Getting the Gist Across Is Enough for Informed Consent for Acute Stroke Thrombolytics

Acute stroke thrombolytics greatly reduce post-stroke disability, both in clinical trials and in clinical practice.^1-3 Too few patients, however, take advantage of these life-improving and cost-saving treatments.^{4, 5} In fact, in the United States up to 7.5% of tissue plasminogen activator (tPA) eligible patients refuse this time limited stroke treatment,^{6, 7} and even among stroke patients who receive tPA, up to 20% experience delayed treatment due to patient and family consent.⁸ Currently in the United States, a recommended approach to tPA consent is shared decision making, which involves deliberative consideration of options, risks, benefits, and patient values. While shared decision making is highly valuable and appropriate in many medical decision making situations, we believe that full shared decision making is not well suited for the acute stroke scenario due to the lack of clinical equipoise regarding tPA administration, the time sensitive nature of this acute stroke treatment, and an inappropriate setting for in depth-discussions.

tPA is approved by the United States Food and Drug Administration and the European Union, albeit with more restrictive eligibility criteria than the current guideline recommendations.⁹ Current United States professional societies’ recommendations on tPA administration vary in strength by specialty. For example, in the United States, Neurology and Neurosurgical societies strongly recommend tPA treatment, while Emergency Medicine societies are more tepid in their recommendations. The American Heart Association, with endorsements from American Association of Neurological Surgeons and Congress of Neurological Surgeons, gives tPA administration in the 3 hour window its strongest recommendation.¹⁰ However, in 2015 the American College of Emergency Physicians’ (ACEP) recommendation to administer tPA in the 3 hour window was downgraded to a level B recommendation^{11, 12} and a consent processes grounded in shared decision was recommended.¹² Similarly in their recent focused update, the American Academy of Emergency Medicine (AAEM) recommends discussions regarding benefit and harm and advocates use of a decision aid, iScore (http://www.sorcan.ca/iscore/) to facilitate the tPA consent discussion.^{13, 14} We also note that reconciliation of recommendations across specialties might alone facilitate greater tPA administration by itself by reducing uncertainty.

The incongruence of acute stroke consent and shared decision making

Both shared decision making and decision aids are intended for use when there is clinical equipoise in the treatment decision on the part of the provider and the patient.¹⁵ Shared decision making involves reaching a treatment decision through careful consideration of the patient’s values and the impacts that each treatment option could have on their life in the short and long term. Shared decision making is well-suited to situations where there is no universal ‘right answer’ but a ‘right answer’ for an individual patient. Shared decision is often used in situations such as breast cancer, localized prostate cancer, knee arthritis, and cataracts, where patient preference plays a large role in treatment decisions.¹⁶ By contrast, shared decision making is not appropriate in the absence of clinical equipoise. Thus, given the support of tPA use by all medical societies, though in varying strengths, there is no clinical equipoise in the treatment of tPA to eligible patients.

In addition to the absence of clinical equipoise, the time sensitive nature of tPA and the emergency room setting contribute to our critique of shared decision making and decision aids in acute stroke treatment. When it comes to acute stroke, ‘time is brain’ and a prolonged decision-making process delays treatment, decreasing the opportunity for recovery.¹⁷ Given the extreme time constraints of acute stroke treatment, tPA discussions with patients and families are often short, an average of 2.7 minutes by some estimates, leaving little time for a full discussion of patient values and preferences.¹⁸ Second, these discussions most often occur in the busy resuscitation bays of an Emergency department (ED). This setting is suboptimal for the in-depth, deliberative exploration of a patients’ values central to shared decision making due to the fast-paced and chaotic nature of the ED.

Gist communication

The tPA treatment decision is unusual in that one number encompasses both the major benefit and risk: the 30% relative reduction in post-stroke disability includes the ischemic stroke benefit and the risk of intracerebral hemorrhage and death.¹⁹ Current acute stroke thrombolysis decision aids, including the iScore endorsed by the AAEM, typically provide precise numerical estimates of the risks and benefits of tPA.^{14, 20, 21} In addition, some decisions aids present the risk of disability, intracerebral hemorrhage, and death as individual values.^{14, 20} Listing these risks as separate numbers may lead patients to overestimate the risk of receiving tPA treatment. An important nuance here is that major disability is likely for patients with a large stroke, but the potential of shifting outcomes toward a worse state is very unlikely,²² and many aids may overstate this risk.²⁰

The above concerns about shared decision making and decision aids for tPA decisions are evident in practice. As part of an implementation science research study, we conducted 15 semi-structured interviews with acute stroke providers, including medical technicians, nurses, and physicians in a safety net ED to explore barriers and facilitators to tPA treatment, guided by the Tailored Implementation of Chronic Disease framework.²³ We observed shared decision making taking place. One ED physician noted, “I gave him [tPA eligible acute stroke patient] my whole spiel about it, that I informed by the evidence and my experience, to try to get some buy-in. This is increasingly where we’ve gone with a lot of, not just controversial or debatable areas, but with a lot of emergency medicine is to try to do shared-decision making.” We also found that providers’ emphasis on the risks of tPA were a barrier to treatment. One ED nurse noted, “I think we give too much warning with tPA…I think we tell patients about all the bleeding and that they may get worse or die and then patients get scared. I mean, why do we focus so much on the risks. This is nothing like MI [Myocardial Infarction] when we just take them to the cath lab. Or when we stop and restart someone’s heart. We do not tell the patients all this warning, we just do it. There is no decision. But for stroke we talk too much about all the warnings which scares the patient. We should not give them so many choices. The stroke will be with them for the rest of their life.” Other studies corroborate provider emphasis on the risks of tPA, particularly intracerebral hemorrhage, as a barrier to treatment.^{24, 25}

In light of the lack of clinical equipoise and the unsuitable clinical scenario and setting for shared decision making and decision aids, we believe that communicating the gist of tPA treatment is warranted. Gist discussions focus on qualitative description of the potential benefits, which take into account the risks without describing specific numerical values. There is a fallacy that more information is better, but increasingly recognized is the idea that more information does not always equate to a better understanding of the most decision-relevant information.²⁶ Gist communication also supports autonomy without a time-consuming values discussion. Autonomy requires understanding of what the true choice is, thereby enabling patients to make choices that increase the likelihood of outcomes they would define as positive. In the acute stroke scenario, the most decision-relevant information is the net benefit (which, as noted, incorporates information about risk). Given the time pressure of the acute stroke scenario, patients and families need to understand at a minimum a) that risk exists regardless of their choice (so they are not actually avoiding risk by not agreeing to treatment), b) that the likelihood of positive outcome is significantly higher (but not assured) if tPA is administered, and c) that the net benefit depends on acting as fast as possible. Our proposed process, which is congruent with other gist-based risk communication strategies, focuses on ensuring that every patient truly understands these fundamentals.^27,28

The following summary, developed in collaboration with patients in a safety-net ED and stroke providers, captures the relevant gist: “Getting tPA after a stroke reduces your risk of disability. People who get tPA to treat their stroke have a better chance of recovering without disability and getting back to the activities they love compared to people who do not receive the treatment. All medicines have some risk. With tPA, there is a risk of serious bleeding. However, time is important as well. We have found the faster tPA is administered, the greater the chance that patients will have the best possible outcome. Do you have any questions?”

Given the clinical efficacy of tPA, the time sensitivity and the setting, the acute stroke clinical situation is not optimal for shared decision making and decision aids. Communicating only the salient benefits and risks of tPA to patient and families may decrease tPA refusals and improve treatment times, an important predictor of outcomes.