Thursday, November 20, 2008

NHS Evidence

NHS Evidence have published a briefing document. It's pretty predictable and certainly doesn't look radically different from what the NLH aims to achieve (click here to see the NLH's aims/objectives).

Although I knew about it already, it was still difficult to see them explicitly stating that a dedicated Q&A service was out (see page 7).

I forwarded the following quote to Dr Gillian Leng (who I met early on in NHS Evidence's existence to discuss a Q&A service):

"Despite attempts to evolve search into something more human friendly, there’s still a big hole there. As useful as Google is, it doesn’t answer questions very well..."

The quotes comes from the industrial strength TechCrunch (click here for the full article). My reason for sending the quote is my brand of logic, which goes like this:

  • No-one doubts that clinicians generate a large number of clinical questions.
  • Technology alone does not answer clinical questions.
  • Therefore, you cannot rely on technology alone to answer clinical questions.

So, I'll be fascinated to see what a busy clinician with a burning clinical question gets via NHS Evidence. A few 'easy' facts:

  • NHS Evidence will concentrate on information from accredited sources with a focus on synthesised content (e.g. systematic reviews, meta-analyses).
  • Our analysis of 327 dermatology questions showed that such 'synthesised' content answered 2% of the questions.
  • Therefore, even if a clinician uses a half-way decent search they will still get to content that is unlikely to answer their question.

For the sake of the patients and the clinicians I really hope NHS Evidence has got something up it's sleeve.

3 comments:

Martin said...

Though I understand your objections against NHS's plans, I would not be so negative about using synthesized knowldege to answer clinical questions. I thin it will not work in dermatology: How many SR are there in dermatology? Cochrane Skin group lists a mere 31! That's nothing, but other fields like cardiovascular medicine have much more, making NHS's plan feasible...

Jon Brassey said...

I agree that not all Cochrane groups are created equal. From experience I would say most are poor at creating SRs that answer the questions of primary care clinicians. A notable exception is the Cochrane Airways Group, who have a different approach to SRs (in topic selection) that others.

Currently, I do not believe synthesised evidence is focussed on the 'target audience' of clinician's with clinical questions. Again, a notable exception is the CKS guidelines (formerly PRODIGY).

My concern is that if you focus exclusively on secondary evidence you'll not produce a resource to help clinicians answer questions.

Cheers

jon

Martin said...

Yes, I'm with you on that such a collection can only supplement a Q&A service.
From my experience, 1/3 of the Q&A work can be done using own knowledge and product information sheets, 1/3 using systematic revies and for the remaining 1/3 there's only scare evidence (if any). (but this may be specific to our pharmacology Q&A)