Wednesday, November 29, 2006

TRIP in articles

I came across two fairly recent articles that have used TRIP as part of a literature search:

It's nice to see that we get mentioned in the literature as well as being so highly used.

Goodbye Google Answers

Google is shutting down its Google Answers service. As the article syas it might well be to re-launch with something new and innovative in the New Year.

Interestingly, a Q&A 'service' will be one of the core features of Gwagle...

Tuesday, November 28, 2006

TRIP Leaflets

We're near to finalising the design for the TRIP leaflets and they're looking pretty good. If you want any let us know via 'Contact Us' on the TRIP Database - they should be available December or early January.

Monday, November 27, 2006

Ambient Findability

Ambient Findability, by Peter Morville, is a great book. Only half way through but already lots to think about.

Friday, November 24, 2006

Podcasts - few hooked

A news article on the BBC website 'Podcast numbers show 'few hooked'. The article reveals:

  • The survey by the Pew Internet and American Life Project found 12% of US people online had downloaded a podcast.
  • Earlier this year, a survey by the same research group found that just 7% of online Americans had downloaded a show.
  • But despite the growth, just 1% of respondents said that they would download a podcast on a typical day.
  • This figure remains unchanged from the February survey.

Thursday, November 23, 2006

Merck manual

I’m not sure why, but I have an affection for the Merck Manual. It’s not massively popular on TRIP (eMedicine is more popular). But it’s recently been updated to the 18th edition (most updates are 2005); the previous edition – 17 – was in 1999.

There’s something verging on ‘old fashioned’ about completely re-writing a book every 5-6 years. I suppose this reflects the ‘old media’ style approach to things. Some clinical areas move very rapidly while other move much more sedately. Surely it’d be better to have more frequent updates of the ‘rapid’ areas. I suppose the problem with that is that it pushes back the development of the ‘sedate’ chapters.

Still, it highlights the problem of ‘central control’. Merck retains control of the content – making it very expensive (no doubt) to update, making it rapidly out of date etc.

eMedicine, another eTextbook, takes a different approach. It retains control, but updates happen with fair regularity. It appears that a senior clinician is given overall control. I imagine if s/he hears of new information they’ll want it added to ‘their’ chapter.

Then we have wikis such as Ganfyd. Lots of doctors contribute and ‘own’ it – there is no central ownership, but likewise little resource. So contributions are made in the docs own time. As it stands there appears to be a lot of titles (they’re easy) and relatively few ‘solid’ articles. It’ll be very interesting to see how Ganfyd ‘matures’ over time.

Let’s compare the entries for hirsutism:
Merck Manual
eMedicine
Ganfyd

Hardly a scientific test but the Merck Manual will possibly stay the same for the next 5 years, Ganfyd will change slowly, while eMedicine seems the most comprehensive/useful! In fact the Merck Manual (recently published) was last updated November 2005, while eMedicine's last update was July 2006.

Anyway, that aside, I’ve got the Merck Manual to add to TRIP – only another 1,000 links to manually add!

Wednesday, November 22, 2006

Gwagle

Formerly referred to as Socrates, the new project I'm working on is called Gwagle. Still a bit early to post details as they are still being finalised. But we've now secured the domain names - so can reveal the name at least.

The project is a combination of TRIP's skills in a few things clinical and Sequence's (the web people we use for TRIP) expertise in all things web. I have two formal meetings with third parties to gauge their reaction and take on board comments before a final specification meeting on the 4th December. After that the programming for Gwagle will commence. I'm hoping that things might actually be 'out there' on the web early 2007 (Feb/March).

But what is Gwagle about? It 'sort of' combines all the things we're into (e.g. search, Q&A) as well as a few broader concepts. If you know what Gwagle 'means' then you're a little further along the path of knowing what it's about!

Tuesday, November 21, 2006

YouTube

I'm really liking YouTube at the moment and I'm currently listening to the classic Beck song Loser. I'm also really liking 'Set the fire to the third bar' by Snow Patrol and Martha Wainwright.

Aside from my own musical interests I'd be interested to see YouTube being used for online training. You can upload up to ten minutes - the maximum attention span for many people. You could have quick topics on a number of topics, such as:

  • Searching TRIP
  • An introduction to Medline
  • Clinical Queries in Medline
  • What is a PICO

Well, the list could be very long....

Just a thought!

Libel ruling boosts net providers

An interesting decision on libel laws and the internet (click here)

Saturday, November 18, 2006

Search engine failure?

Many of the answers we give on the Q&A service have only one reference. In other words we feel that the answer has been adequately answered with that single resource. Often these are from excellent sources such as PRODIGY. I'm meeting the clinical lead for PRODIGY, Sharon Smart, on Monday to hopefully forge closer links between PRODIGY and the Q&A services.

Aside from that, what does it mean when the user doesn't find the information we do?

  • Are they 'unmotivated' to try the search?
  • They try the search and can't find anything?
  • They try the search, find something promising but can't find the information within the document?

One problem, linked with the last point, is that documents such as PRODIGY is their sheer size. It's fine to find the document you're interested - but in a document that size how do you find the 'needle in a hay stack'? The PRODIGY guideline is split into many sections. For instance the PRODIGY guideline on antiplatelets answers many specific questions e.g.

  • How do antiplatelet drugs reduce cardiovascular risk?
  • How should I manage people who develop an ulcer while on low-dose aspirin?
  • What should I do if a person on aspirin is at increased risk of gastrointestinal adverse effects, or develops severe dyspepsia?

My view is that these should be separate documents. As well as a global 'Antiplatelets' guidelines why not have small documents such as "How do antiplatelet drugs reduce cardiovascular risk?" with just the specific text associated? PRODIGY have moved towards that with their structured view. However, it's not particularly user friendly and doesn't allow users to get straight to the 'good stuff' - the fewer the clicks the better.

This principle could easily be extended to other documents e.g. other guidelines, NICE documents, systematic reviews ie what questions do these documents answer?

I remember Ben Toth (his blog) talking about this a good few years ago! Always a great source of information and advice!

Wednesday, November 15, 2006

Socrates

At the weekend my broadband broke so I was left with my thoughts. On Sunday morning I has a 'Eureka' moment and worked for about 3 hours on a new project, codenamed 'Socrates'. I presented my work on Tuesday morning to a potential partner. Today they called me to say they were 'in'.

So work will start on Socrates almost immediately.

It's too soon to give out details and these will filter out, via this blog, over the coming months. But I'm hoping that it'll be released early 2007. It has the potential to be truly massive.

Watch this space.

Web-based diagnosis

Following on from the earlier post on Google diagnosis I have come across a paper using the Isabel system. The paper Diagnostic omission errors in acute paediatric practice: impact of a reminder system on decision-making is available as a provisional PDF. This gave the following results and conclusion:

Results: Subjects attempted to access the diagnostic aid on 595 occasions during the study period (8.6% of all medical assessments); subjects examined diagnostic advice only in 177 episodes (30%). Senior House Officers at hospitals with greater number of available computer workstations in the clinical area were most likely to consult the system, especially out of working hours. Diagnostic workups construed as ‘unsafe’ occurred in 47/104 cases (45.2%); this reduced to 32.7% following system consultation (McNemar test, p<0.001).

Conclusions: We have shown that junior doctors used a Web-based diagnostic reminder system during acute paediatric assessments to significantly improve the quality of their diagnostic workup and reduce diagnostic omission errors. These benefits were achieved without any adverse effects on patient management following a quick consultation.

A thought provoking article, especially given the recent debate surrounding the Google diagnosis story.

Tuesday, November 14, 2006

New look NLH

I wrote on the 7th August about the NLH's plans to revamp their site and more specifically their search. Well it has just gone live (click here). They've introduced the 'title' and 'title and text' distinction. As I mentioned previously I'm flattered with them copying the old style TRIP method.

But I'd be interested to see the rationale for the changes. Our move from 'title' and 'title and text' to the current method has dramatically improved our search results.

Aside from that I do like the new layout and colour scheme.

Friday, November 10, 2006

Googling for a diagnosis

This article will no doubt feature highly in the blogs (click here) and create much controversy! It reminded me of the Isabel decision support tool.

Basically, the authors took a years worth (n=26) of diagnostic cases from the NEJM and, after reading the article, they selected 3-5 pertinent search terms. Added these to google and somehow selected the 3 most likely diagnoses. They then compared these 3 diagnoses with the actual result from the NEJM.

Overall result - 'correct' in 58% of cases (CI - 38-77%).

I'm not sure if I find this disturbing or not, it's certainly too small to make conclusions. But once again Google (not deliberately on their part) have set a standard.....

Tuesday, November 07, 2006

Another record

Another record, this time in Q&A. Today - my birthday - saw the NLH Q&A Service answer and published 22 questions. The previous record was 17.

Happy birthday!

Friday, November 03, 2006

October - Record Month

It has been two whole months since we went free. Before going free we averaged 22,500 searches per month. In September we were searched 132,538 times. For October, another huge increase - 191,581 searches. This represents a increase of nearly 50% since September and over 800% increase from our subscription days!

Onwards and upwards!

October Zeitgeist

  1. Chickungunya fever (Ganfyd) – viewed 396 times.
  2. Guidelines for the administration of blood components (ANZSBT) - 342
  3. Failure to thrive (Ganfyd) - 276
  4. Guidelines for Blood Grouping & Antibody Screening in the Antenatal & Perinatal (ANZSBT) - 272
  5. Asthma (PRODIGY) - 254
  6. Rheumatoid Arthritis: Diagnosis and Management (British Columbia Medical Association Guideline) - 251
  7. Nutrition support in adults (NICE) - 214
  8. Moxibustion (CAM) - 214
  9. Diabetes services for adults (RCN) - 201
  10. Good practice in infection control (RCN) - viewed 197 times.

Wednesday, November 01, 2006

What is a wiki?

Found by Ben and a very useful intro (click here). I'm seriously considering using a wiki in at least one new project I've got in the pipeline, possibly TRIP Answers.