> For example, patients who had undergone antibiotic exposure within a 6 month period preceding their admission to the ICU had a 20% increase in mortality rate.

That seems like a very obvious case of correlation vs causation (people who need medicine are more likely to die than those who dont)