Debbie Spencer, Academic Liaison Librarian for Nursing and Allied Health, University of Chester
Introduction to the project
In December 2023, one of our research professors approached me to ask if I would help with a search strategy on a project he was involved with. His team had just been awarded a grant from the United Kingdom Health Security Agency (UKHSA) to conduct a rapid review on the ‘unintended consequences experienced by inclusion health groups in institutional settings due to the implementation of non-pharmaceutical interventions (NPIs) during the COVID-19 pandemic in the United Kingdom’. My role at the university involves teaching and supporting nursing and allied health students, so this isn’t something I do on a regular basis, but I was keen to be involved.
The project had a particularly tight time constraint; a search of relevant databases, screening of the articles and writing the report needed to be completed by the end of March 2024. The intended aims of this report are to inform the development of policy and advice for institutional settings and support cross-governmental decision-making on pandemic preparedness.
Creating the search strategy
I met with the professor and other university researchers to discuss the keywords and synonyms we thought we should use; different examples of NPIs for example, and which inclusion groups the project would focus on. We then sent our draft search strategy to the UKHSA for scrutiny, and several online meetings with their team ensued (participants included the UKHSA Head of Evidence and Professor Andrew Hayward!). They provided us with feedback on the search strategy, and there was lots of debate around the terms both teams wanted to include and exclude, and which databases should be used. A sensitive search approach was decided upon, combining only search terms for Covid, NPIs and within the Ovid databases, a UK filter.
We used recommendations for search terms including the NICE developed Covid search strategy as a starting point, incorporating different syntax for each database. Once the keywords had been approved by both teams, I met with their information scientist, who peer reviewed my search strategy and we decided that she would run the searches in Ovid Medline & Embase, and I would run in EBSCO CINAHL, PsycInfo and SocIndex. The final search strategies were complex, averaging around 150 search lines for each database. We then had two further meetings with the UKHSA team before the protocol and search strategy were signed off.
Running the searches in the databases
Part of the criteria was that the review would include studies published from January 2020 to 4th January 2024. This meant searches in all databases must be run on the 4th of January 2024. As the searches had been constructed and saved in the relevant databases in December, this was just a case of rerunning the searches. I met with the information scientist to do a final check on the morning of the 4th of January, and then we ran the searches separately. The searches yielded 5,990 results in total. We used Deduklick software to sift out duplicates, leaving 4673 for the team to screen.
I then exported the results to our research team so they could begin to conduct the review using Covidence software. They managed to remove a further 10 duplicates and initially screened the results by title/abstract, before looking at 80 key articles in more depth and deciding on 13 to be included in the review.
Concluding thoughts
I was excited to be part of the project, and the team have kept me updated with the process as they’ve progressed through it. I learned a great deal from the UKHSA information scientist and made some excellent contacts within the research team. Additionally, my contributions will be recognised with a credit in the impending publication. I can highly recommend if you ever get the opportunity to be involved in something like this!
Health Libraries Group 