WSJ on Stroke: TPA is a wonder drug for stroke!

Summary: Neurologists are excellent with acute stroke, EM docs are stupid, and anyone against TPA is backward.

I have read this twice, and am of the opinion the reporter was assigned the topic “stroke” to write about, then talked to a neurologist who was a big fan of TPA for stroke. It is filled with “they’re stupid if they don’t embrace…”, TPA for stroke (which I’ve written about before), and then it just keeps getting worse:

Stroke victims are often taken to wrong hospital

Monday, May 09, 2005
By Thomas M. Burton, The Wall Street Journal

Christina Mei suffered a stroke just before noon on Sept. 2, 2001.
Within eight minutes, an ambulance arrived. Her medical fate may have
been sealed by where the ambulance took her.

Ms. Mei’s stroke, caused by a clot blocking blood flow to her brain,
occurred while she was driving with her family south of San Francisco.
Her car swerved, but she was able to pull over before slumping at the
wheel. Paramedics saw the classic signs of a stroke: The 45-year-old
driver couldn’t speak or move the right side of her body.

Had Ms. Mei’s stroke occurred a few miles to the south, she probably
would have been taken to Stanford University Medical Center, one of the
world’s top stroke hospitals. There, a neurologist almost certainly
would have seen her quickly and administered an intravenous drug to
dissolve the clot. Stanford was 17 miles away, across a county line.

But paramedics, following county ambulance rules that stress proximity,
took her 13 miles north, to Kaiser Permanente’s South San Francisco
Medical Center. There, despite her sudden inability to talk or walk and
her facial droop, an emergency-room doctor concluded she was suffering
from depression and stress. It was six hours before a neurologist saw
her, and she never got the intravenous clot-dissolving drug.

Woah, pardner. There has to be more to the story than this; if not, there’s an idiot for an ER doc. As the vast majority aren’t idiots, the above makes no sense.

However, the ‘..almost certainly would have..administered an IV drug to dissolve the clot…’ is only about 1/10th of the story. However, just reading this article, I’d be ticked off if I didn’t get TPA for a stroke.

Stroke is the nation’s No. 1 cause of disability and No. 3 cause of
death, killing 164,000 people a year. But far too many stroke victims,
like Ms. Mei, get inadequate care thanks to deficient medical training
and outdated ambulance rules that don’t send patients to the best
stroke hospitals.

Uh, deficient medical training? Based on what fact is this assertion made? There exists a lot of controversy about the use of TPA in stroke, and that isn’t because of ‘deficient training’, it’s because very smart people have looked at the literature, evaluated the risks, combined those with their clinical experience and then came to differing conclusions.

Over the past decade, American medicine has learned how to save stroke
patients’ lives and keep them out of nursing homes. New techniques
offer a better chance of complete recovery by dissolving blood clots
and treating even more lethal strokes caused by burst blood vessels in
the brain. But few patients receive this kind of treatment because most
hospitals lack specialized staff and knowledge, stroke experts say.
State and county rules generally require paramedics to take stroke
patients to the nearest emergency room, regardless of that hospital’s
level of expertise with stroke.

There’s a paragraph about stroke care needing to be like Trauma care, deleted.

Eighty percent or more of the 700,000 strokes that Americans suffer
annually are “ischemic,” meaning they are caused by blockage of an
artery feeding the brain, usually a blood clot. Most of the rest are
“hemorrhagic” strokes, resulting from burst blood vessels in or near
the brain. Although they have different causes, both result in brain
tissue dying by the minute.

Several factors have combined to prevent improvement in stroke care. In
some areas, hospitals have resisted movement toward a system of
specialized stroke centers because nondesignated institutions could
lose business, according to neurologists who favor the changes. In
addition, stroke treatment has lacked an organized lobby to galvanize
popular and political interest in the ailment.

Nobody I know wants to ‘prevent improvement in stroke care’. And the idea that an organized lobby is needed to push this agenda is ludicrous.

A big reason for the backwardness of much stroke treatment is that many
doctors know little about it. Even emergency physicians and internists
likely to see stroke victims tend to receive scant neurology training
in their internships and residencies, according to stroke specialists.

“Surprisingly, you could go through your entire internal-medicine
rotation without training in neurology, and in emergency medicine it
hasn’t been emphasized,” says James C. Grotta, director of the stroke
program at the University of Texas Health Science Center at Houston.

This is just twaddle. EM residency programs cover neurology of emergencies very well, thank you. It’s part of the EM curricula, and it’s part of the tests and training. IMHO, EM docs see way more acute strokes than do neurologists.

Many hospitals don’t have a neurologist ready to deal with emergencies.
As a result, strokes aren’t treated urgently there, even though short
delays increase chances of severe disability or death. Even if doctors
do react quickly, recent research has shown that many aren’t sure what
treatment to provide.

True, most hospitals don’t have neurologists available to treat emergencies (and those that do have trouble getting them to come in). That ‘lack of neurologists’ has nothing to do with a delay in stroke care. Delays are delays, and every hospital has a way to prioritize tests and treatments. Lack of neurologists doesn’t equal lack of priority.

For example, a survey published in 2000 in the journal Stroke showed
that 66 percent of hospitals in North Carolina lacked any protocol for
treating stroke. About 82 percent couldn’t rapidly identify patients
with acute stroke.

What? Couldn’t rapidly identify patients with an acute stroke? That’s just nonsensical. Any EM trained doc, or an ER nurse with 6 months’ experience can identify people with ‘possible stroke symptoms’ and start the eval right then. As for the lack of protocol, well, there’s a whole argument about checklist medicine, and whether protocols are helpful or harmful.

As with other life-threatening conditions, stroke patients are better
off going where doctors have had a lot of practice addressing their
ailment. A seven-year analysis of surgery in New York state in the
1990s showed that patients with ruptured blood vessels in the brain
were more than twice as likely to die — 16 percent versus 7 percent —
in hospitals doing few such operations, compared with those doing them
regularly. A national study published last year in the Journal of
Neurosurgery showed a similar disparity.

Okay, here we start mixing our apples and oranges; that was (probably) about aneurysm surgeries, very different for the vast majority of ischemic CVA’s.

Another major shortcoming of most stroke treatment
, according to many
neurologists, is the failure to use the genetically engineered
clot-dissolving drug known as tPA.
Short for tissue plasminogen
activator, tPA, which is made by Genentech Inc., has been shown to be a
powerful treatment that can lessen disability for many patients. A
study published in 2004 in The Lancet, a prominent medical journal,
showed that the chances of returning to normal are about three times
greater among patients getting tPA in the first 90 minutes after
suffering a stroke, even after accounting for tPA’s potential side
effect of cerebral bleeding that can cause death
. But several recent
medical-journal articles have found that nationally, only 2 percent to
3 percent of strokes caused by clots are treated with tPA, which has no
competitor on the market.

Some authors of studies supporting the use of tPA have had consultant
or other financial relationships with Genentech. Skeptics of the drug
point to these ties and stress tPA’s side-effect danger. But among
stroke neurologists, there is a strong consensus that the drug is

If they were pushing, say, Bush administration talking points and got caught taking money for it, they’d be pariahs to the press. Double standard for those pushing medical interventions, I suppose.

But, let’s not gloss over the side effect: taking an injured brain and making it substantially worse, in 6-9% of cases. That’s not an intervention to rally around, it’s one to be feared, and to be engaged only when the benefits very clearly outweigh the risks.

The article is very much longer, and reads like the neurologist dream agenda. The slant of the article is so pro-neurologist and so anti-EM doc, though, and gives short shrift to the other side of th TPA controversy, it’s impossible to take seriously.


  1. Bad Shift says:

    oh man this makes me ill…

  2. Obviously someone at Genentech pushed for this article! I say the neurologists change places with the ER doctor. That way when they have to tell the family of the pt. that just received TPA that they are now unconscious and unlikely to survive the bleed…they know what it feels like!

  3. Goat Whacker says:

    I read this article this morning and actually sat down to write a letter to the WSJ editor, but unfortunately real life got in the way. But I have seen the light! TPA for all! Hemorrhagic strokes! Migraines! Ruptured aneurysms! Week-old CVAs! Hypoglycemia! Whatever it is, TPA will make it better!

  4. Belinda says:

    See how hard it is to be a layperson trying to make sense of all this?

    I believe Ms. Mei’s main problem was being a 45-year-old female. For those of us “too young” to be having brain disorders, and especially if we’re women, depression/stress seems to be something of a default diagnosis. And in my experience, neurologists are the worst offenders in this regard.

  5. A friend of mine had a stroke about a year ago, 46 years old with extremely high blood pressure. I’m not sure TPA would have helped but I know that the hospital they brought him to could not administer the drug. I do not know the reason. His best friend knew the head of neurology at nearby hospital but could not get permission to allow another diagnosis during the critical time period.

    Today he is paralyzed on one side and has has blurred vision. It is difficult to know as a layperson if TPA would have helped.

  6. Gruntdoc,

    I read that article on the emed list, and its poorly done, but I do think you’re banging the wrong people. What’s your beef with neuros? This piece is a typical WSJ pro-big business drive by shooting. They are pro-Genentech, that’s what they are, not “pro-neurologists” (what does that even mean?).

    I do agree with you that one or two of the neurologists interviewed seemed to have made some unfortunate remarks, eg. Grotta saying that you could get through internal medicine residency without any “training in neurology”. That’s technically correct, I suppose – it is possible to do that in the shittier programs – but rather unlikely. His comment about EM programs is, in my experience, completely untrue. Although to be fair to him, Grotta might not be talking about current programs (which on the whole provide good exposure to the common emergency neurologic presentations), but programs in the past which produced the bulk of physicians staffing ERs today (most of whom are not even BCEPs). (How much formal training have FP programs provided in emergency neurology?) But I do think that if that is what he meant, he should have made it clearer. I am not an expert on FP or EM program pedagogy, so I cannot comment more than that, and defer to the experts.

    With respect to your claim that “EM docs see way more acute strokes than do neurologists,” I think you’re being as inaccurate as the article you’re complaining about, GD. Does an EM doc staffing a Level I academic center see more acute strokes than a Movement Disorders sub-specialist practising primarily outpatient neurology? Of course. Does an EM doc staffing Generic ER, Anywhere, USA see more acute strokes than the neuro-intensivists at Johns Hopkins or the General or some other Level I center?

    With respect to the actual heart of the article (why EM docs don’t like tPA), I completely agree that it’s pretty shitty: a huge part of the reason that many (most?) EM docs are very wary of giving tPA is simply that they are professionaly unconvinced that it’s a safe drug. It’s not that they don’t have the cajones/ are stupid. That’s BS. Most neurologist do not see this as some sort of “neuro vs. EM” issue and would have been embarassed for the writer for his naivete. You are being very unfair to them by saying this article is a “neurologists dream agenda”.

    Part of the reason many EM docs remain unconvinced is simply the way medicine is structured and the way strokes evolve. An EM doc sees a stroke patient for a very short interval of his disease – the acute part when the guy is in the ER. If an EM doc is in the habit of prescribing tPA, he will likely see only two outcomes: 1. Nothing much, or 2. Shit hitting the fan.

    The reason is that all the bad stuff with tPA happens right at the start – if you bleed from it, you bleed from it right there, in the ER. No one bleeds from tPA 30 days later, or even 3 days later. So if anything bad happens, it is the EM docs who “endure” it. However, the vast majority of the time, EM docs don’t see patients at the other end of the timeline, 3 months later, when not only have many of them survived, but thrive, almost as if nothing happened. The people who see both ends are the stroke neurologists. The PM&R folks see only the tail end.

    Because each of these specialists see the stroke patient at different stages, their personal opinion of the drug is often quite dissimilar. If you are told to give a drug to patients that most of the time you see either doing nothing much or else doing something terrible, it’s going to be hard to convince you it’s actually worth it.

    Now, whether tPA is “worth it” is something I suspect we’re all going to be arguing about for the next few years.

    I saw your last post on tPA, but didn’t open my mouth even though I disagreed. But since I’m posting here on this one, what the heck.

    (The following is addresed to your general audience, not specifically to you, GD).

    My opinion is that tPA is a dangerous drug that does however have a clear net benefit if prescribed in the correct manner. Lets unpack that sentence.

    Anyone who wants to understand tPA needs to read at least three papers, IMHO. From start to finish. These are:

    1. The original NINDS report in the NEJM Dec 14, 1995.
    2. Joanna Wardlaw’s 62 page Cochrane review 2003.
    3. The tPA Review Committee’s 85 page December 2004 re-analysis.

    For those of you who want a summary of sorts, continue reading.

    The first thing that people need to understand about tPA is that it does not behave the same way in all patients. This is actually true of every drug, but with tPA the difference is phenotypically very stark.

    When you give tPA to some stroke patients, they will die. There is no getting around this, and no other way to say it. Anyone who tries to talk down this aspect of tPA is not being honest with you. tPA directly kills some patients, and it does so in a very unpleasant way. It causes them to bleed into their brains. The mechanism by which tPA does this is by making it impossible for the patient’s blood to clot, and so tiny arteries in the brain spontaneously bleed uncontrollably, and the brain dies.

    However, tPA also does other things. It causes many patients to improve tremendously. They improve in much greater numbers than you would expect if you just sat there and held their hands (which, generally, is about what you can do if you didn’t give tPA). The extent of the improvement and how many improve depend on how you measure it, but in general you can see that the drug helps a lot of people.

    Nothing means anything without the numbers, so lets have a look at those, shall we?

    If you give 1,000 selected* stroke patients tPA:
    1. it will cause about 64 patients to get a serious bleed in their brains.

    2. about 170 patients, including many of those 64 patients who bled, will die.

    3. about 500 patients, however, will go on to do extremely well from the point of view of independent function. This means that these 500 patients will be able to feed themselves, bathe, dress, groom themselves, use the washroom, be continent of urine and stool, walk more than 50 yards, use the stairs both up and down, and transfer themselves from bed to chair. They will be able to do all this completely on their own, ie. independently. That is to say, these folks will basically be able to lead independent lives.

    4. about 160 others will be able to do about half of the above independently, but will need help with others.

    5. about 170 other patients will need a lot of help with the above, and will essentially be close to complete dependency (ie. completely sucky outcome).

    If you sit down and hold 1,000 stroke patients’ hands, and do not give them tPA:

    1. About 6 will get a serious brain bleed.

    2. About 210 patients will die, mainly from the stroke itself.

    3. About 380 will go on to do extremely well from the point of view of independent function, and be able to do all those things I listed above, independently.

    4. About 230 others will be able to do about half of the above independently, but will need help with the rest.

    5. About 180 others will do very poorly from a functional point of view, being almost totally dependent.

    As you can see, the main conclusions here are:

    A. More people bleed with tPA,
    B. More people die from the stroke itself without tPA,
    C. More people recover extremely well with tPA.

    What tPA seems to do is push people out from the “dependent”, disabled groups into the high-functioning, “independent” group. It unfortunately also kills a significant number on the other end through brain bleeds. However, the total number of dead seems to be the same, whether you give tPA or you don’t: this is because while it kills some patients by causing them to bleed, tPA also saves others from dying from the stroke itself. So the death rate balances out, ie. no difference. But there is a net advantage from a functional point of view because tPA pushes so many people through to the high functioning group from the “dependent” groups.

    Bottom line, if you get tPA, you will have a better chance of getting through the stroke without disability. You will also have a definite risk of dying. If you don’t get tPA, your chances of dying are the same, but you will also have a lower chance of escaping disability.

    Now, what are the important caveats to bear in mind? Look at the heading for the tPA list I wrote out. Why the asterisk* for “selected” patients?

    Because that’s crucial. The people in whom this tPA calculus works are not a random sample of stroke patients. They are a highly selected bunch. If you gave tPA to every single person who walked through the ER doors with a stroke, you will have a massacre on your hands. There is only a narrow population of stroke patients who are eligible for the drug, so that the chance of benefits will outweigh the risk of harms. If the balanced is tipped, you will kill more patients with the drug than if you didn’t do anything, and that’s bad.

    This is why you cannot run a stroke program at any old place. The people need to be trained. A good radiology service is a must. A formal neuro department with well-trained neurologists, taking emergency call is highly desirable. Some of the early community stroke programs that tried to emulate the NINDS trial result went completely off the tracks because of these issues. The Cleveland area program had a protocol violation of 50%, and their bleed rate was 15.7%, more than two times the NINDS trial rate, and these guys were basically killing more patients by giving them tPA than they would have if they offered nothing.

    So I agree with some who suggested in the earlier thread that this drug cannot simply be given by every ER. It can’t. It needs a highly specialized team of neurologically trained folks to do it right, and its unfair to hoist the inherent risks unto every average community ER.

    I do disagree with those who say that tPA is “not proven”, or should not be used. It can be, but it needs to be done in the properly in a center with a formal stroke program. When the Goat guy said in the other post on tPA that it actually couldn’t benefit patients, and that all the improvement was coming from resolving TIAs, he was basically talking out of his posterior. (Look at the figures above and compare the two groups)

    It is also an issue of informed consent. IMO, every candidate for tPA needs to have the odds explained to him (if he can understand it) or his decision maker, and they need to come to a decision.

    For me, personally, it’s not even a contest: if God forbid someone I love (or I) got a stroke, I’d advocate tPA without a moment’s hesitation. Because I know what a horrendous existence it must be to live completely paralysed on one side of your body, completely aphasic (ie. unable to express or understand human language), dysarthric, incontinent, with a defective visual field and sensory loss, unable to do anything on my own. A drug that did not increase my chance of death but did increase my chance of complete recovery is a no-brainer.

    Of course, I’d also like to be taken to Hopkins to get it.

  7. DensityDuck says:

    Of course, we could have been treating strokes with OFNE by now, but neurologists were all “you want to put TUBES? Into the BRAIN? And FLOUROCARBON?! What are you, some kind of MAD SCIENTIST?! Get away with your leech-craft and let our patients’ brains dissolve in peace!”

  8. Note, in the above commentary, I did not go into details about the controversy over the NINDS trial itself (immediately after the trial was published, it was pointed out that there was a slight baseline imbalance in stroke severity between the tPA group and the placebo group. A lot of people felt strongly that this negated the result, the most famous probably being Jerome Hoffman of UCLA). I completely respect this criticism, and in fact made it myself when I read the report.

    However, the issue has been re-analyzed to kingdom come, the most thorough going over being the tPA Review Comittee whose report I noted above, and their figures bear out the original conclusions: in a select group of patients, tPA has a net benefit in clinical outcomes.

    I have not talked about that in detail because I have no time.

  9. I believe Ms. Mei’s main problem was being a 45-year-old female. For those of us “too young” to be having brain disorders, and especially if we’re women, depression/stress seems to be something of a default diagnosis. And in my experience, neurologists are the worst offenders in this regard.

    With respect, I think this comment can most accurately be described as a display of ignorance.

  10. Goat Whacker says:

    “When the Goat guy said in the other post on tPA that it actually couldn’t benefit patients, and that all the improvement was coming from resolving TIAs, he was basically talking out of his posterior.”

    Thanks for your post, it’s very good. While I oftentimes talk out of my posterior, I was referring more to many ER folks relying on anecdotes to justify use of TPA, not that TPA never works. Sorry if I was unclear.

    I have been to several lectures by Jerry Hoffman and that obviously colors my faith (or lack of) in TPA.

  11. Belinda says:

    A MD, I most sincerely hope that my comment is a display of ignorance, and that the cases which occasioned it are truly flukes.

  12. There are some truths to the article, though it was written with the common layperson “ER” theme that 90% of those with catastrophic presentations survive the trauma room to make it to the next Budweiser commercial. I work in a “thatched hut” community ER 50% of the time, and in an “academic mecca” the other 50% of the time. I smell the smoke from my medical license burning every time I step foot in the “thatched hut”. To give TPA in the community hospital, we have to call in the Neurologist from home. God forbid you decide to Stroke after business hours. Once the Neuro goes Beddy-Bye, nobody is a TPA candidate. They decide that from home. From 9-5 almost every pt. (or family) is given the option of TPA. At Academic Mecca Hospital, a Stroke Fellow meets the ambulance in the ED. Both institutions are equally designated “stroke centers”. Another example of how we are the only specialty that truly operates 24 hours a day.
    Also, Did anyone else notice Genentech stock went up 3% the day of the article?

  13. I suppose it’s karma that my longest post draws my longest comment. There have been many excellent points made in here, and I thank all of you.

    I suppose I’m not too taken with the ‘Stroke Center’ designation as, IMHO, it is sought by some big hospitals for PR purposes without enough buy-in from the staff who have to do the work. Incidentally, it also gets a lot of hospitals who don’t want it off the hook to get better in their stroke care. I mean, really, it takes a timely CT and a careful exam and history. Every ED in the US has TPA, so what’s the magic in a ‘stroke center’?

    People are being flown or transferred to ‘stroke centers’ for eval and possible TPA from ER’s that are, IMHO, entirely capable of making that determination themselves.

  14. Unfortunately our hospital is wanting the “Stroke Center” title again. We had it a few years ago and the lack of interest caused it to fade away. Now the Admin is back on the bandwagon. Our typical treatment of acute stroke is a noncontrast brain CT and CT angio. If there is an occlusion on the CT angio, and no hemorrhage on the CT then off they go to Interventional Radiology for intra arterial thrombolysis. We don’t give TPA in the ED. We do alot of the intra arterial treatment, and it works well in selected cases. The majority of our Neurologist are AGAINST giving TPA in the ED. I am not saying this is the right way to do things, just our way.

  15. GruntDoc says:

    that actually seems like a more reasonable way to go about it. Were I to need TPA for a stroke myself, I’d much rather have it dripped slowly onto the clot directly than the current squirt and pray method.

    Does your radiology dept insist on waiting for a creatinine to do the ct angio?

  16. The radiologist (more specifically our Interventional Neuroradiologist as they liked to be called) do like to see a creatinine, but that can usually be obtained while the patient is being prepared for CT. Not usually a rate limiting step. We have done the CT angio without the Cr level back, but they like to see it before the intervention. If the Cr is high we can always treat with Bicarb, fluids and N-acetylycystine (that could be a whole other topic for discussion). For the most part this works very well. Our Nurse Admin for the ED is living proof that it works. She came in with a deficit, no abnl on Ct angio, but a 4 vessel showed a small clot that was succesfully intraarterially thrombolysed. She is as good a new. We have had several other successful cases and no intracranial hemorrhages that I am aware of.

  17. CardioNP says:

    To “A MD”:

    Don’t be so quick as to dismiss the observation by Belinda.

    I personally (a 43 yo F) have polyneuropathy with documented B-12 deficiency; unfortunately my neuropathic symptoms worsened despite aggressive B-12 supplementation. During this time I relocated to Sacramento from Los Angeles. Went to a private practice neurologist in Sacto who has been recognized by the AHA for his outstanding stroke work. Without even doing any workup, he had the gall to suggest that my symptoms were due to fibromyalgia and basically dissed the evaluation by my Harvard educated/trained internist from LA by saying that he probably did not do 2 pt discrimination on my feet in the correct manner. Needless to say, did not go back for follow-up. Have since seen an academic neurologist at UC Davis, have EMG documented demyelination in BLE, diffuse evidence of small nerve fiber damage and am being worked up for possible CIDP.

    Were I not a sophisticated medical consumer, I would be left with the wastebasket diagnosis of “fibromyalgia” suggested by the first idiot neurologist that I saw.

    The AHA has the “Go Red” program for a reason – women with CAD were not getting appropriate evaluation and treatment.

    While you may not perceive a bias in the way that women are treated in healthcare, I can tell you that I personally have experienced it, and it does exist.

    Fortunately for me, I have enough education to know better and seek a different provider when necessary.

  18. It often amazes me how poorly reasoning is taught and used in our society.

    To respond to CardioNP & Belinda in what I hope is the last exchange of this sort on this thread:

    I have had the displeasure of knowing, over many years, scores of extremely rude people who were obstetricians, nurse practitioners, corporate executives, at least one school teacher, molecular biologists, blacks, Asians of all kinds, Jews, whites of Western European descent, women, men, school dropouts, PhDs, joggers, sailors, bra-wearers, tie-wearers, and shopping assistants (this is of course an extremely abridged list). Some of them have had the further distinction of being both rude and stupid. I am sure every one of us can compile any number of such lists on any number of characteristics and attributes.

    I have, however, never, ever felt it at all edifying to make statements such as:

    “In my experience, blacks are the rudest people in this regard.”


    “In my experience, nurses are some of the stupidest people around.”

    The reason I think most reasonable people do not make such statements is plain: the data on which they are based – your limited personal experience of a few people – are meaningless, and do not bear the conclusion implied by the statement. For instance, I have met many rather stupid NPs, but happily know of many delightful and intelligent ones. And even if I hadn’t, I wouldn’t presume that my limited sample of a few NPs that I knew from personal experience was at all representative of them as a whole to warrant such hasty pronouncements.

    Ditto every other category of person mentioned above.

    I have no quarell at all with the suggestion that women have faced systematic prejudice in health care. Indeed, if you had asked my opinion on the matter rather than jumped to conclusions, CardioNP, I would have been glad to point out that prejudiced attitudes toward women in medicine is not merely some vague hypothesis – it’s a fact, and we have studies that show it. To quote just one example, we are only now starting to recruit trial participants that more acurately reflect societal demographics.

    I was merely pointing out the absurdity of implying that neurologists had poor attitudes towards women, based on a couple of unfortunate personal interactions with people who happened to be neurologists (one in the case of CardioNP, an unspecified handful in the case of Belinda).

    Now, Grundoc’s post here is on t-PA for the treatment of acute ischemic stroke, and I have no wish to spoil his thread with further debates about how very misogynist neurologists really are. In fact, I did not say a word when Belinda even replied, so charmingly, and apparently without a trace of irony, that she “most sincerely hope[d] that my comment is a display of ignorance, and that the cases which occasioned it are truly flukes.” (Think about that statement for a bit.)

    As such, I will not respond further to any posts directed at me on this topic, and will be very glad to let you go on believing that neurologists are misogynists, Santa Claus lives in Chile, or any other class of moonshine that piques your fancy.

    One is only sorry that Quine’s Elementary Logic has never been made required reading for all kids age 10 and up in our society. It would do so much good.

    GoatWhacker, Darren and Gruntdoc have made some good points that I should like to respond to over the weekend, and I hope we can keep this thread free of further distractions.

    Kind regards.

  19. Unfortunately, more and more people, fostered by the news media, are getting into the blame game, thus stoking the fires of litigation for everything.
    Presumption: well of COURSE, this woman would have done better given tPA. So let’s go after anyone who might have kept her from getting it.

    As a neurologist, and someone who sees that there is some value of tPA when used appropriately, I don’t see fear of litigation or being chastised in the WSJ as a useful way to educate people or change their way of managing illnesses.

  20. Belinda says:

    This is almost funny…..

    1) I never said “neurologists are misogynists” or anything close. It’s pretty well known even to non-physician doofuses like me that an individual who presents with certain sets of symptoms TENDS to be looked at less closely if they are young, slender, and/or female.

    2) I said without irony that I hope I AM the ignoramus I was called “with respect” by A MD because I would love to believe that my experience is not representative. And, guess what? I KNOW it’s limited, entirely anecdotal, and unsupported by empirical evidence or peer-reviewed studies.

    Phooey – I posted on the initial TPA thread just because I was interested to see the down-side info, and am trying to be informed on this issue for reasons involving my husband’s history and my own. It’s funny that those of you who are physicians can lament the fact that certain of your brethren are actually uninformed enough to send people to the ER “to get that clot-buster drug” – but that when a non-physician hazards an unflattering comment, all hell breaks loose.

    I have learned some interesting things in the substance of these entries, and – sorry – I’ve gotten some more anecdotal evidence concerning your profession in general…don’t worry – i won’t post anymore. Thanks for the information despite the other.

  21. “It’s pretty well known even to non-physician doofuses like me that an individual who presents with certain sets of symptoms TENDS to be looked at less closely if they are young, slender, and/or female.”

    In this crazy world we live in, so driven by litigation, it’s the young ones that we fear (and test) the most! When a ” young, slender, and/or female” pt. comes to an ED, and has a bad outcome, someone’s going to pay. It’s the missed diagnosis of stroke in a 40 year old that will lead to a huge settlement, where our careers can be ended! So we CT scan every vague complaint (ie “dizziness” could be a cerebellar stroke) in every young person, “just in case”, to the point that our healthcare system is the most expensive on the planet. Elderly patients have the greatest likelihood of disease, they are much more likely to die if we miss anything. But we order thousands of CT scans on young people, irradiate thousands of young bodies, so we won’t miss one CVA, one pulmonary embolism, one appendicitis. Each time we know the likelihood of disease is much less than 1 percent. I don’t make the rules in this wacky health-care system, I just enforce them.

  22. Goatwhacker:

    Just wanted to say that your impulse to question whether the improvement seen in the tPA group in NINDS was due to resolving TIAs is actually a very good one. It?s a question we should all ask whenever we evaluate stroke trials. However, in the NINDS trial this clearly was not the case, because if you compare the control group and the tPA group, you see significant differences in outcomes. If tPA was truly ineffective (ie. no better than placebo), and all improvemt seen in patients was actually due to resolving TIAs, then you?d expect to see the same thing in the control group (as long as the trial was properly randomized and had a good sample size). This didn?t happen in the NINDS trial, and it would have been unlikely for it to have occurred by chance.

    DS: I agree.

    GD raises a couple of issues. 1. Intra-arterial TPA as a first line thrombolytic strategy 2. why should having neurologic backup make a difference when the decision to give tPA essentially rests on being able to do a good history and exam, and take a CT scan?

    To take the first issue first. It certainly makes intuitive sense that IA tPA would be better than IV. The mechanism that we believe tPA uses is basically binding to fibrin in a thrombus and converting the entrapped plasminogen to plasmin, causing fibrinolysis. It is postulated that most large vessel disease is due to one or two thrombi blocking flow to a major vessel, and it makes sense to want to limit the drug locally rather than introduce it systemically, since its so dangerous. However, we have no Grade 1 evidence for this approach yet. PROACT looked at pro-urokinase, and provided encouraging data for IA urokinase up to 6 hours post stroke. The IMS group has provided observational data with historical controls that IV+IA tpa is better than IV tpa alone. However, there are a LOT of small IA studies out there showing no significant difference or even disastrous outcomes. Until we have Grade I evidence on the role of IA tPA in the management of patients with stroke, it is not wise from a public health perspective to send patients for (or undergo oneself) IA tPA, outside a research program specifically studying it.

    The second question is more interesting. Why indeed should there be a difference? In fact, why does there seem to be a difference in so many areas in medicine?

    At the famous Shouldice hospital in Toronto, the surgeons perform a very simple class of surgery, the one that all general surgeons begin their training with – hernia surgery. Hernia surgery is usually remarkably straightforward and generally safe. But for all its simplicity, it?s also something of an enigma as it is difficult to do without getting complications in at least some of your patients. At the Shouldice, for 60 years, they have done these surgeries better than anyone else in the whole world. The lowest complication rate, the best results. They are better than all the eminent surgeons at Harvard, at Hopkins, at the Mayo, at Stanford, Columbia, Wash U St Louis, Cornell? you name it. If you need hernia surgery, Shouldice is simply the best place on planet earth to get it. Period.

    But why? The operation is simple enough. The techniques that the Shouldice surgeons use are the same ones used by surgeons elsewhere. The instruments are the same, the materials are the same. So why the difference?

    Similarly, look at stroke units. (I?m talking here about stroke units per se, not with any reference to tPA). Study after study has found (Grade I level evidence) that patients who are admitted to specialized stroke units (whether or not they got tPA) do better than patients treated under other physical arrangements. If you admit the patient to a general medical ward for their duration of stay in hospital, they do not do as well. Even if you admit them to a general medical ward but consult neurology to help with management, their mortality and morbidity is higher. Why? Shouldn?t it be a matter of simply being able to take good histories, do good examinations, being familiar with the literature, and prescribing the right drugs? Stroke units don?t have any drug that a normal ward doesn?t. So why do they outperform general internists-staffed wards every single time?

    Most people who study complex systems think it’s most likely due to specialization (ie. knowing how to do a few things really, really well). Taking one or two well-defined processes and perfecting them beyond everyone else, so that error rates fall, and complications are not only dealt with earlier and more effectively, they are often prevented from happening in the first place.

    Now, with respect to emergency stroke management, it is apparent that the process of identifying, diagnosing and treating a stroke patient is complex problem. The differential diagnosis of even a hemiparetic stroke is substantial. The neurologic exam is easily the most complex part of the clinical exam repertoire. The decision pathway of prescribing tPA adds further complexity. Although on the face of it the multiple tasks involved look simple when viewed individually, every multifaceted process becomes more complex when it has to be carried out flawlessly all at once. Time constraints add to the difficult equation.

    The reality is that tPA is a rather dangerous drug, and can only be prescribed within very narrow limits. There are 15 exclusion criteria to begin with. The NIHSS has to be done in its entirety. The differential diagnosis is crucial ? this is not a drug that you want to give someone with a Todd?s paralysis or a complex migraine, items on the differential which are easily confused by many for a stroke. The CT has to be read, and hemorrhage excluded. (Do you know what the sensitivity of excluding hemorrhage on brain CT is for EMdocs, general radiologists, and neurologists, using a neuroradiologist?s read as the gold-standard? There have been a few small studies, with sobering results).

    In each of these processes, there are numerous places where an error can be made. In most of the post NINDS phase IV studies where there have been excess bleeds (and mortality), the protocol violations were very considerable.

    It is for these reasons that I think it is more likely that outcomes will be better at places where there is a specialist stroke team in place to handle CV emergencies.

    Now, be that as it may, does that mean all patients must be taken to stroke centers? This is more difficult to answer. In the first place, it is currently impossible, because of both manpower and institutional shortages. In the second place, restructuring acute stroke care delivery will be exceptionally expensive and complicated ? do we really want to do it based on an evolving technology? What if someone discovers a safer, simpler thrombolytic that can be given within the first 9 hours, instead of 3? At MGH, IA tPA is already given for up to 12 hours post (with posterior circ up to 24 HOURS post). It may soon turn out to be entirely unecessary to evacuate patients to the nearest university center by chopper.

  23. DensityDuck says:

    Or we could do something to directly oxygenate the brain tissue, but that’s just crazy talk.

  24. Bad Shift says:

    Meanwhile, a billboard recently placed at a bus stop outside my ER reads, “Time lost is brain lost,” followed by a suggestion that any person with tingling, numbess, weakness, or even a headache is eligible to receive a clot-busting wonder-drug.

  25. A MD has been very thoughtful in his comments about TPA for ischemic stroke, and has made several salient points. I’d like to clarify something I said earlier, and explain why I still am leery of TPA.

    First, I said I didn’t understand why a stroke center was needed to give TPA, or something similarly simplistic, pointing out that what is needed is a good history and PE, and the requisite CT and labs. I stand by that, but don’t want to give short shrift to the amount of time it takes to do all the history (that Stroke Scale seems like a terrific research tool, but is very very time-consuming to get through). If you have a neurologist who can help, terrific! We can’t get ours, at the ‘stroke center’, to come in. (My anecdote: the one time I’ve given it, I called the neurologist, laid it all out, and said I thought TPA was indicated. Response? “Let me know how it goes”.) There are a LOT of hospitals (not stroke centers) with good neurologists who could come in, do the initial eval, and give the TPA if indicated. THEN, if necessary transfer the patient to the ‘stroke center’. Perhaps on the coasts with high population densities and zillions of hospitals having a few stroke centers is the way to go, but even here in the DFW Metroplex the drive from the fringes of town to the ‘stroke center’ will chew up an hour, driving past several hospitals with EM docs, CT scans read by rads in real-time, and they all have TPA. The emphasis on Stroke Centers doing the initial eval is IMHO, detrimental.

    An aside. Somehow, it’s gotten into the lore around here that you need to have a neurosurgeon at your hospital to give TPA, ‘in case they bleed’, which is the single dumbest thing I’ve heard about TPA for stroke. If TPA is given and the catastrophe occurs, there will be several hours of fixing the coagulopathy before a surgeon will even touch the patient with something sharper than their wit, even supposing it?s a bleed they can do something about (subdural/epidural vs. intraparynchemal). I mention this to docs who want to transfer their 2 hour old strokes to us ?for possible TPA?, but it falls on deaf ears.

    Second, let?s get into where I, and most EM docs, decide against this treatment: commission vs. omission as a cause of patient injury. I?ll take the numbers cited above by A MD as a given (for purposes of this discussion). There is a difference between deaths that happen when I order and give a drug I know can cause intracranial bleeding, and deaths that happen when I don?t. It?s not the same to me that people die of stroke, and that I gave a medication that caused them to die.

    Nowhere else in EM is there such a horrible risk/reward with any medication, and that?s the crux of my problem with TPA for ischemic stroke.

  26. One of the things I have learned that I try to share with other physicians is that we need to share the risk, share the decision-making with the patients, and typically with tPA, with the patient’s family. They NEED to help with the decisions. If they cannot help with the decision, then the risk should not be taken.

    What this means of course is that you have to be able to explain the pros and cons in nonmedical, translated terms, which means that typically you need to know things really well to be able to do that. It is hard work, but doable because strokes are coming to the ER in a pretty steady stream, so you can get a lot of practice.

    I think it’s important not to be an advocate for tPA or be opposed to tPA by definition. You have to think about what you would want if it was your mother, your spouse, or even yourself having that stroke. If the timing is appropriate, the type of stroke is appropriate, you must be able to offer the choice.

  27. Greg,
    We’ll have to agree to disagree. I have had the chat with BOTH my patients who were candidates, and explained it all. I didn’t bias it, and we made a gropu, informed decision.

    And, when it’s my parent, I know. No TPA.

  28. Gruntdoc: Good post. What you say about why EM docs do not like TPA is very true, and I commented on that sentiment in my first post. I remember being a bit puzzled about it very early on, because their objections seemed rather more earnest than you’d expect from a disagreement about interpreting the data.

    But if you go down to the ER and stand in an EM doc’s shoes, you begin to see things from their perspective very quickly indeed. It can be very difficult to prescribe a drug which you usually see either doing nothing or else killing your patient, even if you intellectually understand that the overall mortality is unchanged etc, etc. The argument that there is a difference between giving a drug that directly kills (which, make no mistake folks, is exactly what TPA can potentially do) as opposed to letting a patient die of his stroke, also has something going for it. However, with this particular argument, one has to thread lightly – for aren’t most of our drugs that way? Aspirin can directly kill some people too. The difference with tPA of course lies both in the numbers (the frequency with which this happens), and in the dramatic way in which it happens – but no one should mistake that tPA alone behaves like a double-edged sword. Far from it.

    I am a strong advocate of letting docs on the front lines – whoever they are, EMdocs or neurointensivists or stroke specialists – decide on their own whether they are going to give tPA in a given patient. No lawsuit that you read about in the papers annoys me as much as some of the tPA related ones. It takes a great deal of hubris to argue that papa would have lived “if only Dr. EM had given him tPA”. Now, I do think that each stroke patient deserves consideration (of whether tPA should be given), ie. I have a problem, based on the data, with a prospective decision to not ever consider the use of tPA in any patient, but once a decision is reached, second guessing the guy in the trenches is very poor form.

    Lastly, about the numbers used in the first post. These are the numbers found in the original NINDS paper for the Barthel index. As most of you know, the NINDS trial used four different methods to measure outcomes – I like to use the Barthel figures because they are easier for most folks to understand (as compared with, say, talking to them about the NIHSS scores).

    However, do note that if you employ the summary estimates from meta-analyses (the most important is Wardlaw’s Cochrane review), your figures are going to be lower (because the other thrombolytic trials did not have outcomes as good as NINDS). I have no argument with using the NINDS data as the basis for my decisions as opposed to summary data from meta-analyses (which is what I think should usually be done), because there is clear heterogeneity between the trials – the other trials used stroptokinase, or enrolled up to 6 hours post, for example. The validity of the NINDS result was, IMO, affirmed by the Ingall/O’Fallon reanalysis.

  29. PS. GD, I may be missing something, but it seems to me that you and Greg don’t have to agree to disagree, because you both agree. You’re both saying candidate patients should be encouraged to make an informed decision. (Except when it comes to parents, LOL).

  30. A MD points out, correctly, that all drugs have downsides. Granted. Aspirin can cause GI bleeds, can contribute to excessive and sometimes life-threatening bleeding.

    However, it’s one thing to prescribe a cheap drug with known but acceptable side effects (rare is the ASA induced lethal GI bleed) but it’s entirely another to push a drug with a 6-9% (at best) complication rate.

    I suspect we could go over this forever, and agree on separate points but not agree it’s a terrific drug. I am entirely OK with that.

    An aspect of stroke treatment I’d like to engage you in, though, is the ‘Stroke Center’, and whatever to do where a lot of America lives not near a ‘Stroke center’, in the Hinterland.

    What should the EM doc, alone in an already busy ED, do with a ‘Classic Witnessed Stroke’ (no complications, let’s make this fairly straightforward) who arrives in the ED 45 minutes after onset? A CT can be obtained within 30 minutes (after blood drawn) and is read by teleradiology within 10 minutes of completion with a faxed report within 15 minutes. (No info about the rads quals, just like most docs practice). It’s two hours to get the patient transferred to a ‘big center’ by helo, minimum. No neurologist on staff, but several IM docs who are helpful, and you could get one to come in if called to help.

    So, what to do? Forgo the TPA because transfer takes too long? Go a it John Wayne and do your own exam, history, and give TPA while readying the transfer? Explain to the family that there’s a drug they’ll sue the ED doc for not giving, but as you cannot get to a Stroke Center in time, so they aren’t eligible?

    This is where many ED docs live and work, and their patients arrive. What to do?

  31. I wrote these posts (here and here) before I saw this post by GruntDoc. How well they match up. Great job GD.

  32. Goat Whacker says:

    I did a little poking around – it looks like this is still the current position statement from the AAEM:

    It is the position of the American Academy of Emergency Medicine that objective evidence regarding the efficacy, safety, and applicability of tPA for acute ischemic stroke is insufficient to warrant its classification as standard of care. Until additional evidence clarifies such controversies, physicians are advised to use their discretion when considering its use. Given the cited absence of definitive evidence, AAEM believes it is inappropriate to claim that either use or non-use of intravenous thrombolytic therapy constitutes a standard of care issue in the treatment of stroke.

    An interesting point/counterpoint on the issue:

    and another, this looks a bit older:

    Feel free to clean up my links, I’m not good at them

  33. LOL. You sure know how to ask em, GD. Here’s what I think:

    We’ll presume, in addition to the details in your question, the ED doc is the “average” ed doc in a community hospital. Family med residency, seen quite a few strokes, not so many acute strokes, saw tPA given a few times, never given it himself. We’ll assume the hospital has no written institutional policy on tPA and stroke. He has not had much of a relationship with the stroke guys at the big academic place, and is not precisely certain what they offer etc. I think this is the scenario you’re looking for, right?

    I would say there is very little he can do, where tPA is concerned. There is now very considerable evidence that giving tPA outside of a formal stroke program (that is able to closely approximate the clinical trial milieu of NINDS) can result in disastrous results. That’s the reason why organizations like your ACEP etc have come out strongly saying EM docs should not be faulted for not giving tPA if they are not in an appropritate setting.

    There is insufficient evidence at this time to endorse the use of intravenous tPA in clinical practice when systems are not in place to ensure that the inclusion/exclusion criteria established by the NINDS guidelines for tPA use in acute stroke are followed. Therefore, the decision for an ED to use intravenous tPA for acute stroke should begin at the institutional level with commitments from hospital administration, the ED, neurology, neurosurgery, radiology, and laboratory services to ensure that the systems necessary for the safe use of fibrinolytic agents are in place.

    That’s the sound of your behinds being covered, guys.

    Still, this still leaves that nice ol man lying back in Rm 10. What will I do if I were the above doc?

    Well, I’ll first do what all EM docs do better than anyone else. Assess the patient for any emergency and make sure he’s stable. I’ll get a good HPI. I’ll do a solid exam, ie. the usual + making sure that he does seem to have a pronounced neurologic deficit that is referable to a vascular territory that makes sense, and is of a general size/severity as to warrant consideration of lytics. I’ll then head back to the nurses desk, and pull out one of those tPA list type things and tick off as many as I can. We’ll assume no contraindications on the list (if there are your case ends here).

    I don’t really do the NIHSS, so I skip that bit, and pick up the phone to call Mr. Stroke Guy at Big Hospital. I’ll say “I have sitting here Mr. Nice Ol Man, who I think has a stroke. His story is ______. His main findings on exam are _____. Labs _______. I didn’t do the NIHSS, but he does have a big stroke. Our hospital doesn’t have a policy of offering tPA for stroke. I’m just wondering what you guys might be able to offer? The timing of patints arrival is______, soonest we can get air transport is _____, etc. Do you guys do any 6 hour IA stuff, etc? I just want to know the options before I speak to the fmly.”

    Head back to the Missus. Give her the basic stroke talk. Then talk to her about this “clot busting drug you may have heard about”. This part is very important, so I explain very carefully what the data show about its risk:beneit ratio, why its important that this drug not be prescribed simply to anyone who walks in etc. I tell her what options are available, now that there is approx 1:45 minutes left on the clock.

    She will ask why cant I “just give the tPA here”. I’ll explain that this drug is an uncommon, complicated treatment. The company ads on tv tend to stress only one side, but every single one of the studies have shown that there is a bad side in increased brain bleed risk. Becuase the balance is quite fine, it is best for the drug only to be administered by a special team who have experience in optimising the treatment. We do not have such facilities in this hospital to determine if your husband really can “safely” take this drug,, and we most certainly do not want to harm him.

    So lets say desoite everything one month later you get a letter from Mrs. Nice Ol Man’s Not so Nice Lawyer. You are being sued for not giving tPA.

    1. Pull out all the Emerg society guidlines. There are at least two American, one Canadian.
    2. Pull out all the comm studies showing tPA disasters where there were protocol violations.
    3. Pull out your nice chart notes, where you documented absolutely everything.
    4. Call Jerry Hoffman. He can make tPA look like cyanide.
    5. In addition to the above, at the deposition you must also make clear that tPA simply cannot be given safely in the face of protocol violations, as the harms very easily overetake the risks. It is difficult for a smaller hospital with no BC Neurologist to do this properly; in fact, unless the ED staff know how to administer the NIHSS it cannot be done, for it may not be easy to tell if the score is >22. No one is going to testify that doing the NIHSS is the standard of care for a EMdoc out in the sticks.
    6. After the nightmare is over, call together all your ED Docs, hospital risk management etc for a sit down, and get them to craft an institutional policy on tPA for stroke.

  34. So here’s the question. The “thatched hut” community hospital I’ve worked in developed an “institutional policy on tPA for stroke”. Thus far, almost every patient that would have qualified since the “institutional policy on tPA for stroke” was instituted has sat in the waiting room so long they no longer qualified. So many Stroke Victims have vague symptoms, and you can only “bring right back” so many chest pains, R/O CVA’s and suicidal patients. Welcome to 20th century overwhelmed Emergency Medicine. #1: How do you explain to a family that Grandpa doesn’t qualify for TPA cause he was sitting in the WR watching “Desperate Housewives” while time went by, and #2. Who pays? The Hospital or the ED Doc? If its a double coverage ED, do both docs get screwed or just the one that was stupid enough to pick up the chart?

  35. Goat Whacker says:

    ds –
    Having worked in thatched hut ER’s myself yours sounds a little worrisome. Most patients with mild or ambiguous enough symptoms to warrant sitting in the waiting room wouldn’t qualify for TPA anyway. If someone with CVA symptoms is sitting watching Desperate Housewives you’ve got a triage problem, or if you are too busy to see such patients promptly you have a staffing problem.

    As far as “who pays” if sued, I would guess the lawyers would go after the hospital plus any doctors named on the chart.

  36. brandt delhamer says:

    The comments offered here in response to the WSJ piece have demonstrated in spades what we should all already appreciate as EPs-that tPA is an arrow in the EM quiver, to be used judiciously and with serious consideration of the risk/benefit ratio before committing your patient and yourself to this potentially disasterous therapy.

    Having trained with Dr Grotta I can assure you that his appreciation for the role of EPs in treating stroke is unquestioned and I suspect that his comments to the WSJ author, when taken in full instead of in part, would reflect this attitude. Speaking for my training program specifically we were accustomed to seeing at least one to two stroke patients nearly every day, often outside the window and ineligible for tPA anyway. Many patients did receive the lytic, however, and I have personally seen all of the described outcomes, from complete resolution to partial improvement to no improvement and needing intra-arterial tPA to intracranial bleed causing death. However, this may not be the case at every academic training program, let alone the average community ED.

    Despite the somewhat abrasive tenor of the article we would all have to admit that, thanks to gruntdoc, we and those whom we have communicated with about the article are all more aware of the way the lay media present the issues; which is to say fairly biased-not that anyone is surprised by this.

    My 2 cents.