 6 years ago now as a junior reg.

    Gentleman in his early 60′s admitted overnight for workup following a TIA earlier that evening. No beds on the ward so he was in an ED cubicle overnight. At around 4am he was found collapsed on the floor with a dense hemiparesis, facial droop, slurred speech , the works. His obs had been fine half an hour earlier. CT showed no bleed so neuro came along and thrombolysis was started. When I was going off shift at around 8am he was sitting up in bed having breakfast. He was a nice guy, unmarried, living alone, and had no close family. He had no real recollection or appreciation of what had just happened/ or indeed almost happened to him. Perhaps had this happened at home he would have spent the rest of his days in some facility completely dependent for even the most basic tasks like going to the toilet and getting dressed; or maybe he wouldn’t have made it to hospital at all!

    BUT…. perhaps he would have recovered spontaneously without any treatment. Maybe the tPA did nothing for him -- he’s just lucky he didn’t have a bleed as well.

    I was lucky enough to spend 3 months on rotation with a large neuro/stroke unit when I was in Oz. Lucky I say because as you pointed out, they are a smart bunch and also I found a very nice bunch. Until that is you raise questions or concerns about the evidence for thrombolysis in stroke -- and we all know that there are questions -- what the answers are we’re not 100% certain -- but everyone must accept that there are questions

    But if you raise these with them, it is as if you have said something offensive about their mother!

    I agree with the above concerns that there is not enough evidence to justify this as standard practice -- and this policy puts us in an awkward position. My message to the neurologists is that the data is the data and we should not be afraid if it does not happen to show what we hoped it would show. I think they bring a bias to the table and have performed some statistical jiggery-pokery in the case of IST3 to make the results look better than they are.

    I am also concerned about the closeness of the pharma industry in all of this and their involvement with the trials and the senior people who are involved in running them.

    Equally as EPs though we need to be unbiased about this. I wonder, being devil’s advocate here, if at times we bring our own bias to it, ie being so sceptical about it that we just decide we want the whole thing thrown out and perhaps take the polar opposite view to the neurologists. It’s in our nature almost to be at loggerheads with some inpatient team (of “experts”). The answer lies somewhere in between.

    Either way it seems that there is really good discussion and debate on this at EM conferences and in the #FOAMed blogosphere mostly among emergency physicians. And maybe the strokologists have their own really good debates/discussions in their own forums (??fora) coming form their own point of view. But the 2 in my experience rarely seem to meet.

    I think there is a good (#FOAMed) opportunity here -- maybe we could invite some of these nice, smart neurologists into the fold and perhaps have a google hangout/online debate. It may be good to have one/some of these heads at SMACC2013 when Domhnall gives his talk. And if they have an Irish accent too, well then that’s a bonus.

  10. says

    Great comment John. I agree that when I have attempted to have sensible discussions on this with my very nice and very smart neurological colleagues, the response has universally been anecdotal stories about the Lazarus effect rather than a real discussion around the data. Story trumps evidence. Woman’s Weekly beats a decent RCT hands down it would seem. So…we need better stories!

    • John Johnston says

      Just as a brief aside, it may interest people to know of a perverse real world example where there have been adverse outcomes as a result of blind acceptance of the “best practice ” argument for tPA.

      To set the scene we have a large active stroke service in inner eastern Melbourne but no neurosurgical service on site. Ambulance services in Victoria have been a part of the state health department stroke initiatives particularly in ensuring transport to stroke centres for stoke unit care and to minimise time to thrombolysis. They have supervising clinicians that enforce protocols mandating all “stroke” like presentations go to the nearest stroke service. Unfortunately this also often includes clear cut neurosurgical emergencies like the hypertensive 50 year old man with sudden onset headache, vomiting and altered conscious state. A number of these patients, who were certainly never going to benefit from tPA, have done poorly which may have been preventable with reduced time to neurosurgical care.

      Yes we do try and divert them when they are called in as stroke signals!! Especially the young previously well clearly candidates for neurosurgical care.

      Hopefully increased awareness of the evidence, or lack of it for tPA, will help at the higher level discussions previously dominated by enthusiastic neurologists, to at least enable some discretion in our situation.

      This is probably an anomalous situation due to the location of our services, but serves as an example of how the evidence, if used in a rigid manner with no common sense and can result in bad outcomes.

    • John Johnston says

      Just as a brief aside, it may interest people to know of a perverse real world example where there have been adverse outcomes as a result of blind acceptance of the “best practice ” argument for tPA.

      To set the scene we have a large active stroke service in inner eastern Melbourne but no neurosurgical service on site. Ambulance services in Victoria have been a part of the state health department stroke initiatives particularly in ensuring transport to stroke centres for stoke unit care and to minimise time to thrombolysis. They have supervising clinicians that enforce protocols mandating all “stroke” like presentations go to the nearest stroke service. Unfortunately this also often includes clear cut neurosurgical emergencies like the hypertensive 50 year old man with sudden onset headache, vomiting and altered conscious state. A number of these patients, who were certainly never going to benefit from tPA, have done poorly which may have been preventable with reduced time to neurosurgical care.

      Yes we do try and divert them when they are called in as stroke signals!! Especially the young previously well clearly candidates for neurosurgical care.

      Hopefully increased awareness of the evidence, or lack of it for tPA, will help at the higher level discussions previously dominated by enthusiastic neurologists, to at least enable some discretion in our situation.

      This is probably an anomalous situation due to the location of our services, but serves as an example of how the evidence, if used in a rigid manner with no common sense can result in bad outcomes.

  11. says

    Great post Michelle. Thank you. The rest of you already made just about every good additional point you could ever think of too.

    So, I only have one thing to add which is to disagree that the experts in a field (in this case neurology) necessarily have the best judgment on these issues. I think EM docs on this particular list (#foamed followers) are as likely to have read and understood the literature in question, especially with respect to NNT.

    Specialists tend to have fealty to “standard of care” and in some cases have financial conflicts of interest are impeding their ability to read these studies without bias. We aren’t biased by that because we aren’t the ones billing for these procedures.

    So I think it’s not necessary my default to defer to the judgment of specialty teams in areas where the data are in opposition to their plan. Then again my hands are 100% tied!

    • Anand Senthi says

      100% agree. EM physicians lack the conflicts and are equally capable of interpreting the evidence. Absolutely no reason to have to defer to neurologist’s opinion on this issue.

  12. says

    Thanks Michelle. You have summarised this most eloquently. It has long discomfited me that tPA has become accepted therapy, and, as is pointed out “policy”, at least as far as the Australian and UK governments are concerned.

    I’d be less kind than you to the neurologists. I’m less happy than you that they are prepared to do an often incomplete assessment (i’ve never seen a full NIHSS done, except when I’ve downloaded the form and done it myself), often from a distance and then prescribe a drug with a high complication and mortality rate (NNH 13 and NNKill 27, from NINDS, which is what we use, IIRC), whilst we pick up the mess if it goes wrong.

    I’d also question the validity of consent taken when there has been a “brain attack”. Sure I know you’ve had a hemispheric stroke with motor symptoms, but can I be truly sure, especially in an elderly patient population with comorbidities that your cognitive function isn’t impaired, even slightly. The time sensitive nature of consent for tPA is also problematic for me -- with other similarly fraught procedures in similar populations -- proximal femoral fracture surgery, for example -- there is the time to think about the decisions, discuss them with relatives and for medical roblems to be corrected -- we have none of these in acute stroke.

    I tend to practice my granny medicine. Would I want tPA for my 82 year old granny? Probably not. As regards my personal choice? At 37, if I had a dense dominant hemispheric stroke, probably. But I’m still not sure and I’ve thought about this a lot and I’m quite bright. Is it fair to ask our patients, who haven’t disected NINDS and IST3 at journal clubs to make a similar, risky decision?

    Thanks again,


  13. Minh Le Cong says

    remarkably TPA is licensed by FDA and TGA for stroke lysis , but TGA rejected the application to extend to 4.5 hrs time window.

    BUT TXA is not licensed for trauma care in both countries, USA and Australia!

    on the face of it, if the Prime Minister of Australia, was rushed into your ED with an acute stroke, And she met NINDS criteria, would anyone stay the hand of the person who wishes to deliver the TPA?

  14. Minh Le Cong says

    equally, I bet you all still administer adrenaline in cardiac arrest, despite the lack of proven mortality benefit.

    my point of course is that when the situation is dire and the stakes are high, we make judgement calls.

    as clinicians, the evidence swings both ways.

    • says

      Minh, I think adrenaline in cardiac arrest is a bit different, in that the person is already dead, so the drug doesn’t make them deader.

      The main harm of adrenaline in cardiac arrest is ROSC without subsequent good functional recovery, and all the resource intensive ICU care that ensues, including the unmeasurable harm to relatives etc.

      The harm we worry about in stroke lysis is that the patient we give the drug to bleeds and/or dies as a direct result of the drug we give, who may have had improvement without this drug. If the benefits are clear, then we can consent for that and wear occasional bad outcomes in the business of a high-stakes game. But the benefits are at best not clear, and the harms are undeniable.

  15. Mya Cubitt says

    AliG makes a great point re consent and competency issues. I questioned a neurology reg on this topic once and she said she was following her ‘duty of care’ and didn’t require consent for thrombolysis of stroke……….ooooooo.

    • Seth Trueger @mdaware says

      Once at a joint Stroke/Neuro/EM conference, when asked about consent, the Director of the Stroke Unit said: we don’t get the patient’s consent — we inform them what the treatment is going to be.

      That being said, I do believe that most Neurologists believe that they are helping patients. I don’t know whether tPA works or not. But some smart people who question the evidence behind tPA seem to make compelling arguments!

  16. james cuthbertson says

    Shout out to Kesh -- that was a robust little discussion, wasn’t it…

    My thoughts:
    - it probably helps some people, but studies to date haven’t identified who they are yet

    - as such, I would suggest the current population-based use is of little benefit, significant cost and big-ticket (ie. new death) harm

    This conspires to put ED staff in a difficult position, as Michelle so eloquently described.

    I tend to offer my opinion/advice, but only if specifically asked. I do make sure to be present during consent, to encourage the Neuro guys to be complete. And I specifically ask patient or NOK about their wishes re. Intubation/’life support’ if things go bad -- since I’m the one that may have to do it. I think this helps dissipate some of the ‘clotbusting wonderdrug’ aura/urgency and bring the patient/family discussion back to a more realistic risk-benefit thing.

    I also point out that if ICH does occur, there is little that can be done except to stop the infusion and see what happens. I think the consent process the Neuro chaps use needs to mention that, if the side effect does occur, it is pretty much curtains.

    As an aside, I have also found that when engaging proponents in a discussion about the evidence base, it often ends in a Lazarus anecdote. I tend to resist pointing out the cases I am aware of that ended in bleeding and death and disaster -- but I do mention that if a population-wide intervention is best advocated by (admittedly nice) success stories, then things may need a rethink.

    “tPA -- not exactly TXA”

  17. Minh Le Cong says

    Michelle et al, thankyou for this stimulating discussion as it has made me go over the whole issue and consolidate my views on it. I even went so far as calling up my brother in Adelaide who is a consultant general medicine physician at one of the major teaching hospitals. He says his neuro colleagues run a strokelysis protocol along with the ED service and it is pretty much standard of care in his shop. There does not appear to be concern amongst the internal medicine service there as to the merits of a stroke lysis protocol. I discussed the debate as it is presented here in Michelle’s expose piece ad he admitted the evidence does not show a mortality benefit, but the rationale is it will improve functional outcomes in some, and there is some increase risk of ICH .

    He however was somewhat surprised by the debate.

    I put this to you all to consider.
    The next time you are running a cardiac arrest code and asking for the tenth dose of adrenaline to be given, if someone current with the latest resuscitation evidence, challenged the notion of giving adrenaline at all, would you quote the occasional Lazarus anecdote you have witnessed or lay the claim that it is standard of care and you are not going to change until you see better evidence?

    if the answer is yes, then the defence rests, your honour.

  18. Kesh Khullar says

    Minh, I hear what you’re saying re adrenaline in cardiac arrest -- but aren’t these patients already dead -- essentially -- whereas our stroke patients with dense hemiplegias -- yes they may not go on to have a significant quality of life -- but are still living.
    I guess the question comes down whether we as physicians wish to instill more harm onto these patients (and I’ve heard it likened to legal euthanasia in some cases!) or wait until the data shows an increase in improvement in functional outcomes in the correct subset of populations.
    At present, as had been highlighted by many above, we still don’t know who those are, and I think we all would agree that identifying this group, in order to have a meaningful population with improved outcomes following stroke would be something that we’d all support.
    Ultimately it comes down to selecting that right subgroup -- which as yet -- from the current data -- is still yet to be decided. Until then, watching neurology registrars consent patients who have mild weakness, or transient symptoms with some improvement, still being included in the subgroups for t-PA makes me wait in trepidation for that call from a nurse saying they’ve dropped their GCS.
    Improved outcomes in stroke is something we all want -- I for one just want more trials to assist in selecting the right group at the right time, before we marvel at this “wonder drug,” and consider it “standard practice.”

  19. Minh Le Cong says

    Thank you Kesh
    That’s a value judgement you are making
    If a patient places more value in quality of life over death , then who are we to judge ?
    This is an approved therapy … It is not even off licence use
    Sure we need to keep the research up
    It is standard of care for ischemic stroke just like adrenaline is in ALS

    You call the neuro reg , you call for the adrenaline
    This is what you do
    Sure it’s not perfect

  20. james cuthbertson says

    I buy the dense hemi/arrest/adrenaline argument -- but I am not sure this stands up when you consider lysis in those with less severe deficits. When the “stakes are lower”, or the deficits are improving, its harder to justify a treatment as a Hail Mary or salvage treatment. Asystolic people may have nothing to lose (or gain) from adrenaline, but I am uncomfortable in stretching this to justify lysis en masses.

  21. Sascha Berning says

    Dear Michelle
    and all the other folks thank you so much for the debate!
    I am not sure if I‘m ready to confess it -- but: YES -- I‘M A NEUROLOGIST -- one of these chaps playing with rtPA for 7 years now in a large German Hospital. Things are different here, we actually don‘t have a EM-doc, but a 24/7 Neurologist within the ER.
    Three comments:
    - I have to admit that I was dazzled and was not aware of the thin ice I was playing on -- until the debate went of this year within the EM community ;o)
    - For some reason we have not encountered as many relevant bleedings as stated as long as you don‘t lyse into uncontrolled hypertension, effective anticoagulation, demarcated stroke and the severe subcortical arteriosclerotic encephalopathies (and they are for sure within the 3h or up to 4,5h with a perfusion/diffusion miss-match). I know that doesn’t replace EBM!*?
    - The Neuro folks here have recognized, that we cannot lyse large vessels (= ICA or MCA) and thus we are doing them in a Neuroradiology-cathlab for almost 3 years ➝ much much better! This will come up and I think might cool down the debate (which is fruitful).

    Take care

    • says

      Thank you so very much, Sascha, for your insights and thoughts on your experiences (and your bravery ;-) ). In a mature and developed system such as yours, it sounds as though you are indeed finding the subset of patients that do well from thrombolysis. And maybe that’s the nubbin of the whole debate -- how do we, in the current climate, be as sure as we can be that we are selecting the right patients, without doing harm to those that will not benefit?

  22. Dean Powell says

    Hi folks,
    Came late to this conversation, but here are two thoughts:

    What would happen if fairly good evidence came out that showed definite harm? How easy do you think it would be for the “Standard Practice” to be stopped? I suspect there are few truly open minds in either camp and we would be fighting people who want to protect their reputations aggressively.

    Secondly, regarding the consent issue -- Although neuro do the consent, I think we can stipulate that they use OUR standard proforma when obtaining that consent in our ED. So we can be sure at least the registrar doing the consent gets the paucity of evidence and maybe even the patient and their family.

    Just a quick few thoughts.

  23. says

    Since I’m just a stroke-addled survivor that is probably alive because I got tPA within 90 minutes, you can ignore my non-medical comments. tPA doesn’t stop most of the neuronal cascade of death processes. I know Dr. Michael Tymianski has stated that 1000 neuroprotection trials have failed in humans that worked in rodents. That is where I think the money for research should go. Especially since its now been proven that mouse inflammation does not have same factors as humans. Prevention of the neuronal cascade of death would probably reduce the dead and damaged neurons enough that spontaneous recovery and neuroplasticity may be enough to get close to 100% recovery.

  24. says

    Like with any intervention, my approach is to let the relatives decide through a careful explanation of the situation

    1) inform the relatives of the estimated risk:benefit based on what was demonstrated in the NINDs trial.
    2) Explain the probability of survival to independence based on initial severity without treatment based on TOAST data
    3).Advocate strict criteria use of the NINDs criteria i.e No lysis > 3 hrs, age>80, hypertensive, on warfarin, time onset not known etc. If these are deviated then I explain that the estimations will be inaccurate and can increase the risk:benefit.

    I would find a table such as this helps explain the probabilities (and ranges) assuming average NIHSS 14.

    Noting that meta-analysis of all stroke trials would indicate one less green box and one more red box