Journal Jam 10 Thrombolysis & Endovascular Therapy for Stroke Part 1

In this 2 part EM Cases Journal Jam podcast Justin Morgenstern, Rory Spiegel and Anton Helman do a deep dive into the world's literature on systemic thrombolysis for ischemic stroke followed by an analysis of endovascular therapy for stroke. We elucidate the important issues related to p-values, ordinal analysis, fragility index, heterogeneity of studies, stopping trials early and conflicts of interest related to this body of evidence. While "calling a code stroke" is now considered standard for most stroke patients and tPA for stroke is considered a class 1A drug, a close look at the literature tells us that the evidence is not as strong as our stroke protocols suggest... Podcast produced by Anton Helman, Justin Morgenstern and Rory Spiegel; sound design and editing by Anton Helman; EBM bomb by Anton Nikoline. Written Summary and blog post by Anton Helman July, 2016 Cite this podcast as: Morgenstern, J, Spiegel, R, Helman, A. Thrombolysis & Endovascular Therapy for Stroke Part 1. Emergency Medicine Cases. July, 2016. https://emergencymedicinecases.com/thrombolysis-endovascular-therapy-for-stroke/. Accessed [date]. The systemic thrombolysis for stroke RCTs Two out of 12 systemic thrombolysis studies suggest a benefit: NINDS-2 and ECASS-3. NINDS-1 tested neurologic improvement at 24 hours and found no benefit. NINDS-2 subjects in the thrombolytic arm experienced milder strokes than those in the placebo arm. Outcome measure = “chance of a good outcome”  12% better (even though goal was to show 20%) Overall: Benefit = NNT of 8 for post-hoc "favorable outcome" measure MAST-I 1995 - <6hrs, increased death (OR 2.7), slight decrease disability (OR 0.5) ECASS 1 1995 - <6hrs, no difference in disability or death (included big bad strokes) ECASS-3 Three to 4.5hrs window; more favorable outcomes with tPA, no mortality difference NNT=15 for “favorable outcome” -  again, milder strokes in lytic arm MAST-Europe 1996 - <6hrs increased mortality and ICH stopped early ASK 1996 <4-5hrs window, slight decrease disability but increased mortality at 3 months; stopped early ECASS-ll 1998 - <6hrs (20% <3hrs) no difference in favorable outcome (modified Rankin) at 3 months ATLANTIS-B  1999 3-5hrs window, favourable outcome at 3 months, increased ICH, slight increase mortality, stopped early ATLANTIS A 2000 <6hrs improved NIHSS at 24hrs but 1 month favored placebo, increased ICH and increased mortality at 3 months stopped early DIAS-2 2008 - 3-9hrs window, notable inclusion is reversible ischemic penumbra on MR or CT; no difference in favorable outcome IST-3 2012 0-6hrs window, short term 1wk increased mortality, no difference in primary outcome (% alive and independent at 6 months) Secondary ordinal analysis showing a “shift” in outcomes favoring thrombolytics Overall harm (symptomatic ICH) NNH: 1 in 20 Issues with the thrombolysis for stroke literature The modified Rankin Scale used to measure outcomes in most stroke trials is subjective. Even among trained neurologists there is variability in categorizing patients into the scale. The modified Rankin Scale has been shown in a systematic review to be unreliable. There is no consistency in the definition of intracranial hemorrhage between trials. Ordinal analysis used in many stroke trials makes the outcomes difficult to interpret. P-values in the studies have been misinterpreted. P-values don't convey the truth, they simply alter the post-test probability. A decent p-value only tells us that a trial should be repl...