How do you arrive at the questions? Using contrived questions will introduce an academic bias. As you know we answer lots of genuine clinical questions and find academic resources such as Cochrane pretty rubbish. My view is that the topics selected by Cochrane have little relevance to clinicians 'at the coalface'. A recent analysis of our dermatology questions - 327 in total - Cochrane reviews answered 2!
Outcomes - very interesting! Do we use 'real' clinicians to rate if the search engine helped them answer their question? If so that doesn't mean the answer is correct. So do you have two outcomes - one relating to user satisfaction and the other to answer accuracy? Again a problem is what's the right answer? I'm sure you saw the NEJM results when they presented a hypothetical clinical scenario and got readers to vote on the suggested management - wide variety of results suggesting there is no clear answer.
from Jon Brassey
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment