From an epidemiological perspective, one of the hardest things about evidence for critical care is that ICUs are really, really hard places to conduct studies.