In reply: To a carpenter, everything looks like a nail, and to a stroke thrombolysis sceptic, every study seems to confirm their opinion. A result cannot be “clinically important” yet statistically not so. We found no difference in symptomatic intracranial haemorrhage rates between our study and others using the same definition. The most widely accepted estimates of the true number needed to harm (for functional outcome at 3 months, following treatment within 3 hours) is about 30, as opposed to a number needed to benefit of around three.1
Publication of your online response is subject to the Medical Journal of Australia's editorial discretion. You will be notified by email within five working days should your response be accepted.