Last week marked the launch of the What Do You Say? case series, and what a great first week it was. This month’s case focused on patient education regarding tPA use in ischemic stroke. We explored how to navigate the evidence with patients, how to approach the situation with surrogate decision makers, and how to deal with possible conflict from consultants.  There was an excellent amount of commentary from numerous, diverse perspectives in stroke care, running the gamut of learner levels and practice settings.

This week’s wrap-up includes two expert responses to the case, as well as a summary of the community discussion from the blog comments and twitter.

Responses to Case 1.01

This case's expert responses were provided by:

Expert 1

Anand Swaminathan, MD, MPH

EM Assistant Residency Director at NYU/Bellevue Hospital Center

Expert 2

Ryan Radecki, MD

Assistant Professor of Emergency Medicine at The University of Texas Health Science Center at Houston

What We Know About Alteplase

All medical interventions, whether they are medications or procedures, have risks and benefits. Thus, it is vital that we are adept at communicating those risks and benefits to our patients. This is an integral part of our work. The less well we understand the risks and benefits of a certain intervention, the more difficult this task becomes. The use of alteplase (tPA) in ischemic stroke is a good example of such an intervention. Since the NINDS trial in 1995, the data on this topic has been consistent only in its inconsistency. A full discussion of the risks and benefits is beyond this post but check out my pro-con debate with Andy Jagoda on the EMCrit podcast back in January of 2014 and the EM Lyceum blog on the topic from August 2013.

While some data demonstrates that alteplase leads to better outcomes in patients with ischemic stroke, the exact group that is most likely to benefit is unclear and the stakes are high. The downside of alteplase? Increased intracranial hemorrhage and increased early mortality. It’s vital that we discuss both risks and benefits of this drug as best we can. There are a number of other aspects that are critical to communicate. The patient may improve without medications, even though it looks like an ischemic stroke, we are often fooled by stroke mimics in whom there is no benefit to alteplase. Additionally, there are other risks with alteplase aside from intracranial hemorrhage, including anaphylaxis.

Working with the Consultant

What do you do when your assessment of the risks and benefits is different from that of the consulting neuro team? This is a tricky situation. Your hospital is very likely to have a policy in place on the management of CVA and that policy likely involves the use of alteplase. Often, though, the neurologist consulting for your patient may be more junior than you, they may have seen less strokes than you, and they may not have seen the consequences of alteplase related intracranial hemorrhage. The first thing to remember is that it is not about you and the neurologist; it’s about the patient. Even if the consultant is taking on care, you have a duty to the patient as long as they are in your department. As an advocate for the patient, I make sure that my dissenting opinion is heard prior to the discussion with the patient.

I make a point to be in the room with the patient and the neurologist so that I can hear the conversation from the consultant on risks and benefits. Basically, my job is to keep them honest. Make sure the absolute and relative contraindications have been appropriately reviewed and make sure the patient is told about the potential risks as much as the potential benefits. If the neurologist falls short in their informed consent process, it’s our job to step in and fill in the gaps.

Communicating with Patients and Families

Ideally, the conversation and informed consent process (which is not standard care for alteplase therapy, but ought to be) should be conducted with the patient and their designated healthcare proxy or close family/loved one. However, this is not always possible. In these situations, it is vital to gauge the patients’ ability to make an informed decision. Even patients suffering from devastating strokes may retain decision-making capacity. If the patient is able, they should be engaged in the decision. It’s trickier when the patient is unable to be part of the discussion. In these cases, you should act in the best interest of the patient and this may simply be asking yourself the question, “If I was in his/her shoes, would I want the Alteplase?”

Ultimately, the conversations surrounding the use of Alteplase are difficult because the risks and benefits aren’t clear. We as informed physicians have a difficult time interpreting the data and deciding what is best. Imagine how hard this would be for a patient or family of a patient suffering from an acute ischemic CVA.

What Do I Say?

How I phrase the conversation:

“Mrs. Desi, it appears that you have had a stroke. This is typically caused by a blood clot. One of the medications that we have to help treat this disease is a clot-busting drug called Alteplase. The best studies we have show that this drug does not save lives but does increase the chance that you will have a more complete recovery. However, it comes at a cost. Patients who receive the clot-busting drug are more likely to suffer serious bleeding around the brain and are more likely to die earlier after a stroke. There is also a chance that you will get better without receiving this medication. This is a difficult decision to make, and we do have some time constraints on when the medication can be given. We want to answer any questions you may have.”

Representing the Risks and Benefits

I represent the risks and benefits of tPA for ischemic stroke as being controversial and uncertain .

The risks, in general, are relatively consistently observed across trials.  Approximately 1 in 12 patients, on average, suffer neurologic worsening or die as a result of intracranial bleeding.  This risk can be further tailored slightly to the individual patient in an imprecise fashion, as risks for ICH increase in the elderly, strokes of greater severity, patients with hypertension, and patients with diabetes, among others.  Similarly, patients absent these features are at lower risk.  Mrs. Desi, with her age and presentation is very similar to the average patient across stroke trials, and she would be expected to have a risk of ICH.

The benefits are substantially more difficult to communicate.  In stroke trials, and specifically those utilizing tPA within 3 hours of symptom onset, there was approximately a 10% absolute increase in patients alive and independent at long-term follow-up.  The challenge, as has been discussed many times previously, involves the reliability of the data.  The biases described – the open-label nature of IST-3; trials stopped early due to harm or futility; the possible unblinding of investigators due to differences in appearance between placebo and active infusion, or by early minor/mucosal bleeding; and the pervasive involvement of the trial sponsors – all dramatically favor the tPA cohort.

Compounding these issues, despite the recognition of stroke as an extremely heterogenous disease, there are no validated tools available for rapid risk/benefit-stratification of patients.  As with ICH risk, a variety of factors are known to affect outcomes, including patient cerebrovascular substrate, stroke etiology, and stroke subtype.  Many patients will see no impact on recovery despite even the theoretical benefit of tPA, while others may manifest favorable vascular conditions for dramatic improvement following tPA.  But, unfortunately, precise estimates of chance or benefit cannot be individualized at this time.

Mrs. Desi has similar characteristics and presenting symptoms to the average patient in stroke trials.  As such, it would be reasonable to consent the patient by first cautioning her tPA is a controversial topic, our best evidence has troubling limitations, and describe this as an active area of research.  Then, I would describe general outcomes estimates based on the published data.

Dealing with Disagreement

Disagreements regarding the use of tPA should rarely be negotiated on an individual-patient basis.  When two departments are involved in the acute care of ischemic stroke patients, staff from each department should cooperatively meet and establish a general institutional practice.  This may involve offering tPA narrowly to those meeting only the NINDS enrollment definition, it may include a subset of the ECASS III enrollment definition, or it could involve a wider population based on institutional research initiatives.

Regardless of the eligible cohort, every effort should be made to minimize on-shift hostilities – and an even greater effort made to settle any disagreements away from the patient.  I recommend a pro-active approach, reviewing the CT and case with the Neurology team apart from the patient and family.  This is the preferred setting to raise any particular points of concern, before addressing the risks and benefits with the patient.  In this case, it appears the Neurology team has primary institutional responsibility for assessment of eligibility for tPA, including informed consent procedures.  I would only intervene in a setting in which I felt the consent was substantially deficient or the administration of tPA atypically dangerous.  Witnessed cases of sub-optimal minor details of care would lead to follow-up discussions regarding eligibility or adequacy of informed consent.

Communicating When Patients Can’t

My communication strategy changes very little when discussing therapeutic interventions with a surrogate family decision-maker.  Every effort is made to ensure the surrogate attempts to best represent the true wishes of the patient, and to reduce their own individual bias.  Many surrogate decision-makers benefit from the affected patient having previously clearly expressed their wishes regarding treatment in the face of near futility, while others struggle to identify such directives.

However, the use of surrogate decision makers for such decisions is typically associated with either severe pre-stroke disease or devastating neurologic insult.  These patients generally have poor cerebrovascular substrate, and are at higher-than-average risk of intracranial bleeding and death.  Likewise, outcomes are likely to be dismal, regardless of treatment status – in the range of 3 in 30 with treatment and 2 in 30 without.  Individual patient factors and relationships substantially affect the decisions of surrogates, but many opt for less-aggressive care.

What Do I Say?

In this instance, my consent would roughly take the following form:

“tPA is a clot-busting therapy approved by the FDA for treatment of stroke.  Its use is controversial, and researchers are still investigating the appropriate patients for its use.  The data regarding its use comes primarily from several small trials sponsored by drug companies, and a few of these trials were even stopped early because it was harmful or of no benefit.  That said, many physicians feel this is promising or proven therapy for stroke.  In the drug company-sponsored trials, about 3 in 10 patients like Mrs. Desi were alive and able to care for themselves after treatment with tPA, compared with about 2 in 10 for those who were not treated.  Unfortunately, in these same trials, approximately 1 in 12 patients like Mrs. Desi were made worse or died because of bleeding side effects from treatment.  Regardless of whether Mrs. Desi receives tPA, every stroke patient receives the best possible care and rehabilitation.”

This summary was produced as a qualitative, thematic review of the blog and twitter discussion created by the Ed in the ED community. All community commentary was aggregated, and the following common themes were identified in a majority of responses.

1. Preparation is key.

The community response was overwhelming clear: the bedside is not an appropriate place for a debate between services, so preparation before talking to the patient is essential. Insight on how best to prepare for this discussion came from two distinct levels:

Prepare at the organizational level:

If not already in practice at your institution, consider implementing interdepartmental meetings in which all specialties involved in the treatment of stroke can be present to discuss best practices. These meetings should not just outline treatment guidelines, but ought to include a discussion of the current evidence for tPA and how to communicate with patients about this intervention. Regular discussions can prevent possible communication errors, disagreements in clinical settings, and lead to more cohesive care for patients. Interdisciplinary simulation exercises can take this approach one step further, and can include nursing staff and pharmacists to fully simulate particularly difficult scenarios working as a team.

Prepare at the shift level:

Prior to the conversation, you should gather all of the personnel responsible for taking care of the patient in order to discuss the plan, review the information to be presented, and solve any disagreements. Ideally, this group of decision makers will head to the bedside to talk to the patient about tPA together so that everyone is informed of the discussion and has the opportunity to participate.  After the discussion, be sure to follow up with the patient and with healthcare providers who have spent further time with the patient (medical students, nursing staff, etc.) to ensure all of the patient’s concerns are being managed.

2. Don’t approach the conversation as obtaining “informed consent.”

In this situation, patients rely on your medical knowledge and experience for a recommendation just as much as information for making their own decision.  Expecting patients or families to fully understand the risks and benefits of tPA in order to make an informed, independent decision is unrealistic and potentially devastating. Community members stressed the need for a shared decision making approach, in which physicians help patients explore the reasons why he or she might choose tPA therapy or forgo it.  Such reasons include patient values, willingness to take risks, quality of life concerns, current function, future goals, risk factors, and desired outcomes. Help the patient to weigh these factors, and do not be afraid to inform the patient of your stance on the literature if asked.

3. Communicate risks and benefits in a clear and simple manner.

Start the discussion by assuming a low health literacy level and present any basic statistics, like the number needed to treat, in plain language. To explain the potential benefits and harms of tPA, some community members advocated for the use of a visual aid to represent outcomes. Another participant found success with a phrase such as “some patients get better, just as many stay the same, and some get worse.” Without a well-studied means of representing tPA risks to patients, it is important to use a method of communication that is both accurate, easy to understand, and comfortable for you.

4. Set realistic expectations.

“In a setting where families grasp at straws and hear what they want to hear, it’s important that someone says ‘death from brain bleeding’ out loud.” – Dr. Pik Mukherji

Throughout the discussion, you have a responsibility to moderate the expectations and hopes of the patient and family. It is crucial to explicitly outline what patients can expect to happen in all scenarios, including the establishment of a realistic time course for any return to function. It is especially important to emphasize the risk of intracranial hemorrhage—this needs to be explained concretely as a bleed in the brain that can cause devastating neurological damage and death. Also, the possibility of intubation, surgery, and life sustaining care needs to be explained to the patient so that a conversation about the patient’s wishes can take place before such a need for this level of care arises.

5. Emphasize that there is no “right” answer.

While the shared decision making process may have identified reasons why a patient might or might not want tPA therapy, it is still important to let patients and families know that this is an incredibly tough decision with no clear answer. Remember that your patient will have to live with this decision, not just the result of her care. As respondent Dr. Carey Cuprisin stated, your objective should be to “try to help the patient or family make a decision that they’ll be unlikely to regret even if the patient has a catastrophically bad outcome.”  Encouraging a positive attitude towards post-stroke care, regardless of a patient’s decision about tPA, can also help.

6. Ask surrogate decision makers to explore patient’s values.

When the patient does not have the capacity to make medical decisions, ask family to reflect on how the patient likes to live his life. How does the patient handle uncertainty—is he a risk taker or does he prefer to play it safe? How would he feel about living with a disability?  Try to get the surrogate decision maker to tap into the core qualities of the patient to aid in making this decision.  Additionally, you should seek out advanced directives and any other pertinent information from EMS, nursing homes, or other sources to ensure a complete picture of the patient’s wishes. Your approach to the conversation with surrogate decision makers should otherwise follow a similar course to a conversation directly with the patient.


Talking to patients about tPA is an incredibly difficult task: it is hard enough for physicians to make the decision to utilize Alteplase given the current evidence, let alone educate patients to participate in that decision. While there is no one-size-fits-all discussion to have with patients suffering an ischemic stroke, these important points will hopefully provide the basis for your approach to helping patients navigate this difficult time.

Thank you to all of the participants:

Matt Astin (mastinmd), Patrick Bafuma (@EMinFocus), Carey Cuprisin (@CareyCuprisin), Jeremey Faust (@jeremyfaust), Anton Helman (@EMCases), Nikita Joshi (@njoshi8), Michelle Lin (@M_Lin), Ken Milne (@TheSGEM), Pik Mukherji (@ercowboy), Eve Purdy (@purdy_eve), Sameed Shaikh (@SynthShaikh), Ryan Stanton (@EverydayMed), Seth Trueger (@MDaware), and Manrique Umana (@umanamd).

Check back soon for a downloadable PDF of the case and responses.