This is a, yet another, study in Stroke of folks claiming it is “safe” to use thrombolysis on patients who are found to have suffered a stroke while sleeping – the so-called “wake-up stroke” population.
The specific claim made is “This retrospective analysis of data in thrombolysed consecutive acute ischemic stroke patients shows no significant differences in mortality, functional outcomes, or bleeding rates between WUIS patients with no early ischemic change on CT and those treated within 4.5 hours of stroke onset.”
…because their sample size is so small the absolute differences are still within the statistical variation expected by chance. This is, unfortunately, a recurring theme I see in these stroke publications, many of which are retrospective registry reviews. Their groups are statistically not different, but this is owing to failed statistical power in study design, as opposed to clinically meaningful equivalence. This is a major difference between retrospective and prospective studies – in which prospective studies choose specific absolute differences necessary to define clinically meaningful equivalence, and then perform power and sample size calculations based on these constraints.
Their outcomes are, incidentally, also simply terrible. They publish a figure comparing outcomes with their wake-up stroke population to their 0-4.5 hour thrombolysis reference group – a 326 patient reference group with 18% mRS 0-1 and a 26% mortality. But then, they further break out the 197 patients from that group that received tPA within the ECASS III license criteria, showing that compliance with guidelines leads to 32% mRS 0-1 and 18% mortality. This therefore implies the other 129 patients – the ones who received tPA outside the license criteria – had utterly dismal functional outcomes and frighteningly high mortality.
Someone needs to go down to King’s College and check up on them and make sure all this off-label use isn’t just costly killing fields.
“A Case-Controlled Comparison of Thrombolysis Outcomes Between Wake-Up and Known Time of Onset Ischemic Stroke Patients”
http://www.ncbi.nlm.nih.gov/pubmed/23723307