Trip Database Blog Liberating the literature

Time 2021-08-20 18:07:11

Web Name: Trip Database Blog Liberating the literature

WebSite: http://blog.tripdatabase.com

ID:188375

Keywords:

Blog,Database,Trip,

Description:

What next for Trip? After over a year if back end development we re looking to the next set of developments for Trip. Below are our current ideas and we d like some help on which to take forward and/or prioritise.Citation data would dramatically increase our ability to link articles. This could enhance our SmartSearch feature. So, if a user clicks on articles #1 and #5 we could identify more, closely related articles, that are relevant to their search intention. Another use could be highlighting the relationship between RCTs and systematic reviews. So, if you find an RCT we could link it to systematic reviews it might have been used in. Conversely if you find a systematic review we can highlight the individual RCTs.Problem Trip is an EBM search engine and we like to connect users to the best available evidence. However, just because something is linked to from Trip it doesn t mean it s not without flaws.Proposed solution we currently, using RobotReviewer, automatically assess if RCTs are likely to be of low risk of bias or not. We have done most of the work to assess guidelines for how evidence-based they are. We have also started some preliminary work to assess systematic reviews for bias and could develop this further.Currently Trip allows users to search and then restrict the search by evidence type, clinical area, year etc. But is this set-up in an efficient way and does it support users to find what they need?Problem users might not be able to find the documents they needs as quickly as possible.Proposed solution (1) introduce an additional category of filters based on clinical question type e.g. therapy, diagnosisProposed solution (2) introduce a step-by-step search guide to really allow users to focus their results. It might have the following steps:User types search termsWe potentially suggest additional terms and user finalises searchWe ask if the user wants a particular type of evidenceWe then ask if there is a particular recency they re interested inIf it s a query about diagnosis, therapyIf there s a particular clinical areaAt each step we show the number of results and when the user is happy we show the results. It s sounds long-winded but it should be quick and should deliver really focussed results.So, it s now over to you to let us know what you think. Please use the poll below (or let us know any alternatives you may have):The new site is out, getting positive reviews and is very stable this is excellent. There are a few teething issues which we re dealing with (95% of them have already been fixed). So, our attention turns to the next steps. The immediate area we re working on is the indexing (as laid out here (with an explanation as to what indexing is)).The step after that which needs considerable planning is to work on improvements to the site. To that end we want to try a new approach to convene a number of small, online, group sessions with users. These are likely to be grouped by user type e.g. health professional and information specialist. These sessions would allow us to better understand how users access evidence, the problems that they encounter and also for us to pitch our current ideas for improvements. The subsequent discussions should allow us to better plan the next steps.Introduce citation data in to Trip.Improve the search introducing greater control over the sensitivity and specificity of the search.Add more document grading e.g. guidelines and systematic reviews.Create specific filters for topics such as therapy, prognosis and diagnosis.Introduce a search wizard to allow an easy way to find the documents users need.Love or loathe those ideas, here s your opportunity to influence. Send an email to jon.brassey@tripdatabase.com if you want to take part.Automated via mechanisms such as APIs and RSS. Once set up this requires no effort from Trip it just grabs new content on a daily or weekly basis.Semi-automated. A number of organisations supply us their new content via email (typically a .csv file) in the correctly formatted way. We then add this to the third way of updating content:Manual. Over a week, always in the middle of the month, Trip manually visits around 300 sites to manually find new content added since the previous month.This month we just uploaded 650 records from the semi-automated and manual method the majority via the manual way. Of these 86 were dated from 2020 or earlier and the rest were from this year. Of the older one these were typically from us having to updated outdated URLs or if we found new sources of evidence. The records from this year covered a large number of sources e.g. NICE, AHRQ, IQWiG, WHO, NIHR, ACOG, NCCMT, FDA, EMA, Prescrire, PHE, PHW, SBU (so many initials)!In addition we manually tag all articles, from 2021, if they re useful for primary care (automatically tagging articles for primary care isn t great). Some examples include:While it s hard work it s always great to see so much wonderful evidence being added to Trip.It s been just over ten days since the new site went live. It has been remarkably smooth. There are a small number of issues but the site appears stable and there have been no critical issues. So, we now look to the future and this is a brief update on that.Firstly, we have a list of around ten issues that need fixing. These have been prioritised and we ll work through those as quickly as we can. Only a few of these I d class as significant.The actual search component the Trip brain, this is working wonderfully and is very modern (so, no updating needed)!The interface this is how people interact with the site. It s the design, the interactive bits etc. It s what we ve spent the last 12 months re-writing.Indexing the search component needs to search content. This involves:Publications being added, either manually or automatically. This is minimal information such as the title of the document, the URL and publication date.Spidering we send a spider (AKA webcrawler or a bot) to the URL and grab the content. Processing our system adds the content to the search component but also does a number of other bits and bobs e.g. tagging the articles by clinical contentOur indexing system is as old as the interface was (some bits being nearly twenty years old) so we re completely rewriting this to make it better (quicker, cheaper etc).Thirdly, planning the next upgrades to the site. We ve got a few projects that are coming close to delivery e.g. guideline grading. However, there are some significant projects we re thinking about and we need to decide which ones to take forward and for that we need to engage with our users. In the past we ve relied on web-surveys and these have been really helpful. However, for this next step we d like to have a more engaged process one where we can have more of a two-way dialogue. To that end we re interested in setting up a series of online small group calls. We d briefly discuss our ideas and then engage with the group to unpick our ideas, what sounds good, what sounds bad and how we might implement these ideas. If you re interested in taking part (one hour session) please let us know via development@tripdatabase.com.I was delighted to recently receive an email from a prominent EBMer who declared himself a fan of Trip! However, his email was more about a teaching session he ran on EBM sources of quick answers and the feedback from the students.He said the more EBM focussed the more they liked Trip however for others Trip was less favoured. He suggested this was because they had an extra cognitive load (in short, more effort) in selecting the resource most likely to be useful to them. Other sources (the likes of UpToDate) tend to give one main result so there s a lack of choice, search and click. For Trip its:While that extra step seems fairly minor, clearly the more time-pressured you are the more of a big deal this is.This is really important and interesting feedback. One for Trip to reflect on but if you have any thoughts, please let me know.We understand how important access to Trip is for mobile and tablet users and as such we have invested significantly in this area with a full-responsive site. There is no requirement for an app (our previous app is no longer supported and will be withdrawn shortly) and users should navigate, via their phone/tablet s browser to the usual Trip page www.tripdatabase.comThe site works well and looks beautiful (while I may be biased I do think it looks great) and there are a few screenshots below. Any comments either email me (jon.brassey@tripdatabase.com) or leave them in the comments.Trip introduced the freemium business model over 5 years ago. It was our approach to remaining both viable and independent. The fact that we re still here (and doing well) is some validation of our approach.Subscriptions are available for individuals at $55 per year while institutions can subscribe, with costs dependent on both their size and organisational type (click here for current prices). But what do you get for the subscription? Below is an overview of the differences and as we develop new features these will mostly favour Pro subscribers:Buying an institutional license is a cost-effective way of bringing the power of Trip to your organisation. Trip can be useful in many different ways, for instance:Supporting clinical care by helping clinicians to easily find robust, evidence-based, answers to their clinical questions.Trip is widely used by systematic review producers.Easy access to hundreds of thousands of full-text articles.The evidence-based content of Trip supports the writing of clinical guidelines.Preparing research grants is made easier with easy access to pre-existing research.PricingPricing is based on size and type of institution.  While not ideal it acts as a starting point for negotiations and if you feel you’re disadvantaged by this method then please contact us to discuss further.  NOTE: figures are in US Dollars and is the annual cost. Discounts can also be arranged for multi-year arrangements.Very small ( 1,000 FTEs) $875Small (1-5,000) $1,400Medium (5-15,000 FTEs) $1,990Large (15,000+ FTEs) $3,500Very small ( 50 beds) $525Small ( 250 beds) $1,075Medium (250-1,000 beds) $1,990Large (1,000+ beds) $2,890Very small ( 20 staff) $340Small ( 75 staff) $1,110Medium (75-250 staff) $1,990Large (250+ staff) $2,925For institutions based in the lowest World Bank income classification (Low-income economies) Trip is free while those classed as Lower-middle-income economies or Upper-middle-income economies generous discounts are available.  For further details contact lmic@tripdatabase.com.AuthenticationTrip currently supports a number of methods but the principle method is via IP authentication.  We can easily arrange free trials of Trip Pro, to request one please email subscriptions@tripdatabase.comAfter over a year of hard-work we have just released the latest version of Trip. We had to re-write ALL the website, replacing code that, in some places, was over 15 years old!We ve tested this extensively so we re hoping any issues will be minor, but if you spot an issue then please let me know: jon.brassey@tripdatabase.com.

TAGS:Blog Database Trip 

<<< Thank you for your visit >>>

Liberating the literature

Hot Websites