Troubleshooting

issues with your searching.

How many results are enough?

This of course depends on the search subject, and the requester. There is no perfect number.

Can you glean from your requester how many results they would like to receive? KnowledgeShare’s online search request forms have a field for requesters to state an amount (but it is not a mandatory/required field). If you meet their number by adding limits or any contrived means, make sure you tell them what you have done and make it clear your results are not comprehensive. Offer to rerun the search without the contrivances.

You may not be able to meet your requester’s expectations if they have unrealistic ideas of what is out there on a topic, but they can help you know whether to expect a high yield or not.

Some requesters will be happy to sift through hundreds of results, others will not. If you can establish a dialogue with the requester, and establish that the searching process is iterative, this can help manage their expectations.

If you are confident that your search strategy is robust, then your yield just reflects the body of evidence, whether you have “too much” or “too little”, but you can adjust your returns using the methods below.

What if my search is retrieving too little, or nothing at all?

Do a quick sense check. Have you…

Asked the requester if they have had a look themselves already? What did they find? Would they expect there to be so little out there?
Spelled your text terms correctly?
Used text searches as well as subject headingg searches?
Used appropriate subject headings? Check the scope notes for the headings you have used.
Exploded all your subject heading searches?
Combined searches in the correct way?
Checked that there are no limits on your search that you did not mean to impose (i.e., full text only)?
Imposed limits on your search? If so, can you relax/remove them within your requester’s criteria?
Looked everywhere? Are there more sources you can search?

Can you remove phrase-searches or remove/widen adjacency operators to make your search more sensitive?

Have you included regional variations in terminology amongst your terms, e.g. physiotherapy (UK) as well as “physical therapy” (US)?

Have you included UK and US spellings of text terms, e.g. haemorrhag* as well as hemorrhag*? Most databases allow the use of special characters to skip letters and cover both spellings, e.g. h?emorrhag*, but the special character used for this varies between databases.

Have you truncated text searches wherever possible? i.e., could you go from using “testicular cancer” to using testic* adj3 cancer* to using testic* cancer*?

Can you introduce more synonyms? i.e., (teste* or testic* or scrot*) and (cancer* or neoplasm* or oncolog* or tumor* or tumour*)

Can you search within all fields of records, rather than just the title and abstract fields?

Can you search a level above your current subject heading level, thus searching more broadly, without retrieving too much irrelevant material?

If you have searched using a broad subject heading combined with another subject heading, could you try using a subheading with that subject heading, or vice-versa? i.e., if you have searched using the subject headings exp "TESTICULAR NEOPLASMS"/ AND exp DIAGNOSIS/, have you also tried exp "TESTICULAR NEOPLASMS"/di?

Does a prospective or retrospective citation search retrieve anything?

Can you (with your requester’s input) broaden the focus of the search and retrieve more general results? Or can you search within the text of more general results? Are there chapters or sections within books/ebooks with a broader focus?

Could it be that you are searching within an area where there has just not been very much research published? If so, can you encourage your requester to conduct some (using library services in the process…)

Consult colleagues locally, or with a mailing list enquiry (see below).

Whichever of the methods you use, be sure to inform your requester what you have done and why. What if my search is retrieving too much?

What if my search is retrieving too much?

Do a quick sense check. Have you…

Used appropriate subject headings?
Combined searches in the correct way?
Used appropriate text terms?
Truncated text terms in the right places, or has a term truncated at the wrong point opened the floodgates? For example, if you were looking for testicular evidence, testic* would be appropriate truncation, while test* would retrieve a huge number of irrelevant results.

It could be that you are just searching within an area where there is a large body of research evidence published. If your search topic is broad, or you are searching within an area that is rich in research, can you (with your requester) narrow the subject down?

Can you increase the specificity of your search using limits, justifiably? i.e., can you add date limits, and/or population limits, and/or language limits to the search? If you have already applied some, can you limit these further? i.e., changing your date limit from the last 10 years to the last 5 years?

Are any of your text terms too “noisy” and retrieving irrelevant results? i.e. common terms like exam* or clinical* etc., or terms with synonyms unrelated to your subject. You can test your truncated terms with a crossword solving website like https://www.wordhippo.com/: select “Words Starting With” and enter your truncated term to see what words your truncation is returning.

Could you use adjacency operators or phrase-searching to make your text searches more specific without being too specific? i.e., could you go from using testic* cancer* to using testic* adj3 cancer* or “testicular cancer”?

Can you justify searching for your terms within just the title fields of records, rather than the title and abstract fields?

Could you search without exploding subject headings? Could you search using only major subject headings? Could you search using subheadings with your subject headings, to home in on certain aspects of your topic?

Could you limit your search to just high tier evidence, and retrieve only reviews/meta-analysis and other high tier evidence for your requester?

If you are searching Embase as well as Medline, consider using the “remove MEDLINE records” limit when you search Embase.

Consider searching using focused exploded subject headings within your Embase search too (rather than just exploded subject headings – though this is not appropriate when undertaken searches for a Systematic Review).

Consult colleagues locally or with a mailing list enquiry (see below).

Whichever of the methods you use, be sure to inform your requester what you have done and why.

Checking the rigour of your search

If you have a conducted a search of Medline, try a search of PubMed. If your strategy is robust, in theory you should not then retrieve anything relevant from PubMed that your Medline search did not retrieve.

Similarly, consider a search of the KLHub at the end of your search process as a way of double-checking you have not missed anything potentially relevant.

Have a look at the Peer Review of Electronic Search Strategies (PRESS) 2015 Guideline Evidence-Based Checklist at table 1 here. Though it is intended to guide peer review, it is a useful aide-memoire for searching. Have you followed all the appropriate steps it outlines?

Test your search strategy: identify key papers (by asking your requester to nominate some if they did not already, or by searching Google Scholar and choosing the top few papers). See if your search has returned these key papers by searching for their titles using ctrl+f within your evidence search report, or within the references you’ve compiled in Refworks or other software. It should appear amongst your results if your strategy is comprehensive.

Alternatively, once you’ve identified a key paper try entering its full title as a title field only search within one of the databases you searched. Combine this search line with the final line of your search strategy in that same database using the AND operator: if your strategy picked-up this key paper you will retrieve only one result, if your strategy did not pick-up that key paper you will retrieve no results.

Local groups of colleagues that you can contact for help

The East of England’s Searching and Training Special Interest Group meets quarterly online, contact chair Sophie Easey on [email protected].

The Kent Surrey Sussex Searching and Training Forum (KSS STF) meets twice a year online, usually in February and September. For more information, contact the chair Tom Kelly on [email protected].

The London Searching and Training Forum (LSTF) meets quarterly online. For more information contact chair Adam Tocock on [email protected].

The North East London Area (NELA) Searchers group meets infrequently online to informally discuss searches and share best practice, with a rotating chair/host. Contact chair James Hurst on [email protected].

The Midlands Regional Trainers Forum is chaired by Kate Jones on [email protected] or committee member Derick Yates on [email protected].

In the North East, Health Libraries North (HLN) can be contacted at [email protected]

In the North West, The Library and Information Health Northwest Network (LIHNN) meet regularly. There is a charge for memberships (organisational or individual). Please see for more information, or contact chair Susan Smith on [email protected]. The LIHNN website also hosts the Library Knowledge Services (North) Current Awareness directory.

A sub-group the Clinical Librarians, Trainers and Searchers Group also meets. Contact Olivia Schaff on [email protected].

In the South West, the SWIMS Searchers and Trainers Forum (SWIMS STF) meets quarterly online. For more information, contact the chair Sarah Rudd ([email protected]) and vice-chair Caroline Green on [email protected].

The Yorkshire and Humber Health Libraries and Knowledge Network (YOHHLNet) is chaired by Heather Steele, contact [email protected], Heather Steele on [email protected], or follow their Twitter handle.

The Health Library and Information Services Directory (HLISD, often pronounced as “health Lizard”) allows you to look up libraries and staff by area, Trust/organisation, or county.

Other help available

A literature searching peer review buddy scheme is run in the London, Kent Surrey Sussex, and South West regions, pairing searchers so they can informally discuss a search they each submit. It is open to colleagues from all regions. For more information or to join the scheme, please contact the organiser Caroline Green on [email protected], She is an outreach librarian at Royal Devon University Hospital.

A very informal Small NHS KLS Support Group exists to help library staff who work as part of small teams within (or adjacent to) the NHS. For more information contact Adam Tocock on [email protected].

The Mental Health Library Network is an informal and safe peer-led space for librarians working across mental health and learning disabilities specialties that meets bi-monthly. For more information contact Federica Bianchini on [email protected]

A Systematic Review Searchers Community of Practice exists for those performing searches for systematic reviews. Search for the email address [email protected], at Jiscmail and request access. Please contact member Sarah Rudd if you experience difficulties joining the community: [email protected].

FutureNHS is an online collaboration platform available to all NHS staff. When you register for a free account you can search for communities of practice to join using the “Find a Workspace” option from the dashboard.

There are many of these workspaces/groups for knowledge and library services staff, including:

The combined workspace for the Searching and Training Forums of London (LSTF), the South East and (KSS STF), and the South West (SWIMS STF).
Primary Care Knowledge Specialists
Sustainability in NHS Knowledge and Library Services

Mailing list enquiries

If you are struggling to get to grips with a search, if you cannot find any results or retrieve too many, you can consult national mailing lists of librarians who will give you suggestions.

Visit Jiscmail and create an account using your nhs.net or work email address.

Subscribe to the LIS-MEDICAL and CLIN-LIB distribution lists, the LIS-PUBLICHEALTH for public health searching queries, and any others relevant to you, by searching for them by title using advanced search.

Once you are subscribed you can send requests for help with tricky searches to hundreds of colleagues by emailing your queries to [title of distribution list]@JISCMAIL.com etc. (You will need to set up rules/filters routing other members’ emails to these to lists into a folder your email client, or your inbox will get too crowded).

If you would like to access a paper or resource that your institution does not have access to, you can try emailing the HEALTHILL list with the details.

If HEALTHILL cannot provide you with the paper you require, consider emailing the corresponding author directly to request a manuscript copy.

You can also consider directing your requester to the author or the paper on ResearchGate, an online community of researchers sharing their work (do not share material from this site directly as this is against copyright legislation).

Subscribe to other regional and national mailing lists to send/receive emails to 100s of librarians in your region.

Contact the NHSE Knowledge for Healthcare team on [email protected]. They can give you access to the appropriate knowledge for healthcare regional mailing list:

When requesters don’t respond to your query emails!

If queries arise when you are searching that require your requester’s input, but your requester does not respond to your emails or calls, you are well within your rights to pause the search until you hear back, as long as you have made every effort to get in touch and progress is prevented by their silence.

You could send the requester a small sample set of initial results and state that you will not search further until you hear from them.

If you perceive there is an urgency to the search request, you could make an educated guess about how to develop the search. Document the assumptions you have made when sending the results, and offer to re-run the search to focus it differently if needed.

Remember that clinicians have limited access to computers, and while some make regular use of their personal mobile phone, not all can or do, so they may not be able to reply to your messages.

Dealing with technical difficulties If you experience performance issues when using bibliographic databases, check if there is a local network issue within your Trust via colleagues, your Trust intranet, or alert emails from your Trust IT department.

If there are no local network issues, check to see if there is a wider issue at the providers’ end via any notices on the database’s homepage, or on the homepage of the Knowledge Hub Service Desk at or on the Content mailing list. If there is a wider problem, that group will let you know.

If there are no wider issues, check the device you are using: are other programs or websites working? Can you change browser, or change PC/device? “Have you tried turning it off and on again?”

If you do not see your issue reported, report it to the KLHub Service Desk and to the Content mailing list.

If technical difficulties are preventing you from completing a search then let the requester know; anyone and everyone who has to contend with NHS IT provision will likely sympathise.

Dealing with technical difficulties

If you experience performance issues when using bibliographic databases, check if there is a local network issue within your Trust via colleagues, your Trust intranet, or alert emails from your Trust IT department.

If there are no local network issues, check to see if there is a wider issue at the providers’ end via:

any notices on the database’s homepage
KLHub Service Desk
Content discussion list - if there is a wider problem, that group will let you know

Consider Down for Eveyyone or Just Me to see if the site/resource you are looking at is down for everyone, or just you.