4 ms·

What is the effect in this study? The linked article has no info, not even the doi.

• OJFord 3y ago
  I linked it in another comment, but that's now a flag-hidden thread, so: https://heart.bmj.com/content/109/Suppl_3/A78.abstract https://heart.bmj.com/content/109/Suppl_3/A78.abstract
  • tgv 3y ago
    Thanks for the link. I'm not impressed by the effect at first sight. I certainly wouldn't want to exclude anything based on it. I mean, it's 13% more on Monday, which makes it "significant", but if you subtract the 3.1% you mention, you get close to the Sunday fraction, which isn't significant (p>0.05, which is a lousy statistic anyway). While it looks there's something going on, it's not enough to ignore the effects of data manipulation.
    • haldujai 3y ago
      You're assuming that the 3.1% for > 12 hours is different between weekdays and weekends. While the study you're quoting grouped after-hours with weekends there were less delayed presentations in this subgroup compared to the M-F business hours group. Other weekday numbers will also have delayed presentations included so you can't just "subtract 3.1" from one day and declare statistical insignificance. What you can do is subtract it from every day as we know that 97% of STEMIs present within 12 hours. As this is just an abstract we don't know what the authors did in this particular example but it's not the first study to suggest Mondays have the highest ACS rates.