#eahil19 – reflections: Reduce / Reuse / Recycle / Review / Respect / Report / Robust

July 3, 2019

Reduce / Reuse / Recycle / Review / Respect / Report / Robust

These are the themes that came through to me during and since the EAHIL 2019 workshop in beautiful Basel.

It’s always great to meet colleagues at conference, and to take time out to think about particular topics and challenges with them. I particularly love EAHIL for giving that space. Check out all the tweets at #eahil19.

inspired by the Rhine, brace yourself – this is a long one……

Basel & the Rhine

Reduce / Reuse / Recycle

The amount of waste in research is a classic quote now. Where we, as librarians, are involved in waste in terms of how much time we spend, creating search strategies from scratch, rather than re-using/adapting existing strategies.

Now there’s something of a circular argument, that research that is reported badly probably wont have the search strategy included which means that it’s not available to reuse. Fair point, and one I’ll come back to.

I’m sure most of us do try to find Cochrane and other reviews which overlap with our topic, and we make use of portions of the strategies which are reported there. And there are other ways to solve this problem: Jane Falconer from LSHTM uses her institutional repository, loading up the search strategies. This has the double benefit of not having to rely on appendices in online articles which might be behind a paywall, and also gives the strategies DOIs of their own. She’s even been so fabulous as to create a guide as to how to do this! What a star!

There are also other banks of search strategies, created by Dutch colleagues and an EAHIL site too. This might raise the question of credit – at what point is a search strategy intellectual property which should be cited in the paper, how many tweeks makes it your own? I reckon that a search filter used without tweeks should be cited.

On this note, the quality of search filters was discussed, at a session facilitated by Alison Bethel and Morwenna Rogers, in particular how difficult it was to develop a filter for qualitative studies. What we noticed in particular was how different the quality of indexing for qualitative studies was between Medline and CINAHL, enough to influence  @wichor‘s future practice:

I also learned (via Lina Gulhane’s session and more recently tweets from @srobalino) about OVID Search Launcher – so long as you remove the numbers from your strategy, you can upload a massive long search string into the OVID interface without lots of copy/paste! Wonderful (though it’s a bit of a drag having to remove the numbers…) @srobalino’s twitter thread also has a suggestion about using OVID jumpstart – suggested by @v_woolf, which I’ll need to play with. I think that @v_woolf‘s blog will be required reading from now on, for expert hints and tips like these.

Add to this list of new toys is SR-Accelerator, and the polyglot section, which can help with translation of strategies between databases.

Review

The issue of librarians involved in peer review was raised at EAHIL in Seville in 2016, and was the subject of much debate in Basel. Ideally poor quality work should be improved much earlier in the research life cycle than at point of publication, but if librarians were more involved at the peer review stage this might help some flawed papers making it into print. Watch the combined EAHIL/MLA/CHLA/UHMLG space for some action on this front.

I know there are some librarians who already are regularly involved in peer review – Wichor Bramer and Dean Giustini (see Dean’s thread about open peer review) are just 2 examples, and I’ve tried it once or twice myself. Personally, I had to take a very deep breath before I did it, but I focused on the bits I knew, and ignored the other bits.

But wouldn’t it be so much better if the quality of the search strategy, and search methodology was improved before screening took place? This is where PRESS comes in, and I enjoyed another very engaging session with Alison Bethel and Morwenna Rogers, where we worked through a PRESS checklist. Another set of eyes to check over a strategy is always valuable, spotting gaps and typos, questioning the decisions we’ve made about databases etc.

My colleagues in the Medical Library and I are going to try to be more diligent about doing this in future (we’ve tried it once or twice, more as a training exercise than anything else – Note to self- could this be used in teaching with students?? ) And for those without willing colleagues close by, there’s the PRESS-Forum which you can register for, to ask the community in the forum to review your searches.

But what I particularly liked was the suggestion that the fact that the strategy had been PRESS’d could go in the methods section, as evidence of rigour. Since wordcount is always a problem, perhaps a short narrative could go in to the appendix to explain some of the decision making around the strategy. This could help with replication too. Nice!

Respect

Sandy Campbell was very clear: “We don’t work for acknowledgements”. The issue of respect for the contribution that librarians make to systematic reviews, and their right to be co-authors was discussed in several sessions. I really liked the form that Sandy and her team uses at University of Alberta, to set out right at the start the expectations of workload.

Librarian as co-author?

They also list explicitly on their libguide “who is an author” – just to make it absolutely clear.

If you don’t ask, you don’t get, and I will be reviewing local practice (ie nicking that form!)

I also had a great conversation with @sandyiverson about her teams recording of “billable hours”. Even if you don’t charge for a service, it can be a powerful tool to say how long a piece of work actually took (and possibly add a ££ which will not be charged). We do a lot “for free”, but there is a cost.

This also raises the issue of how you set up a systematic review service – what are the ground rules? what’s a free service, what’s not? I wasn’t able to attend Hannah Ewald’s session, but it’s definitely something I need to explore further.

Report

The fabulous team at KSR, Caro and Shelley, gave a great session on how librarians can improve the conduct and reporting of systematic reviews. PRISMA and PRISMA-P, and, still in draft, PRISMA-S were all part of the conversation. If methods aren’t reported well, then that raises doubts about the validity of the findings.

We should definitely use our place as co-author to comment on the whole paper that our name is attached to before it is published. We could even acknowledge the limitations (doesn’t every paper have some limitations?) – if not in the methods, perhaps in the discussion section? or in the narrative in the appendix that I mentioned earlier.

In the session we used ROBIS, the Risk of Bias tool, to evaluate papers. As co-authors, we could advocate (and use ourselves) this tool before submitting for publication, to help us see the research as others might see it, and maybe spot gaps/flaws before it’s sent for peer review.

Robust

All these steps make research more robust, and therefore more worthwhile.

Better research = less waste. What’s not to like?

logo for eahil 2019


grey literature – what a ‘mare!

March 20, 2019

I attended an HEE course on grey literature last week, led by Jo Hooper.

Now I’ve always thought that searching the grey literature was a bit of a nightmare – and I’m afraid that I didn’t get any reassurance from the course. It’s pretty much guaranteed to be much more time consuming than searching standard bibliographic databases. But….

I did get a useful reminder of why it was useful to include (the Tamiflu expose (read here and here), and  Reboxetine) and I did get a fresh look at what Grey Literature actually was.

Key descriptions that stick in my mind were that grey literature was “anything that can’t stand up by itself on a shelf”, referring to the pamphlet, slim report, leaflet that makes up a significant proportion of grey lit, and the fact that it might be online, and/or in another format – eg spreadsheets, etc.

I also liked the idea that “grey literature is ahead of curve of changes in how we communicate” This reminds us (me) that social media and discussion forums, and email discussion lists are also included in the mix. This means that the definition of what is grey and what is not, is going to continue to change over time.

But I also got some direction to further reading by the always excellent Simon Briscoe, Claire Stansfield, Alison Bethel, Morwenna Rogers and others about how to record the methodology of searching and including grey literature in a systematic review – really practical, useful advice.

Lots of links to further reading is in the guide I drafted.

PS – I shared this guide with colleagues and immediately got extra links which improve it. V2 is now available! (brace for further changes when more of my omissions become apparent!)

examples of grey literature

Source: UIC Library

 

 


Evidence for Global and Disaster Health – follow up

September 18, 2018

The papers and presentations from the IFLA Special Interest Group – “Evidence for Global & Disaster Health” – have just gone live.

1


Evidence for Global and Disaster Health

August 27, 2018

Evidence for Global and Disaster Health

What does this mean to you?

If you’re reading this, you’re probably some sort of information professional, so the evidence bit is probably pretty straight forward – research literature which tests the effectiveness of interventions, be they diagnostic tests, or treatments, or predictions of prognosis.

The disaster health is probably pretty obvious too, but what about global health.

The Sendai Framework for Disaster Risk Reduction identifies 12 macro level types of threat:

·         Financial shock

·         Trade disputes

·         Geopolitical conflict

·         Political violence

·         Natural catastrophe

·         Climatic catastrophe

·      Environmental catastrophe

·      Technological catastrophe

·      Disease outbreak

·      Humanitarian crisis

·      Externality

·      Other shock

sendai

Yes, externality (the impact of a meteor hit, or a sun spot) might seem a little bit Hollywood disaster movie, but did you include the consequences of financial shocks, or trade? Yet the rise of poverty related illness because of austerity in the UK will attest to the relevance of this as a category.

I attended the inaugural meeting of the IFLA special interest group on Global and Disaster Health. It’s the culmination of tireless work by Shane Godbolt, Anne Brice, and many others. That it happened at all is only the start, and yet it’s very much a sign of how much work has gone one, and continues to go on in this area.

e4gdh

Aim of the E4GDH SIG

To explore, map and strengthen the potential for librarians and information specialists, and their services, to play an enhanced, pivotal role in the collation, organisation, assessment and deployment of information concerning global and disaster health including disaster preparedness and risk reduction.

All the slides will be going up on the E4GDH site shortly, but you may also want to have a quick browse through the tweets: #e4gdh

I knew that as information professionals there was the potential in us all to contribute through the voluntary work I do with Evidence Aid (read more about that here), but the variety of opportunities was a surprise as exemplified by Bethany McGowan. She talked about Using GIS Data and Mapping Parties to Expedite Disaster Relief Response to Vulnerable Places: just as you might contribute to Wikipedia to improve the knowledge of a particular topic, you can make maps better and thereby aid relief efforts. Amazing! Read more here and here

“Humanitarian mapping activities combine open data and crowdsourcing to support disaster relief response and humanitarian aid to vulnerable populations.”

Merlita M. Opeña talked about her work: Philippine Research on Disaster Risk Reduction in Health: Mapping a Research Roadmap and Creating a Framework for Information Sharing Nationally and Globally. She’s working to help improve health resiliency – one of the measures of how bad a disaster is, is how long it takes the community to return to “normal”, all the more reason to help them be better prepared to withstand a disaster. But this needs research, and evidence, and this is where Merlita and her colleagues come in.

I discovered that the language of disaster health is still evolving so research about research is vital to better clarify and unify everyone’s understanding: Diana Wong’s work in Disaster Metrics: A Comprehensive Framework for Disaster Evaluation Typologies achieves exactly that. “This is a framework outlining the different types of evaluation types that can be used in the disaster setting. It will help provide consistency and structure in disaster evaluations and improve the science of disaster health.” Read more here.

The role of librarians in the development of policy is crucial in so many ways, but the work of Prof Daisy Selematsela and Blessing Mawire in South African exemplified this: “The transitionary role of Research and University Librarians/Knowledge Specialists in developing countries in facilitating transformation for sustainable development.” Their work is influencing the way in which the South African government is working to achieve UN Sustainable Development Goals, particularly #3 and #4.

When disaster strikes closer to home – as Feili Tu-Keefner experienced with 2 hurricanes ravaging South Carolina in almost as many years – it’s crucial that public librarians understand how best they can support their communities in times of crisis. Feili has investigated “Lessons Learned After a Disaster: Investigations of Public Librarians’ Health Information Services to the Community and Community Members’ Information Needs Following a Catastrophic Flood”.

I learned so much from Caroline de Brun and Blessing Mawire in their practical session raising awareness of sources of high quality information – their padlet of resources will be part of any future teaching and searching I do. Particular favourites were medbox, and 3ie and openwho, but I’ll also be reading more about the PARIHs framework. All details at:  www.padlet.com/caroline_debrun/e4gdh

vm

Now I’m starting to believe Prof Virginia Murray, when she called libraries “a secondary emergency service”, and I am proud to have been part of the event, and to be involved in the work.

IMG_3489.JPG

 


#hlg2018 – Health Libraries Group, Keele

July 1, 2018

For the first time in 10 years I attended the Health Libraries Group conference. I contributed to the programme in 3 ways:

There was a packed programme, with great speakers. It was great to hear Dr Mark Murphy in EAHIL last year, and he was equally stimulating this year at HLG (you can read a paper of his presentation).

It’s a really important aspect of any conference that it gives an opportunity to meet up with colleagues, and I had the pleasure of chairing 2 sessions too, which is always fun.

There were 2 standout presentations, one of which had been inspired by a presentation at EAHIL 2016, which I was lucky enough to also attend. Alicia F. Gómez-Sánchez,  Mar González-Cantalejo, Gaétan Kerdelhué, Pablo Iriarte and Rebeca Isabel-Gómez presented their work assessing the quality of reporting of a certain subset of systematic reviews. It’s great reading, if you like reading about just how badly reported most systematic reviews are.

Jane Falconer referenced this paper when she reported on her own survey of the systematic reviews produced by her own institution, London School of Hygiene and Tropical Medicine.

Key points from her talk

  • great respect to LSHTM for being willing to share the fact that Jane found that a large proportion of their publications failed aspects of assessment on the basis of PRESS, PRISMA and AMSTAR criteria.
  • the frustration that all the teaching and guidance that Jane had provided over the years still hadn’t resulted in better outcomes in this audit (though of course we don’t know how much more poorly papers would be reported without her influence)
  • that by carrying out this audit and presenting the results to the leadership at LSHTM, the profile of the library was raised.

If the statistic that 85% of research is wasted is still true, then the cost of poorly reported SRs must surely contribute massively to this, and peer reviewers must take a portion of the responsibility for this – how do these papers get accepted?

I was already looking forward to hearing Kate Miso‘s paper, but after Jane’s I was really fired up!

Echoing aspects of that blog by Iain Chalmers and Paul Glasziou, Kate quoted Doug Altman:

doug_altman

Again, there was a sorry tale of school-boy errors in the reporting and a catalogue of other increasingly tooth-grinding errors. They culminate in significant errors which would have consequences for the interpretation of the results. Any poor method will bias the results, and any bias unrecognised will result in problems if the results are applied in practice.

One of the take-homes for me, was that if the review doesn’t actually have “systematic review” in the title, then start looking for more flaws.

Kate was kind enough to share her slides with me after the event, but clearly they’re not for me to share here.

As with Jane’s presentation, I have to wonder about doing an audit of local SRs in a bid to check that my own house is in order, and I’m also minded to start a process of local PRESS activity with reviews that my librarian colleagues and I produce to ensure that the strategies we put together in collaboration with researchers are the best they can be. Watch this space for how both these projects progress.

Librarians can make a difference to improve reporting SRs, but perhaps we also need to do some work to give our colleagues a wake-up call that they need some help.

 

 


Decision making and bias

July 9, 2017

Last week I went to  ‘The Art & Science of Clinical Problem-Solving’  by Professor Sanjay Saint. It was a great presentation, interactive and engaging, and despite (as well as?) being pretty much the only non-clinician in the audience I learned a great deal.

the_cognitive_bias_codex_-_1802b_biases2c_designed_by_john_manoogian_iii_28jm329

While I’m not involved in any clinical decision making, there’ plenty of decision making that goes with my job, and even more that goes with life – and insight into the amount of bias, conscious or not, that comes with each decision, and the process of coming to a decision can be a huge advantage.

Now I wont go into all the possible forms – there’s Wikipedia, postgraduate qualifications and books aplenty on this. The forms that Prof Saint focused on were:

  • availability bias
    when we just don’t have enough information, yet we might make a decision/come to a conclusion anyway. Sometimes this is because we’re feeling pressure to take action, but it might be because we’ve already decided what we want to do and just want to get on with it, despite subsequent information perhaps warning against it.
  • anchoring bias:
    that we pay too much attention to one piece of information (likely to be the first piece of information we receive), and then this skews our decision making from here on. (this is Blink by Malcolm Gladwell in a nutshell)
  • confirmation bias
    when we only look at new pieces of information as evidence to confirm our first conclusion.
  • premature closure
    that when we come to a conclusion we stop hearing new information, which might otherwise require us to change position, or to at least reconsider our first conclusion.
  • framing effect
    This is how we come to a conclusion based on how the information is presented, rather than on what the information is.
    We can use this to our advantage (particularly when we want someone to do something they might be otherwise reluctant to do), but you have to know what will ring the other persons bell if you want it to work.

I’m afraid to say that I know myself well enough to know that I often fall into the trap of several of these – and that I don’t spot (or ignore)  at the time and fail to work harder against my natural tendencies….. but maybe with the next decision I make it will be different, eh?


#uhmlg17 – Happy birthday UHMLG – here’s to the next 10 years

July 9, 2017

A tenth anniversary for any organisation is quite a milestone, and not least for UHMLG – University Health and Medical Librarians Group. I’m lucky to have been on the committee for the past year, and thoroughly enjoyed putting together this years programme with my colleagues.

All the slides from the event, and a bit more detail about the even are available on the UHMLG blog, so here’s some of what I considered my highlights/take-homes from the event.

Alison Day spoke about her leadership course. I thought she introduced a tremendous discussion topic, by asking us to have a super-speedy knowledge cafe on “what makes a good follower” – great to have the topic turned on its head, and a useful reminder that a good leader needn’t be in a management position. Nor does being in a management position automatically make you a leader – plenty to think about here.

There were several standout moments for me from Adam Young and Taryn Jackson’s talk about what primary school children are taught. The level of computational thinking that they encourage was fascinating, and perhaps needs to be taught retrospectively to those who’s primary school years are long behind them. Perhaps the whole concept of “process fixing”, or process mapping means at least some of these skills are ingrained already, but I don’t think it would do any harm for me to seek out some further training in this area.

Amy Icke gave me a light-bulb moment when she said that her secondary school students thought that books carried more weight (ahem!) than journal articles: that they didn’t really understand the difference, and considered “magazines” as being of less value. This  prompted me to consider the starting point of the dissertation sessions we run for our undergraduate students – who’ve primarily only had book-based reading to do up till that point. We march in, talking about pubmed and searching, and referencing, but should maybe start from the point of “why should you be looking for articles as well as/instead of books?”, “what’s the difference?”, etc.

We also got a great example of how a lecturer has embraced a new way of teaching their students, when Tim Vincent introduced us to the way that one of the lecturer’s has completely transformed the way he delivers his teaching – by using interactive tools, by chunking up the presentations into shorter blocks, and using video. Tim included lots of different tools into his own presentation, and the possibilities are really exciting. I used Padlet recently to get feedback from a lecture hall full of students who were working in small groups, and I think it’s something I’ll be exploring more.

Tim also pointed us to a short video that I think should be compulsory for all presenters – 5 Things Every Presenter Needs To Know About People

Do watch it, it’s a really useful reminder of stuff we know, but sometimes don’t put into practice.

Finally, I was really excited by our first “peer assist” – well at least a version of this knowledge management technique. I asked everyone to bring along a challenge that they were facing (anonymised to protect the innocent), and then shared them out to get the hive mind to suggest solutions. While we’ve not shared this on the UHMLG blog, the collected problems and proposed solutions were shared amongst all attendees, and I found it fascinating as a process. This knowledge management malarky is actually pretty useful!

There was plenty more I could say, but mostly I’d say join us! If you’re a health / medical librarian working in HE in the UK, please join us. We have a spring forum: 23rd March 2018, and a summer residential event, and together with the email discussion list, they’re all great ways of getting together with HE colleagues to discuss common problems and topics that are interesting to us all.

It’s free to join, and I’ve got a lot from all the events I’ve attended, as well as from the (continuing) experience of being on the committee with a great bunch of colleagues. Happy birthday to us!

Hope to see you at the next UHMLG event.