Trip Database Blog

The TRIP Database has an extensive synonym function. However, we still get terms we don’t fully understand. This time it was a question received on the NLH Q&A Service asking about ‘candidal overgrowth’. How do you search for this? We think we got it right with candidiasis – but how do you verify this? The requesting doctor didn’t know.

We’ve now discussed the specification with our web people, got the quote (ouch) and a rough timetable. The next steps are as follows:

1. Final specification meeting (sometime next week).
2. Web-people to produce specification document including wireframes.
3. A robust test version should (!!) be available by mid-May
4. Fingers crossed, the beta test version will be made publicly available at the start of June.
5. Although free accessible the beta will not become official (for contractual reasons) till September 1st.

Why not create a TRIP Database toolbar with an embedded search box.

I spoke with my webpeople and they suggested creating a bookmarklet as an interim measure to see if there is any demand. For search examples see bookmarklets.com

Had one of those interesting questions today:

“In patients who are totally blind from retinitis pigmentosa and who have erectile dysfunction, do the usual contraindications to viagra apply?”

What’s interesting is that viagra is contra-indicated in people with retinitis pigmentosa as it may lead to blindess (or that’s the assumption). So what do you do if the person is already blind?

Interesting looking at the feedback on the Q&A site. For 2006 we’ve been rated as follows:

• Excellent – 88.3%
• Good – 10.4%
• Average – 1.3%

I tend to view responses other than ‘excellent’, negatively. So I’ve tried to look at our answers to questions which have been rated as ‘good’ or average’ to see if there is a trend. There appears to be two main reasons:

1. When there is little or no evidence to answer the question. From the users perspective they come to us with an uncertainty and sometimes we can’t help. Anecdotally, the lack of evidence can frequently be useful/helpful. However, in a some situations its not as well received – understandably.
2. Not comprehensive enough answer. Although only mentioned twice in the 2000+ questions we’ve answered I guess that might constitute a trend! Our remit is to provide quick responses to the literature. Occasionally we get a GP using us who has already done a good search asking us for information. Unfortunately, they don’t tell us what’s they’ve done so we answer it as normal ie give brief information.

Still with less than 2% of people rating us at less than good – we must be doing something right!

A few posts ago I reported on the mis-spelling frequency on the TRIP Database. I’ve just been looking at the stats and the most frequently ‘mis-spelt’ term is ‘hypertension’ and ‘diabetes’. To clarify that, it’s not a count of the number of actual times it’s been mis-spelt, more the number of different ways it’s been mis-spelt – 27 (see side)

These are confirmed mis-spellings in that people have clicked on the ‘did you mean: prostate’ to confirm spelling correction!

Nice article on search Can we know everything?

The ATTRACT site contains a wonderful counter that counts how frequently an answer has been viewed. The top five being:

1) What are the risks of flying while pregnant?
2) Is there any information on the Novasure system for treating menorrhagia?
3) Should dianette (oral contraceptive pill) only be used for 3 months in treating acne?
4) Should you treat a low Ferritin (

A post on Dean Giustini’s blog highlights Wellcome’s invitation to tender for a UK version of PubMedCentral. See Dean’s blog (click here) for more information.