Response to Letter by Sanossian et al
We wish to thank Dr Sanossian et al for their interest in our article and for their kind words. They are certainly right when they say that our study has some limitations: in fact, we have no information on prior vascular events, admission temperature or diabetes of patients randomized in International Stroke Trial (IST). We do not know about baseline dementia, but IST patients could have been randomized if they were independent before their stroke, so it seems unlikely that this is a real problem. Also, premorbid statin use should have been very rare at the time IST was conducted. We would also say that, given the huge number of patients we have analyzed, factors like diabetes, temperature, etc, should result to be equally distributed in the 2 groups, attributable to the effect of randomization; obviously, this does not apply to prior vascular events, as we have discussed in our article. As far as the way we measured stroke severity, to our knowledge there is no direct comparison between Oxford Community Stroke Project (OCSP) method and National Institutes of Health Stroke Scale (NIHSS), and therefore, while waiting for such a study (which will be possible with data from IST 3!), the use of one method instead of the other is a matter of subjective preference; we respect the preference of our American Colleagues, but it is not correct to state that NIHSS is “a higher resolution tool” as compared with OCSP, because there is no proof of that.
Finally, we briefly discussed in our addendum the article by Sanossian et al, and did not actually mention the problem of prior history of transient ischemic attack or stroke, but simply considered that 51% reaching of Rankin 0 or 1 suggests some kind of sample selection. We still think that, despite the above mentioned (and correctly depicted by Sanossian et al) limitations of our study, we can reliably conclude that previous aspirin use has no material effect on stroke severity.