There’s a tale told in the North of a man drunkenly scrabbling on the path at night, met by a neighbour also on the way home from the Hound Of The Roofers pub.

“What’ye doin’ Arthur?” asks the upright man.

“Loookin f’r mi key!” responds Arthur.

“Did y’jus drop it?” asks his inquisitive pal.

“Nay, lad, lost it back t’Hound” grumpily returns the desperate scrabbler.

“Then why y’loookin thar?”

“Is too dark to loook back thurrr, o’course!”

I’m imagining many of us have had a similar experience of looking in the pockets we’ve already searched for a pen, list, pager, battery, phone, specs or similar. And we’ve probably done the same when looking at evidence for the treatment of our patients.

The process of evidence based medicine begins with Ask A Question – in a focussed and patient-relevant way. Now if we only ask questions of the areas we know a little, or where we know there are data to be mined, are we any better than Arthur searching in the light for his key, lost in the dark? Do we tend to ask our clinical questions only under the lamp-posts? And how can we do it differently?

• Archi