EAHIL 2018 – Take away messages

I attended the EAHIL Conference held in the Royal Welsh College of Music & Drama from 11-13 July. EAHIL is the European Association for Health Information and Libraries and it was great to meet and learn from health librarians from across Europe and further afield.

I was presenting in the first set of parallel sessions, which was great as it meant my nerves didn’t have a chance to catch hold and I had a way in to network with strangers (not my strong point) for the rest of the conference. I was presenting in the Richard Burton Lecture Theatre (intimidating or what!?) with Irma Klerings from Cochrane Austria. Irma and colleagues looked at the impact of searching in fewer databases in rapid reviews and compared outcomes found from a rapid or a full systematic review. Her three conclusions were:

  • If decision makers are willing to accept less certainty and a small risk for opposite conclusions, some abbreviated searches are viable options for rapid evidence synthesis.
  • Decisions demanding high certainty require comprehensive searches
  • Impact of abbreviated searches depends on type of intervention, “size” of the topic, and definition of “changed conclusion”

She also found that limiting number of databases was more suitable for reviews of pharmacological interventions and public health topics were more liable to change conclusions if less databases were searched – very pertinent information for us at the School.

As was also drawn out in the Q&A afterwards…

Her study is just published, see Nussbaumer-Streit, B., et al. Abbreviated literature searches were viable alternatives to comprehensive searches: a meta-epidemiological study. J Clin Epidemiol 2018; 102: 1-11. doi: 10.1016/j.jclinepi.2018.05.022.

Then it was my turn. I presented on work myself and Kim Coles did, looking at the quality and reporting standards of systematic reviews published by School authors.

What Alicia fails to mention, is that our work was inspired by work she presented at EAHIL 2016. It was great to meet her and compare notes.

As you can see from Amanda’s tweet, the stage and my slides were huge. The lecture theatre was also really full, which was terrifying and pleasing in equal measure.

Another advantage of being in the first parallel session of the day is that

As the third presenter pulled out due to illness, we ended up having 30 mins of Q&A after my session, which prompted a lively discussion about how we, as librarians, could improve quality and reporting of search strategies – in the venue and also on twitter.

Thank you to everyone who asked questions or spoke to me afterwards, i’m pleased so many delegates found my presentation so interesting.

After lunch I listened to Caroline de Brun from Public Health England talk about information needs of health professionals working in humanitarian settings. She carried out work with HIFA (Health Information For All), which is published on the HIFA website (click on the publications tab). Ruth managed to capture a useful summary slide.

This will provide useful background to the ongoing discussions we are having in the Library with Public Health England around how we can support the UK Rapid Response Team. Caroline also mentioned a resource which was new to me: Medbox. I’ll have to check it out and maybe add it to our list of databases and literature sources.

Next up was Sarah Young from Carnegie Mellon Univ talking about how she supports systematic review capacity building in LMICs.

Although her work was centered in LMICs, she made the good point that resource limited settings can be in any country if a researcher doesn’t have access to a well stocked library. Researchers in LMICs often need to write their own systematic reviews as most are not suitable for their situation, although Sarah highlighted that Cochrane and the EPPI centre are now including low-resource settings in their reviews, which is excellent news.

Sarah and colleagues visited researchers in Africa (sorry, I didn’t note down where) and conducted face-to-face training over five afternoons. Not surprisingly, they found this wasn’t long enough.

She ended her talk by calling for a mentorship scheme for librarians in LMICs, something i’m interested in setting up in the UK.

At the end of the day I gatecrashed the Public Health Special Interest Group meeting where I discovered PubMed is going through a complete overhaul of the backend.

Day two started with a wellbeing morning. I opted to visit Cardiff University special collections and was fascinated by an old European atlas – in particular the map of Scotland.

But no rest for the wicked, after lunch we were back to the conference proper.

Alicia Gomez-Sanchez and Rebeca Isabel-Gomez gave a presentation called ‘Rapid Reviews drive us crazy!’ They tried to find if there was any consensus in recommendations on how to do one. As you can see from the photo in the tweet below, the answer was ‘no’.

Their conclusions show that this field is a bit of a methodological mess.

Next up was – Floriane Muller & Pablo Iriarte who tried to determine how much of PubMed was held in full-text by their library and how much was available open access. To do this, they downloaded PubMed – yes, you read that right – and used PMIDs to compare their local holdings and open access holdings.

To be honest, both were a higher proportion than I thought.

The final presentation of the session was Andrew Booth asking ‘How many search results and enough… and what can we do about it?’

I found this really thought provoking as it tied into questions I was asking in my presentation.

His argument was that as well as being complete, we need to be efficient and came up with his own metric – number needed to read- to measure this. This is the total number of items screened divided by the number included in the review. He argued that as there is no benchmarks for searching, we all do it slightly differently. Therefore the individual/institution carrying out the search has as big an impact to the results retrieved as the topic.

He found huge variation in the number needed to screen, from 3306 (!!!) to 7, and that systematic reviews with an information professional doing the searching typically ends up with 3x more results to screen.

Unsurprisingly to me, he found that public health reviews had much more references to screen on average, and a higher number needed to read than on average. Put that together with the findings from Irma on day one, where she found that public health reviews were more likely to draw inaccurate conclusions if a rapid review methodology is used, and we can see where there may be issues.

His conclusions give food for thought.

Note, that if it’s filled in correctly, the PRISMA Diagram should document this process.

Friday started with a morning Wikipedia editathon. This is something i’ve always fancied having a go at, but have been a bit hesitant to try. It’s really easy! Thinking of putting together something at LSHTM now.

Here’s Ruth from Edinburgh Uni telling us why they decided to introduce it there.

And here’s me having a go.

Finally, I went to hear Anne Brice talk about Knowledge Management in Global & Disaster Health. This is an area we’re increasingly involved in, as noted above, so I found this really interesting. Sendai has now made it onto my ‘to read’ list.

The conference ended with the announcement that my presentation won Best Presentation for a first time attendee.

I was a bit flabergasted and rather chuffed. Thank you to the EAHIL organisers for voting for me and for the prize.

As I tweeted at the end…

A round up of blog posts is being published and will be added to as they become available.

All the slides etc will be available on the EAHIL website next week.