What you want to know is intimately tied to how and where you search. It's a bit of a chicken and egg situation - but you've got to start somewhere!

It is traditional to divide questions into two types:

• background questions - these are general questions about a problem or disorder eg "what causes lymphoedema?" or "how is HIV transmitted"?
• foreground questions - these are very specific questions that focus on the management of patients. They have three or four essential components:

1. patient and / or problem;
2. intervention;
3. comparison intervention (sometimes not relevant), and
4. clinical outcome.

A useful mnemonic for formulating such questions is PICO.

Examples of foreground questions are: "in a child with acute otitis media, does the addition of antibiotics, compared with symptomatic treatment only, reduce the duration of illness? Or "in older patients with heart failure, does adding digoxin to standard diuretic and ACEI therapy yield enough reduction in morbidity / mortality to be worth its adverse effects?"

To answer background and foreground questions requires different levels of searching skills and the information tends to be held in different places. In the realities of everyday practice, most clinical questions asked by clinicians probably exist somewhere on a continuum between these two extremes. Having a clear idea of what you are looking for will enhance your search strategy and hence maximise the number of useful hits and minimise the time spent searching. The tutorials mentioned below cover this very well.

• Netting the Evidence provided by the University of Sheffield
• Introduction to Evidence Based Medicine tutorial from the University of North Carolina