In January 2020, HEE commissioned a piece of research to find out about the needs of expert searchers: individuals who carry out frequent and complex searches of the healthcare literature to support clinical practice, research, service development or systematic review. This followed on from a similar project looking at the information needs of end-users (the healthcare workforce) which was carried out in 2019 and which was key to shaping plans for the development of the national discovery system. The aim of the 2020 work was to find out more about expert searchers in the NHS: who are they, what do they do and why, what technologies and systems they use.
Identifying expert searchers
The first step was to recruit the right people into the research. We predicted that a large proportion of expert searching would be carried out by librarians but we also wanted to make sure that the experiences of non-librarian expert searchers were captured. Colleagues from NICE looked at 2019 HDAS user data and extracted a pool of very frequent users based on the number of saved searches. As anticipated, many of these were librarians. Of the non-librarians, 30% were in pharmacy roles, 14% were medical and dental staff, 10% were in specialist nursing roles including practice development nurses and nurse educators, and 3% were in clinical effectiveness and similar roles. To this pool of heavy HDAS users, we added a list of individuals who are known to be expert searchers through their involvement in systematic reviews, evidence synthesis and similar work for the NHS in England. Some of these individuals work in the NHS but are known not to use HDAS while others work outside the NHS, for example in higher education, and also use other interfaces. All these experts were sent a screening survey inviting them to participate in the research.
The research itself was carried out by Lagom Strategy, independent consultants in user experience research. The project included:
- 12 one-to-one interviews with expert searchers
- A diary study in which three participants logged their daily search activities
- Three field visits to observe expert searchers in action
- A workshop to develop the main expert searcher personas and map their user journeys
- A survey, completed by 169 participants, to validate the user needs captured during steps 1-4.
Expert Searcher Workshop
Practitioner Researcher User Journey
Who are expert searchers?
Expert searches are mostly carried out by librarians. This confirms that knowledge specialists play a key role in the delivery of an evidence-based NHS. In addition to librarians, the research revealed two other expert searcher personas: practitioner researchers and research students. Through the workshop, these personas were characterised as follows.
- Searches on behalf of others
- Takes a systematic, planned approach to searching
- Searches across different sources: HDAS, native interfaces, Google Scholar, specialist datasets
- Uses a variety of tools to manage search outputs and record activities including reference management, KnowledgeShare
- Helps other users through training and support
- Juggling studies with full time work and family commitments
- Need to search on the go/mobile with universal logins
- May need resources beyond core content
- Use keywords rather than full syntactic searching but need more guidance about things like wildcards
- Need to record search strategy and results, often using tools like Zotero
- Like open access content and “suggest articles” functionality
- Some protected time for research/practice development activities alongside clinical commitments
- May be linked with local R&D team and may be linked with local higher education institution
- Need to store search results, sometimes using EndNote
- Share results with peers eg journal clubs
- Promoting evidence-based practice and research awareness with peers eg writing and promoting “critically appraised topics”, upskilling colleagues in research methods
What’s different about expert searchers?
Knowing where to look for what kind of information, is a defining characteristic of an expert searcher. They use a wide range of information sources for their searching. Lots of expert searchers in the sample use HDAS (86% of the validation survey respondents report using HDAS) but they also use PubMed, native interfaces, TRIP, Google Scholar, CRD, OpenGrey, Royal College websites, PEDro and so on.
“Some of my skills are centred around knowing where to look for information” – Librarian, interview
Some expert searchers report having go-to resources based on habit, training or experience while others point to following local or professional guidance.
“HDAS was the first one that I learned, and I [felt] more confident with using this” – Masters student and Consultant Psychiatrist, interview
“[UKMI] creates a list of resources that are recommended for all the MI services to use” – Pharmacist, interview
“I’ve got a list basically that I kind of add to, lots of different sources that I can refer to because obviously there are so many, it’s hard to kind of keep track” – Librarian, interview
Because they are more familiar with these different resources than other users, expert searchers are more able to get the best out of them.
“Using the various limits and all the additional features on the databases but using them to the best advantage… knowing when to use them and knowing when not to use them as well” – Librarian, interview
“Particular fields, tapping into the thesaurus, the taxonomy that’s used on each particular database. Being able to collate the results I’m getting back” – Librarian, interview
Expert searchers are more comprehensive in their searching.
“They’re probably just looking for something quite quickly whereas I think we’re looking for, you know a lot more in-depth and finding everything or everything as far as you possibly can” – Librarian, interview
“We are very thorough” – Pharmacist, interview
“We wouldn’t just look in one database, we would look in several” – Librarian, interview
They are also more systematic, sometimes starting with a scoping search then following a planned strategy.
“Started with exploratory search to assess the number and types of publications broadly related to the question and to familiarise myself with the specific terminology, and gather subject headings and keywords and synonyms” – Librarian, diary study
“I tend to try to come up with a bit of a strategy first, before just sort of diving in. So, kind of thinking around keywords and thinking of like different synonyms, kind of building the sort of structure of the search up” – Librarian, interview
Expert searchers need to be able to record their search strategies, histories, de-duplicated results, full text articles and search reports. They are often searching on behalf of or with others and so need to share their search activities and outputs.
“We have a template that can be adapted, that would include details of who has done it, and what they’re asking for, the question they’re asking” – Librarian, interview
“That [sharing information] would be done, most of the time, writing a report at the end” – Research Officer, interview
“What I might have done is print that article and show it to the psychiatrist before the appointment” – Clinical Psychologist, field visit
Some expert searchers support others to improve their searching skills. This isn’t limited to librarians. Support can be informal (as provided by the practitioner researcher mentioned above) or formal training.
“It’s kind of using my own skills and also sort of trying to upskill everybody else” – Librarian, interview
“I actually provide that training in an IT suite, over one-to-one, with a group of people all sitting at a computer and going through it” – Pharmacist, interview
How satisfied are expert searchers?
Overall, expert searchers are fairly satisfied that they have the tools and resources that they need for their work.
HDAS users like the single platform for multiple datasets which is easy to navigate.
“I use HDAS because it’s a really nice, clean interface” – Clinical Psychologist, field visit
Expert searchers reported benefits to using native interfaces as well.
“I think generally they’re [native interfaces] a lot more robust, they never go down” – Librarian, interview
“I think they [native interfaces] have a lot of features that are really beneficial, things like being able to do more complex searches” – Librarian, interview
But there are things which could be improved about all interfaces.
There was a sense that some of HDAS’s functionality had been downgraded over time.
“Previously, you used to be able to run a keyword search across 5 databases, run another keyword search across the same, as long as they were the same 5, you could then combine them with the AND operator within the search panel. That’s no longer available, you have to either search the databases separately or put your entire search strategy into a single search line and then just run it as a one off, you can’t then add any limits or combine any more” – Librarian, interview
There was also some frustration about the way that subject headings are managed in HDAS.
“I had problems with searching for the corresponding subject headings in HDAS Embase (the search facility there is simply inadequate, it would have been much better to have access to proper Embase thesaurus search)” – Librarian, diary study
“It used to be that you could type in a word, search it in the thesaurus, and it would come up with lots of different things relevant to that. But now you have to type in the exact thesaurus term for it to actually come up, and that’s really annoying because usually you don’t necessarily know that the thesaurus term is” – Librarian, interview
“When I’m using Medline, I always use that with the original MeSH browser from the National Library of Health… I found their MeSH browser much much more useful than the one from HDAS” – Pharmacist, interview
“I do go through the process of knowing what my terms are before I even get to HDAS, then having to go through the whole thesaurus process and having to re-input everything… that is a process I’m willing to do” – Pharmacist, interview
“I have to go to PubMed, to their MeSH headings in PubMed to find out, so it’s like an extra step. I find that really frustrating and annoying” – Librarian, interview
Some users reported more confidence in the quality of the searches they carry out in native interfaces
“We end up having to use the native interfaces, because HDAS just can’t cope with it” [systematic reviews] – Librarian, interview
Expert searchers also liked some of the additional features seen in other platforms.
“It would be quite good if HDAS could do something like that, an in built kind of citation matching tool” – Librarian, interview
“You can do like a related article search on Mendeley and that’s deal useful, especially if somebody sends you a paper and says I want things like this” – Librarian, interview
However, searching multiple resources presents its own problems, especially around recording and sharing search histories and results.
“I need an easier way to compile search results from multiple resources as I don’t have access to reference manager software” – Other library and information professional, user needs validation survey
Keeping on top of differences in functionality across multiple interfaces is also seen as a challenge.
“For me there’s always a little bit of a kind of a learning curve of oh right where’s the box to put the search terms in and oh it’s over here and how do I combine this and so on” – Librarian, interview
“You have to know how to use them and we don’t receive training. Well, I’ve never received training in it, I’ve only ever learned how to use them through my own time and effort” – Librarian, interview
“Easier to use help function… it would be nice if you could divide it up [help content] for your simple user versus your advanced user” – Pharmacist, interview
“There are some new features being introduced by them [native interfaces], which the database providers don’t necessarily explain very well” – Librarian, interview
What happens next?
As well as the rich learning from the qualitative phase of the study which has been summarised above, the research has resulted in a validated “user story backlog” or prioritised list of requirements for expert searchers.
Expert Searcher User Story Backlog
These will all be used by HEE to inform the development of the discovery ecosystem, ensuring that the needs of expert searchers are met or exceeded as new services are introduced.
For more information on this work, contact Lucy Reid.