research Collider bias undermines our understanding of COVID-19 disease risk and severity
This review discusses the challenges of interpreting observational studies on COVID-19 due to biases from non-representative samples, such as collider bias, and suggests better sampling methods to mitigate these issues.


