Trip Database Blog

In the registration process there is a section for ‘Institutions’:

This is optional and can be ignored.  But it does serve two purposes, should you not ignore it:

1. Better links to full-text.  If you are at an institution that subscribes to journals we can use something called a link resolver to link from Trip directly to your institutions full-text (as opposed to the PubMed abstract.  This is a powerful feature but does require a link resolver.  To find out if your institution has one – ask in the library.
2. Premium membership. If the institution purchases a premium account it can be used to connect users to the premium content/features.

But, to repeat, this can be ignored, it is not essential for your registration.

It’s always an exciting time when we release the new site and this time is no different.  In fact, I tend to get more excited, as the site gets better and better.  So, what’s new?

A beautiful new design which has been based on a in-depth analysis of user activity on the site.  So, similar functions are better grouped together.

Three new extra languages supported in Trip.  These allow users to search in their own language and see results in their own language.  Feedback has clearly demonstrated that query formulation is the hardest translational activity for non-native English speakers.  The ability to search in your own language is a real boost.  The new languages are Spanish, Hungarian and Swedish.  These sit alongside French, German and Czech.  This language support has been made possible by our inclusion in the EU-funded, Horizon 2020, KConnect project.

We’ve also added two new categories of results:

• Regulatory guidance which includes NICE, IQWiG, SMC and new content to Trip – FDA and EMA drug evaluations.
• Ongoing systematic review, via PROSPERO.

We have also created a filter for articles on Overdiagnosis.  Once you’ve searched you just click the button and we return documents on overdiagnosis and overtreatment.

We have noticed, from our analysis of site usage, that the ability to refine initial search results by clinical area is really popular.  So, we’ve promoted it to make it more prominent.  It’s another way to refine your results so it now sits alongside ‘refine by evidence type’.

Other highlights include:

• Renaming of ‘Trip Premium’ to ‘Trip Pro’
• Simplified the functionality around each article e.g. tweet this, starring etc.
• Functions such as search history, login are all consistent and grouped together.
• A further simplification has been the different evidence types.  This has meant that patient information is now included in the main search leaving the ‘Pro’ content of Videos and Images available from just above the search results.
• Creation of an explorer function (again located at the top of the search results) which was previously called the search safety net.  At the time of typing this is still being copied over to the new site so won’t be live till later today/tomorrow!

Finally, mobile is increasingly important way of users accessing Trip and as part of the redesign we have significantly enhanced our mobile offering.

I was looking at a post from Cochrane Canada on Facebook relating to a free answer they’re making available from the Cochrane Clinical Answers.  The question they were supplying an answer to was “In people with people with epilepsy and depression, how do antidepressants affect outcomes?”.

So, I thought I’d see how the Trip Answer Engine answered it (our still under development project).  There is still work to do on the algorithm and currently it’s on a standalone site – so it looks poor.  But the functionality is there.  So, to answer the question we went to the test interface:

We then press search and the result comes back:

So, how does this compare to the real answer:

Call me biased but our answer appears pretty good.

A massive advantage of our system is that it’ll always be as up to date as the published literature (it generates answers from the literature in Trip) and will answer loads more questions than any other resource – when we launch we’ll probably be able to answer 100,000+ questions.

Below is the new homepage for Trip.  I’ve blurred it, a bit, to retain some mystery.  The homepage has changed considerably the results page less so.

I’m still hopeful to get a test version by the end of February and then – hopefully – a full release by mid-March.

Click on image to enlarge it!

In 2011 I posted:

“An answer engine.  For commercial reasons I have to be vague on this for now (something I’m not comfortable with) but it’ll take shape over 2012.“

I have revisited the idea many times in the last five years, always trying to solve the same problem: a search engine doesn’t answer questions, it gives you 10-20 results to articles that may answer your question.  I borrowed the term ‘answer engine’ from Steve Wozniak (co-founder of Apple) who used it in relation to the release of Siri.  He said people want and need an answer engine.

Until recently the problem has been scalability – how to extract the answers from the great content held in Trip.  But, due to my involvement in various machine learning initiatives, the missing link(s) has been found and we’re busy developing an ‘answer engine’.  Our initiative works on a number of connected problems areas:

• From a set of search terms can you infer the likely question?
• Can we find an answer to that likely question?
• Given there will be potentially multiple answers in Trip, can we surface the best answer?

We have made great strides in all three areas.  So much so we are currently working on the second iteration of our answer engine.  For the test it has two search boxes based on PICO.  The first box (P) represents the disease and the second box is the (I) intervention of interest. To illustrate what it does is using the P of ‘sore throat’ and I of ‘antibiotics’ – the answer engine returned:

“Antibiotics appear to have no benefit in treating acute laryngitis. Erythromycin could reduce voice disturbance at one week and cough at two weeks when measured subjectively. We consider that these outcomes are not relevant in clinical practice. The implications for practice are that prescribing antibiotics should not be done in the first instance as they will not objectively improve symptoms.“

The above is a great answer.  But we can’t rely on a single test so we’ve been extensively testing it and the current version gets the following results:

• Fail – 19%
• Partial pass – 32%
• Pass – 49%

So virtually a 50% success rate, which impressed me!  But the biggest reason for the partial passes is our system not pulling through the answer (a relatively simple fix) and the biggest reasons for the failures was the inability to exclude additional terms which confused the answer (again, this should be a relatively simple fix).

Our system should be great for a number of reasons:

• It can integrate seamlessly with the existing Trip interface but also act as a standalone product/app!
• It can easily be integrated with our multi-lingual systems to users will be able to search and obtain answers in languages such as French, German and soon to come languages such as Spanish.  No English will be needed!
• It will always be as up to date as the answer are based on all the evidence in Trip.
• It is modular and will launch with a focus on therapeutics.  It will then expand to include medicines information (eg side-effects, interactions, contraindications) and from their I’d like to tackle clinical guidelines or lab tests (depends on the resource availability).
• A user can get an answer in less than 5 seconds, when previously they would have had to scan the first 10-20 results to see which result was most likely to answer their question.  Assuming they get the best article they still need to read/scan it for the answer.  So, considerably longer.

 This is brilliant, it really is…!

We are hoping to get the latest upgrade out by the end of February.  There are a number of changes, including:

• Homepage design has been changed radically and it looks lovely.
• Results page has been tidied up and functional items grouped together.  This is not a radical change, more a subtle one.
• We’ll be adding three new types of data: regulatory guidance, ongoing systematic reviews and overdiagnosis/overtreatment.
• As part of the analysis of the way users used the site we are promoting the clinical areas feature (refining results by areas such as cardiology, primary care) to make it more usable.
• Drug information. The main reason people use Trip is to answer clinical questions to help support clinical practice.  One thing that we don’t do is offer drug information support (e.g. dosage, contraindications, adverse events) but we will be doing with the new design.  It’ll help make Trip much more of a Point-of-Care tool.
• Broken links improvement.  We have a really crude system for updating broken links, but by the end of February it’ll be much improved.  While we’ll not be able to guarantee no broken links, the issue will become less relevant over time!
• Improved demarcation of free versus premium – currently premium members can find it difficult to know if they’ve actually got premium.  The new design should help with that!
• We’re going to be re-naming ‘Trip Premium’ to ‘Trip Pro’!

Finally, we’re also working on an answer engine, a mythical feature I’ve been talking about for years.  We are making real strides in this direction.  While it won’t be ready for the end of February to be fully integrated in the site it might well appear in the ‘Labs’ section of the site – for people to try.

We had a huge number of searches for diabetes in 2015, far too many for us to analyse easily.  But, I wanted to explore the topic – it’s one of our most popular searches.

To keep things manageable I have analyses situations where a user has performed any search that contains the term diabetes and has then restricted the results to a specific clinical area, in this case cardiology, primary care and psychiatry.  The results are below.  It’s fascinating to see the differences between the images as it gives an insight – perhaps nothing more – into the intentions of users.  For instance:

• In the psychiatry screenshot the term mellitus is much more prominent
• In primary care they like the word management.
• For cardiology, not surprisingly hypertension is prominent, but then so is children (more so than primary care) – which I find more surprising

Searches for diabetes and restricted to cardiology

Searches for diabetes and restricted to primary care

Searches for diabetes and restricted to psychiatry

I’ve already posted a breakdown of stats for the site (see 2015 in numbers).  The highlights for me are:

• 3,700,000 – page views
• 900,000 – individual sessions spent on the site
• 650,000 – number of users of the site
• 4,560,603 – number of minutes spent on the site
• 350 – mentions in journal articles
• 881,280 – number of times Trip helped improve patient care in 2015

Apart from our continued impact on health care related to search we made large strides towards financial security with the launch, in May, of the Freemium business model.  Uptake has exceeded our expectations with a significant increase in income.  However, we are far from secure, so any suggestions around additional income streams and/or sponsorship opportunities – I’m all ears!

The one slight mistake we made, in moving to Freemium, was to insist on users logging in to use Trip, which resulted in a significant user backlash.  We listened and removed that ‘feature’ and that resulted in a large increase in usage towards the end of the year.

We’ve been making some great progress in relation to our work around clickstream data.  For instance:

• Clickstream data and results reordering. Still experimental but very solid groundwork for future developments.
• The search safety net

A new site, rapid-reviews.info was launched to help separate out the Trip related content (on this blog) from the content related to rapid/systematic reviews.  Talking of which, my post ‘A critique of the Cochrane Collaboration‘ reached over 20,000 views last year.

As part of our work with the EU-funded KConnect project we introduced some very nicely integrated multi-lingual search function.  It’s currently restricted to French, German and Czech but more languages will come online in 2016 including Spanish.

But what else can we expect in 2016?:

• A big announcement around our freemium business model in the next week or so.
• A significant redesign of the site, hopefully ready by February.
• A first draft of our answer engine concept which is looking really exciting.
• Further work on rapid reviews including working on understanding important regulatory data and improvements to the Trip Rapid Review system.
• The rollout of a new system to help cope with broken links.
• Further development, related to our clickstream data, around analytics and insights e.g. A new analytic ‘toy’ and Searching for hypertension.

Running Trip is full of interest, lots of excitement and helping support care with high-quality evidence across the globe is the icing on the cake.  One thing I’d like to see this year, apart form continuing financial security, is to have a greater academic presence/impact.  I’m not even sure why, I think I’d just enjoy it (based on interactions over the last few years).

I will sign off now and wish you all a happy 2016.