Rather than giving treatment with tPA the Level A recommendation in this go-around, there is a new Level A recommendation (requiring high clinical certainty). It concerns the greatest fears of treatment with tPA, the risk of intracerebral hemorrhage (ICH):

“The increased risk of symptomatic intracerebral hemorrhage (approximately 7 percent compared to a baseline of 1 percent) must be considered when deciding whether to administer IV tPA to acute ischemic stroke patients.”

Essentially, based on a systematic review of randomized trials and observational registry data, the only consistent finding suitable for a Level A recommendation was a recognition of the serious adverse effects of systemic thrombolysis.

Treatment with tPA within 3 hours now becomes a Level B recommendation:

“With a goal to improve functional outcomes, IV tPA may be given to carefully selected acute ischemic stroke patients within 3 hours after symptom onset at institutions where systems are in place to safely administer the medication.”

Along with the downgrade in strength of the recommendation, the language also addresses the systems necessary to administer tPA. One safety concern shared by many emergency physicians stems from the generalizability of trial and registry data collected at dedicated stroke centers staffed by stroke neurologists. Many practice settings do not have access to the same level of subspecialty, radiology, and nursing expertise as the centers conducting stroke trials, resulting in less-safe conditions for treatment.

Use of tPA in the 3–4.5-hour time frame remains a Level B recommendation:

“Despite the known risk of symptomatic intracerebral hemorrhage and the variability in the degree of benefit in functional outcomes, IV tPA may be given to carefully selected acute ischemic stroke patients within 3 to 4.5 hours after symptom onset at institutions where systems are in place to safely administer the medication.”

Finally, the clinical policy authors added a new Level C recommendation, based on expert consensus, suggesting patients must be included in the decision-making process:

“Shared decision-making between the patient (and/or their surrogate) and a member of the health care team must include a discussion of potential benefits and harms prior to the decision whether to administer IV tPA for acute ischemic stroke.”

While this ultimate recommendation seems simply like sound, ethical practice, there are several publications from the stroke neurology literature suggesting patients ought be given tPA under “implied consent” as the “standard of care.” This recommendation reinforces the obvious necessity of a full patient and surrogate involvement when administering a medication with potentially devastating consequences. In other words, informed consent should be obtained for tPA administration.

Guidelines ought to accurately reflect the strength of the evidence, not the collective wishes and hopes of a small cohort of experts. Happily, this new version makes profound strides in sticking to appropriate grading of the evidence.

It should be noted, however, the clinical policy summarized here is only a draft, open for feedback to all concerned parties, as part of the new writing process. In general, it is a laudable effort regardless of one’s personal stance regarding tPA in acute ischemic stroke. Guidelines ought to accurately reflect the strength of the evidence, not the collective wishes and hopes of a small cohort of experts. Happily, this new version makes profound strides in sticking to appropriate grading of the evidence. That said, there are a handful of aspects in which this policy could potentially be improved:

• The policy statement describes shared decision-making and cites two examples of information graphics potentially usable for illustration of the risks and benefits. However, it is very clear from stroke trials the risk-benefit ratio differs depending on many factors, including stroke severity, specific stroke syndromes, and individual patient substrate. Several models have attempted to individualize the risk of symptomatic ICH compared to baseline with uncontrolled diabetes, uncontrolled hypertension, and age the most important predictors. The policy statement alludes to a need for further research necessary to tailor treatment to the individual patient but understates this critical need. Considering it has been 20 years since the original NINDS trial, still having inadequate evidence with which to guide decision-making is nonsensical. This document could be a powerful platform with which to state clinical equipoise and call for additional placebo-controlled trials.
• The Level B recommendation for the 3–4.5-hour time window is difficult to justify based on the stated recommendation criteria. The authors rate the one positive study, ECASS III, as Class II evidence based on potential for bias. Class III data from ATLANTIS and IST-3, however, provide negative data. ATLANTIS was modified several times (including due to safety monitoring) before ultimately settling on a 3–5-hour time window and was then stopped early for futility. IST-3 suffered from an open-label design but was particularly unfavorable within the 3–4.5-hour window, with 31.5 percent having good outcomes given tPA compared with 37.7 percent in the control group. The individual-patient meta-analysis further cited in support of the 3–4.5-hour window derives most of its patients from these three trials, providing limited additive information. This probably does not meet their stated criteria for a recommendation with “moderate clinical certainty.”
• Several statements use vague terminology to describe the specifics of treatment. Each recommendation for tPA use mentions “carefully selected patients” and “systems in place to safely administer the medication.” If the recommendations propose strict adherence to ECASS III or ECASS III enrollment criteria, this should be clearly stated. Likewise, if “systems in place” refers to a certification such as The Joint Commission Advanced Comprehensive Stroke Center, this should also be clarified.
• Ultimately, even though the document explicitly states it is not intended to represent a legal standard of care for emergency physicians, it will certainly be wielded as such to either protect or crucify. To that end, it could use language specifically protecting the clinician who chooses not to offer tPA, in recognition of the persistent uncertainty responsible for the downgrading of recommendations from Level A.

But these are only my initial reactions, subject to the limitations of my own knowledge base and biases. Luckily, this document was open to feedback from all emergency physicians. The comment period closed March 13, 2015. The final policy should be forthcoming.