Developing an answerable question:
Patients or population
You would like to restrict patient variables
to those similar to your patient and relevant
to the outcome of interest.
Uncontrolled variables may affect the
outcome variable in ways that make it
difficult to distinguish their effects from
those of the intervention.