NHS Evidence

NHS Evidence is a very promising start - it's already good enough to be used in favour of Medline and Google for quick searches on clinical effectiveness in the UK. It could do with a really good database of recent systematic reviews because DARE is not up to date.

One odd result so far - a search for BNP (a peptide not the dark side of UKIP) in heart failure worked well, but when the 'diagnosis' filter was applied the results became less relevant to the use of BNP as a diagnostic tool. It's difficult to get get metadata right on search engines (Berry Brothers does a very good job, but that is because of the nature of their business). Maybe not even worth it. Just let users put in the words they want and let the technology de-fangle the query.

The end of clinical coding

One of the reasons put forward by its advocates is that clinical coding is a vital component of a standardised health record, which is itself a vital pre-requisite to sharing clinical records. So the investment goes on, in initiatives, committees, sub-committees, standards, taxonomies, meta-taxonomies, ontologies, to support the wide range of coding systems used in health.

The effort would be worthwhile if coding was essential to the accurate structuring and transmission of the information in a medical record or medical information more generally. One recent incident has shown how fragile that claim is. And there is nothing to suggest that using SNOMED-CT will improve matters, since SNOMED-CT is itself far too large, complex (and oddly incomplete) to accurately and consistently convey the meaning of free-text or even semi-structured text in a compressed format. (Hint for Google - SNOMED-CT is to health information as the semantic web is to understanding language)

"The use of codes ensures the information derived from them is standardised and comparable." is an often repeated claim. The fact that it does neither continues to be overlooked, and there is still an open question about the value of coding in support of decision making, patient safety. Where coding clearly has a role (HRGs, Read codes) is in support of physician and provider payments. Even this role is of questionable value , and there is always a risk of coding being affected by the level of payment associated with a class.

An electronic mess

I'm trying to access the European Journal of Heart Failure, whose website says it is available on Science Direct. But it's not on Science Direct. The catalogue of the University I'm working at says it is available on Science Direct. But it's not. TDNet says it is available on Science Direct...even the journal website says it is available on Science Direct. But it's not.

Rooting around, there was an announcement on Science Direct last August that EJHF was transferring to OUP in 2009.  More rooting around finds an old message:

The European Society of Cardiology (ESC) has transferred the publishing contract for the European Heart Journal from Elsevier to Oxford University Press(OUP). OUP started publishing the title as of volume 26, 2005. All publishing and electronic rights have also been transferred. The archival content is available via the new publisher’s website.

In the majority of transfer cases, Elsevier manages to secure the non-exclusive archival rights, allowing content to be available on ScienceDirect as well as the new website. Unfortunately, in this case, this was not possible and the content had to be removed from ScienceDirect. Whilst such removals are not common, they do happen occasionally and Elsevier is legally obliged to comply.

Maybe ECS/OUP take all their content away from Science Direct. Whatever, it's a bit of a mess.

Semantic Web - the unreasonable effectiveness of data

Not much response yet to The Unreasonable Effectiveness of Data , except for this oddity . It makes the case  against RDF style angle bracket web intelligence pretty effectively and adds to the 'meaning is use ' approach to semantics.

Information Architecture - moving on

Jesse James Garrett

Here are some of the most common terms from past IA Summit programs: taxonomy, thesaurus, controlled vocabulary, metadata, faceted classification, navigation, content management -- and then there was that one year with all the talks about tagging. Like my grandparents, we cling to these things because they are what saved us. They are the tools by which we proved that yes, IA is real, and it is valuable. But that war is over. We won. And now it's time to move on, because those comfortable, familiar things represent only part of what information architecture can be.
So it's time to leave the nest. Thank you, Lou and Peter. Thank you, library science. For getting us off to a great start. For giving us the tools and knowledge to win a place for IA in the world. There will still be a place for library science in IA, but it's only a part of our larger destiny.

DirectGov search

Why does Jayne Nickalls thinks it significant that DirectGov has teamed up with Internet Explorer 8 ? Given MS's desperation to revive it's fortunes on browsers I'm sure it didn't cost DirectGov a penny. But will it have any real benefit for the public? You have to wonder.

Anyway, the real problem with Direct is its search engine. Accelerating people to it isn't going to be much fun:

FAST talk

The best way to follow FAST's fate is through commentator Stephen Arnold, who really does know a thing or two about enterprise search. Here is his latest:

More problematic is the Fast Search & Transfer technology. Perhaps it was once “best of breed”, but now the Fast ESP (enterprise search platform) has become more complex with the addition of new home grown functions, components obtained via licenses or open source, and the integration of sophisticated third party functions from other vendors.

Is anyone talking about FAST in the NHS?

Metrics for enterprise social computing

Consumer social computing - obsessive twittering, midnight blogging, endless xbox live sessions, Wikipedia editing - hardly needs to prove itself. The volume of activity tells its own story.

But what about social computing in the Enterprise? Enterprise Social Computing metrics are in their infancy. Take one recent example. A large pharmaceutical company switches from a range of ad hoc user-introduced social computing tools to a strategic platform. Gartner reports that 7000 internal websites are sitting on the new platform. And says that this has 'fostered information sharing and collaboration'.

Really? Maybe these are 7000 little used silos. What the Gartner report doesn't say is that a site at Pfizer is considered active if it is used once a month. That's not very often for a tool that supports the lifecycle of a team. How many sites would there be if the metric was once a day? We don't know - the question for Gartner and other independent analysts is - what are the reliable metrics about enterprise social computing to base investment and management decisions on?

More strange results from Clinical Decisions

I mentioned a while ago that the Scottish Clinical Decisions search engine was giving some odd results. It's based on FAST, which claims to improve with use, so I went back today and tried some similar searches. The result of a search for 'effectiveness of heart failure specialist nurses' was, if anything, more off the mark than before:

I wonder what's going on? Meanwhile, I'll be sticking with TRIP or Google, or even Medline.

 

Taxonomy to go?

From the BBC

Council leaders have compiled a banned list of the 200 worst uses of jargon, with "predictors of beaconicity" and "taxonomy" among the worst horrors.

At library school

The author Michele Roberts trained at UCL in the early 1970s

For example, all public libraries in Britain used the Dewey decimal system, which categorised human knowledge as a pyramidal structure. Subjects were arranged one to 10, and under them came sub-divisions, and then sub-divisions of sub-divisions. The system locked subjects into fixed places. Under category seven, Sociology, you could find Women, alongside Lunatics and Gypsies. Men were nowhere to be seen: as designers of the universe, and of classification systems, they did not need to be visible.

Serendipity - collaborative curation / research data curation

RSS readers are modern serendipity machines. This morning reports from Jon Udell and Alf Eaton, one about establishing patterns for the collaborative curation of calendar data, the other about bootstrapping transparency in science data .

Broken users or broken system?

Some time ago a librarian blogger chastised users in this case a PhD student) for not using Medline properly and not wanting to be inducted into the complexities of the Medline interface. Here is some more evidence if it were needed that librarians need to adapt their systems to user needs. It will still be possible to return high quality data sets to Google style interrogations of Medline - it just takes a bit more effort, which begins by not wishing users would become au fait with the complexities of MeSH headings and Boolean queries.

In the library

I had occasion to visit the University of Worcester library today. Very nice and pleasant. Wireless network if you had the password. Good coffee just around the corner.

The thing that struck me after being in the library for a little while was that almost no-one was consulting a book or journal. People were either pecking away at a laptop, or chatting quietly in small groups.

EtherPad - for really real-time collaboration

Etherpad  - surrealists would have loved it

Sharepoint - MySite - yawn or groan?

On a Sharepoint portal where apparently I have a MySite I get the following:

The Cochrane conundrum

A recent paper at the Cochrane Colloquium highlighted the need for free access to best evidence. However, despite providing significant funding for their production, the public still has to pay Wiley Interscience for full access to Cochrane Reviews.