Trip Database Blog

Our fourth update on this topic, the last one being in April! I have yet to see any other attempt at highlighting the information/search needs of health professionals, so if you know of any please let me know.

Below are the top 15 search terms used alongside searches for Covid-19 (or synonyms). This is a crude list with no attempt to reconcile synonymous entries.

1. diagnosis
2. PPE or masks
3. chloroquine
4. pregnancy
5. screening
6. interferon
7. chloroquine OR hydrochloroquine
8. surgery
9. dentistry
10. vaccine
11. antibiotics
12. azithromycin
13. treatment
14. vitamin d
15. ivermectin

So, what’s changed?

Firstly, diagnosis is now in top spot…

Secondly, there are increases on the topics of vaccines, vitamin d and ivermectin.

Lastly, there seems to be tailing off of searches on the topic. The graph below shows a cumulative total of searches for Covid-19 (and synonyms) – that’s the thick blue line. The dashed yellow line is a trend line based on the first months of the pandemic. 

The new site is getting ready for testing…

No new functionality, just modernising the code to reflect modern web standards. It’s been a mammoth effort and once released we’ll be able to move much more quickly with new features – and we have a big list of these 🙂

As well as the reprogramming we’ve taken the opportunity to refresh the design. Again, this is subject to testers feedback and alteration. But here’s a very enigmatic sneak preview:

Trip’s aim is to support health professionals to easily obtain answers to their clinical questions. This requires (1) clever systems to interpret search queries and (2) then find the most appropriate evidence. New evidence is published all the time and an important task is getting it added to Trip in a timely manner!  This is done in two main ways:

• Automatically – the majority of our content is added automatically
• Manually – where automation isn’t possible we have to undertake a manual trawl of the sites and look for new content. This is typically done on a monthly basis

We’ve just uploaded our latest manual content which typically takes 24 hours from adding to the site to being searchable!

As well as existing sources we constantly scan the web for new, quality, content (often helped with ‘tip offs’ from our users). This month we’ve added two new publications:

• Austrian Institute of Health Technology Assessment
• Swiss Federal Office of Public Health HTA

Trip, getting better every day…

Trip has now integrated LibKey in to Trip. LibKey is an amazing tool and it makes Trip even better at supporting our users to obtain full-text articles.

One of the traditional methods of obtaining full-text has been to use something called a link resolver. This aims to link a user with their institutions full-text holdings. For instance if a University subscribes to a journal, the idea is that we send the user to the full-text journal article instead of the PubMed abstract. This is great, to a point. The downside is that it’s dumb! By that it doesn’t know if the institution actually subscribes to the journal – we simply insert the link out (using the link resolver) for every journal article! So, if the University doesn’t subscribe you get a disappointed user.

LibKey, on the other hand, is smart! It knows which institutions have which subscriptions and over which time-period. So, now no more misleading linkouts.

For this to work we need to link your institutions LibKey details with Trip and away we go. Fingers crossed your institution uses LibKey!

Our original SmartSearch worked very well. A user selected a couple of documents and we’d suggested closely related articles, based on clickstream data. This worked really well, when we had clickstream data, often we didn’t.  So, that was a problem.

Our new system will work on a similar principle but will include citation data and semantic similarity data – which articles cite which other articles and which articles are semantically similar to other articles? This will allow for a very rich network of connected articles; in turn allowing users to easily find other articles of interest. Think about this, instead of scrolling through loads of articles the system should overcome much of that. Find 1 or 2 articles that interest you – our system finds the rest.

Testing has been really positive.

Initially, this will be just for Covid-19 data but when the new site is released (?September) we’ll look to incorporate it in to the main Trip.

Much of this work was inspired by speeding up evidence reviews. An important second component of this work is to allow rapid quantitative synthesis. This is being planned and we’re currently exploring funding opportunities to get this off the ground. The intention with this is to create SR quality reviews all done within a handful of hours. Longer than our fully-automated system but much nuanced and could be an important step to full automation.

What is Trip up to, apart from:

• Usual business – easy access to the best available evidence
• Covid – adding as much useful content as possible, trying to understand the searches health professionals are using and helping support the Oxford Covid-19 Evidence Service.

We are actually incredibly busy recoding the site! Much of the site’s code is over ten years old (ancient in internet terms) and, due to the way the site has evolved, the underlying code structure is a bit of a mess. This is a 6-month project and we’re probably two-thirds of the way through. The advantages will be massive and will make future developments much easier to implement.

We have also been doing some amazing work around easily identifying content. This is a much more powerful version of out SmartSearch and uses clever things like knowledge graphs. While work started before Covid (?BC) the crisis has created funding opportunities and – via our work with Oxford – has reinforced the usefulness of the approach we’re taking. This work will also form the basis of a larger bid to fund a rapid-review system that can generate systematic review quality answers.

Oh yes, finally, a few years ago we started working with a couple of academic centres to explore improving our search results. It’s taken a while (so much so I forgot about it) but they have recently shared their results – and they have been able to massively improve the search results ‘in the lab’. So, when we’ve got all the coding redone we’ll start to test it on the live Trip.

So, while on the surface nothing much is changing, behind the scenes things are changing massively – as ever Trip is always an exciting environment to work in!

It’s been two weeks since the last update, so what’s changed?

Well, one thing that hasn’t changed is that around 50% of searches are still just looking, generally, for Covid-19 information ie no additional search terms; just the disease name! Oh yes, another search that’s stayed the same is that chloroquine (or hydroxychloroquine) is still the most popular search other than just the disease name.

Topics that have become more popular include:

• pregnancy
• PPE
• vaccines
• dental

Topics that have seen big increases in popularity:

• antibiotics
• young people

I feel I should be adding some extra value by offering some insight, but I’m really not sure what these changes mean!

What a difference a week makes!  In our previous posts generic searches for just Covid-19 (or synonyms) dominated. But things have changed dramatically – as have the number of searches, these too have increased hugely. After simple searches for the disease the top searches look like this:

• pregnancy
• PPE
• diagnosis
• interferon
• screening
• azithromycin
• surgery
• outbreak
• vaccine
• antibiotics

Are users getting past the need for general, background knowledge and seeking answers to more specific questions? It certainly seems that way!

Just under a week ago we posted the first list, so time for an update!

There are still lots of searches just for Covid-19 (or variants, such as coronavirus) but there are now lots more variation in frequent searches.  After simple searches for the disease (and removing country specific searches) the top searches look like this:

1. diagnosis
2. chloroquine
3. treatment
4. blood test nasal swab diagnosis
5. ibuprofen
6. pregnancy
7. critical care
8. pediatrics
9. hydroxychloroquine
10. symptoms
11. vaccine
12. interferon
13. patient transport
14. diarrhea
15. nursing