Searching the medical literature is a problem for orthopaedic surgeons. First of all, mainstream search engines only search as far down as the anatomical joint; but that’s pretty much where orthopaedics starts. This leaves orthopods with hundreds of pages of irrelevant results to sift through. Secondly, whilst these search engines cover traditional research literature, they do not pick up results in other publication areas such as preprints or standards and guidelines, as well as new multimedia resources like podcasts and videos.

In this interview, Andrew Duckworth speaks to The British Editorial Society of Bone & Joint Surgery’s Director of Publishing, Emma Vodden, about OrthoSearch, a discovery tool developed with and for the orthopaedic community.

Can we start with you filling us in on the background of the role and aims of the society?

We are not a big publisher. We have one audience - orthopaedic surgeons, and so we've been able to be completely focused on that community. That goes right back to the start in 1948 when in collaboration with the JBJS, we launched The Journal of Bone & Joint Surgery, British Volume (JBJS Br).

In 1953 the surgeons who were involved with the journal launched a charity - The British Editorial Society of Bone & Joint Surgery with the central aim always being the education of orthopaedic surgeons. And they did that, not through fundraising events, but through the JBJS Br and by selling subscriptions. For 60 odd years that is all we did, we just did the journal. In 2012 though we launched Bone & Joint Research to fulfil a niche that we saw in the research community who were particularly underserved in terms of journal choices. Since then, we've launched Bone & Joint 360, again, with that educational mission; trying to give busy surgeons an overview of the literature.

All have been phenomenally successful and again expanded that charitable mission that has always remained that the core of the society. We also do lots of other charitable activities; sponsoring the ABC travelling fellows, the Mark Patterson Travelling Fellowship, as well as supporting other third-world charitable initiatives.

We further expanded the publishing portfolio last year with the launch of Bone & Joint Open, offering our community another way to publish in open access, and also with a slightly wider focus than The Bone & Joint Journal. We've been, we are, a publisher and we have been doing the traditional publishing thing, until recently.

Can you give a little insight into OrthoSearch and how the idea came about?

We decided that we needed to again, expand what we're doing, and this time to move more into digital education, to not necessarily remain in that firm publisher space. Through much discussion with the community we identified that search was a problem.

Obviously, PubMed's been going for a really long time; it's often the first port of call. The other major search engine is Google. With PubMed we all know that it covers everything. It's got hundreds and hundreds of thousands, probably millions of records, and unless you are an expert in search terms, you often struggle to find the right results, ending up with hundreds of pages of results. They use a tagging system called MeSH, which we're all familiar with but that doesn't particularly serve our community. We decided that that was something that we could address; we created the world's first orthopaedic taxonomy. We've got now over 19,000 terms in our orthopaedic taxonomy, which we are using to tag content. By tagging content correctly with orthopaedic terms, you are able then to create a search engine that responds to very specific terms and concepts, and can importantly, understand them.

We started by deciding what we were going to index and the first step was to get as many orthopaedic titles into the search engine as possible. That's bringing in an awful lot of content, but we also recognise that there's a significant portion of orthopaedics appearing in those larger medical journals where the authors have had to go for a very high impact factor. So, we've developed a relevancy tool, which is allowing us to skim the other literature to bring in only the results that are relevant to orthopaedics. That's a powerful piece of tech that we've developed with our technical partners, Molecular Connections. This was all in response to many, many conversations with the community, because we wanted to develop something that the average orthopaedic surgeon can just use.

It’s like trying to create a Google specific to orthopaedics isn't it! How has that evolved over time and how has it developed into what it is today?

In terms of content, one of the things that's used regularly are standards and guidelines like the NICE guidelines or the AAOS guidelines. But NICE guidelines come out poorly in terms of how you can search them because often they’re documents over 600 pages long. The search behind them isn't particularly powerful or useful. So we've included standards and guidelines in OrthoSearch. They're all being brought into the platform.

We're also indexing. We have brought in the Orthopaedic Proceedings platform on which we publish the meeting abstracts from many, many different meetings. This is already phenomenally well used by the community. At big general orthopaedic journals if your article has appeared on a preprint server, it is not going to be published at that point. However, we recognise with OrthoSearch that preprints are actually an important element of the literature because if you're looking for a research idea you want to be able to see at a glance what's just been published so you don't repeat it, or maybe you do want to repeat it but with more patients.

And we're also indexing podcasts! We've pulled in loads of orthopaedic podcasts from around the globe that we are tagging again, because podcast searches are not particularly useful. Often people will find, you know, a channel that they like, and they'll subscribe to it, or they'll regularly look out for the episodes. But if you were looking for a podcast specifically on shoulder or a particular paper it's impossible to find it. We've tried to make undiscoverable content discoverable.

How is it going to advance on what's already out there for the orthopaedic surgeons across the globe?

This isn't just for The Bone & Joint Journal. This is for everything. It won't preferentially give you our own articles over something else. This is just for the orthopaedic literature and everything else out there. It is independent of all of our publications. Our publications are in there because they're in orthopaedics, but they are not in any way weighted to appear higher in the search engine.

The other thing that we've tried to address here is the fact that there's so much literature out there. How do we know what's good? Or how can we at least give the user some guidance as to what might be a more trustworthy source? By selecting the journals that we've put in there already, we have been able to cut out some of that junk science that you would find in other engines. If you type a term into Google, that junk is just going to come up anyway. In OrthoSearch you won't necessarily get that paper from a publisher that is perhaps a bit shady.

On top of this, we've put in the search engine results on the article listings the impact factor of the journals if they've got them, the cite score of the journal and the Altmetric score of the paper. It just gives you a bit of an idea of the quality of the source of the paper. Altmetric scores are becoming more and more important; if you're interested in looking at what's been popular with your peers, then seeing high Altmetric score shows you that it’s got some attention, particularly on Twitter or in the news. It might be something that you want to look at.

We're not trying to replace PubMed or Google. We understand that if you're doing a systematic search of the literature, then you are going to be looking at multiple search engines anyway. What we want to be is the first place you go, because what you're going to find from the first page to the last page of results is relevant. You are not going to get the random papers you can often get in the other search engines. You can also bookmark your search; if you're using OrthoSearch and you put in lots of filters and you suddenly have to leave, if you bookmark the page you can go straight back to it. Those same filters will still be there. You can make it fit around your work schedule.

I should also mention that this is open for much debate and feedback. We are desperate for feedback. If there's something missing, we want to know about it because we can easily include it, so let us know! Also, it's free to use! It's out there. It's for orthopaedic surgeons.

It also has an update feature doesn't it? If you type in a certain topic like hip fractures let’s say, it will then update you on all relevant publications in that area in a recent period of time.

That's another unique feature yes; as you know, we all sign up to eTOC alerts from individual journals and you'll get the listing showing the 20 odd papers published across the literature. There might be a couple of hip papers you were interested in, but often you’re not interested in the rest. What we have created at OrthoSearch is a unique alerting system; you can put in a keyword and be alerted when literature has been published, or it might be a new podcast or standards relating to that keyword across the literature. It is actually something that doesn't exist anywhere else at the moment. It’s very simple to create an account and very simple to add alerts.

Where do you think it will evolve to? What are your hopes and dreams in the next five to 10 years for it?

That's a good question. I think probably at this point, five years is too far away to see where it's going to be. We want to get people using it, and we will be looking to expand the content types that we're indexing and uncovering. There are also videos being indexed in there from various sources at the moment. We want to listen to the community and develop from there; a practising orthopaedic surgeon who has so many pressures on their time could tell us something really useful but also simple that we could add. We'll take it where the community directs us.

You can try OrthoSearch here and listen to the original podcast recording of this interview here.