What every intensivist should know about common pitfalls in observational research

Pitfalls MethodologyWith the growing availability of large clinical databases, the number of studies addressing research questions using such data has increased substantially.What is often overlooked is that the appropriate methodology for such studies, as well as the valid interpretation of their findings, depends entirely on the type of research question being addressed, whether descriptive, predictive, or causal.A common pitfall in observational research is the failure to clearly specify this research aim, causing analyses and interpretations to fall somewhere between descriptive, predictive, and causal domains, which can lead to false conclusions [1,2].