Wednesday, May 26, 2010

An evaluation of TRIP (sort of)

I've just seen this rather nice paper about TRIP. The article is from an Italian publication, so all I can go on is the abstract. As such I'm not sure how they searched TRIP and what other databases were involved. Irrespective of that it's quite pleasing!

Bibliographic research of efficiency tests: analysis of the validity of the meta-database
Galbiati G, Casati M.
Prof Inferm. 2010 Jan-Mar;63(1):3-8.


Aims: The use of a meta-database as a first approach to bibliographic research can be just as efficient asinterrogating single data-bases of the litature. The advantages and drawbacks of the two strategies are compared .
Method: A comparison of the results obtained using an identical interrogation made using the TRIP meta-database and different single databases (15 of guide-lines, 4 of systematic reviews, 3 prevalently consisting of primary studies) made it possible to analyse these methods as well as to study 4 meta-databases and identify the most efficient one.
Results: Using the same MeSH terms in both strategies, the following results were obtained: 204 publicationsusing TRIP and 475 using different databases. Evaluation demonstrated the pertinence of 142 (69,6%) of the 204 found using TRIP compared to 185 (38,9%) of those elicited by single data-bases.
Conclusions: The TRIP meta-database yields a lower number of documents but with a higher degree of pertinency, meaning that the researcher employs less time finding pertinent documents. With respect to the traditional approach, beginning research by testing the efficiency of the TRIP meta-database proved advantageous.

Monday, May 24, 2010

Facebook Questions

I have a real passion for getting answers to people.

I have a passion for search, but search is of limited value in answering questions. The typical search engine returns 10-20 results which the algorithm believes will answer your question(s). Imagine if search engines didn't exist and you got 100 doctors in a room and asked them what features would they like to see in a system to answer their questions. Do you think any would say 'give me 10-20 results to articles which may answer my question' - but here we are.

Anyway, all that preamble is due to news that Facebook is working on Facebook Questions, see this TechCrunch article for more info - this'll be interesting.

Sunday, May 16, 2010

Monthly update

I've just finished the monthly update to TRIP with 668 new articles added manually and 2,500+ automatically.

When reviewing the manual uploads I often come across themes and this month the stand out, for me, was wound care. We've answered lots of questions on wounds over the years and the lack of evidence if obvious. So, three new reviews are very welcome:

1) Water for wound cleansing (Cochrane)

Conclusion: There is no evidence that using tap water to cleanse acute wounds in adults increases infection and some evidence that it reduces it. However there is not strong evidence that cleansing wounds per se increases healing or reduces infection. In the absence of potable tap water, boiled and cooled water as well as distilled water can be used as wound cleansing agents.

2) Wound cleansing for pressure ulcers (Cochrane)

Conclusion: We identified three small studies addressing cleansing of pressure ulcers. One noted a statistically significant improvement in pressure ulcer healing for wounds cleansed with saline spray containing Aloe vera, silver chloride and decyl glucoside (Vulnopur) when compared with isotonic saline solution. Overall, there is no good trial evidence to support use of any particular wound cleansing solution or technique for pressure ulcers.

3) Silver-Releasing Dressings in Treating Chronic Wounds (Swedish Technology Council)

Main conclusion: The scientific evidence is insufficient to determine whether silver dressings differ from dressings without silver in terms of effects on the percentage of healed wounds, wound size, pain, quality of life, percentage of infections, and use of antibiotics in treating chronic wounds. The reason is that too few studies of sufficient quality are available. The studies reviewed have not identified serious side effects or complications related to silver dressings, but they were not designed to study this specifically.

Thursday, May 06, 2010

TILT

A further update on TILT - it's generating some very interesting outputs. For those of you unsure of what the TILT project is - it's an area allowing clinicians to record and share learning. Some examples of the learning we've received so far:

  • Ropinirole for Parkinson's disease can cause ankle swelling as a side effect
  • For every 1% increase in omega-3 intake, HDL levels rose by 2.5 mg/dL
  • DISH is Diffuse Idiopathic Skeletal Hyperostosis and is failry common in men and represents calcification of ligaments tendons etc principally around the Thoracic spine. It has a classical 'dripping wax' appearence on plain radiography. There may be an increased risk of diabetes and poss heart disease in patients who have DISH
  • Medial arch pain localised to bone rather than soft tissue may indicate an accessory navicular bone! usually unmasked after a twisting injury
  • Both antiseptic cream and silver nitrate cautery are likely to be effective in preventing recurrent idiopathic epistaxis in children Cautery is indicated when there are obvious telangiectatic blood vessels

Lots of feedback from users, many reporting very positive examples of learning from others and others impatient to see it being taken out of 'proof of concept' and some serious development undertaken.

So far very, very interesting

Tuesday, May 04, 2010

Update on TILT

TILT (see previous blog) has been running for a little over a week and so far so good.

We have had around 50 recorded instances of learning and crucially people have learnt from others. In other words they have seen someone elses recorded learning and reported they have learnt from that. Therefore, it appears that:
  • People will record their own learning.
  • People will allow the sharing of this learning.
  • Other people can learn from other people's learning.

The pilot still has another 4(ish) weeks to run so we'll see how things develop. After that a decision will need to be made about if we adopt TILT or not. If we do we'll need to learn from those involved in the pilot. However, the biggest challenge will be scalability - if we have 10,000 doctors recording one item of learning per week that's 500,000+ items per year. So, the issue of findability is vital. Thankfully, we have a number of ideas to help deal with this - if we proceed.

It's not too late to volunteer to take part, let me know via jon.brassey@tripdatabase.com