Training and support are available in the following ways:

• Top tips guides and video tutorials (link opens in new window) on how to use all the major databases
• Introduction to literature searching training for University of Cambridge researchers/students (link opens in new window), or for NHS staff (link opens in new window). For those wanting more advanced searching training our database Q&A sessions (link opens in new window) are recommended.
• You are also welcome to contact the training team at librarytraining@medschl.cam.ac.uk to arrange a one-to-one meeting. We would expect you to have attempted a literature search in advance, and have a clear idea of your research topic and the specific problems you are having, prior to arranging a one-to-one meeting.

Find the material you need at:

• Databases for literature searching (link opens in new window): the Medical Library's curated list of the major medical, healthcare and scientific databases
• Databases A-Z (link opens in new window): the full list of databases available to University of Cambridge students and researchers. A Raven login is needed to access these remotely
• NHS Athens resources (link opens in new window): the full list of databases available to NHS staff. An Athens login is required to access these

While most of the standard medical, healthcare and interdisciplinary databases are appropriate for use in systematic reviews, Pubmed should be avoided. Pubmed lacks certain search functionality (for example the ability to code for proximity operators), making it different to search consistently across multiple databases. However, the biggest problem is that Pubmed searches are not reproducible and transparent. Pubmed:

'uses machine learning algorithms working behind the scenes which are invisible to the searcher. That means that transparency and reproducibility is no longer possible. Transparency and reproducibility are of key importance in scientific reporting and experiments. Without these present in the search strategy, a systematic review falls at the first hurdle when being critically appraised.

PRISMA-S (link opens in new window) was launched in 2020, outlining all the reporting requirements for literature searching in systematic reviews. Item 8 is: Include the search strategies for each database and information source, copied and pasted exactly as run. Note ‘exactly as run’. This is not possible in PubMed. Medline on the OVID platform (or via EBSCO or other aggregator) is preferred.'

Source: Exploring the Evidence Base (link opens in new window).

The requirements for systematic reviews will mean that your literature search will need to be more complex and thorough than previous literature searching you may have done. You will need to come up with many synonyms for your search terms, and make sure that you are searching in the title, abstract and MeSH term fields of every database you use. Make sure you are using the Boolean operators AND, OR and NOT correctly, and make use of other search syntax such as truncation and wildcards to ensure your search is as comprehensive and efficient as possible.

Most systematic review search strategies use PICO to frame their search.

University of Toronto Libraries have written an extensive, detailed guide to the risks and possibilities of using generative AI tools such as ChatGPT to write a systematic review search strategy. We recommend you read this guide in full, but focus in particular on its final conclusions.

In this section of the guide, University of Toronto Libraries share a journal article in which the authors have used ChatGPT to generate a Pubmed advanced search strategy (Shuai Wang, Harrisen Scells, Guido Zuccon, and Bevan Koopman. 2023. Can ChatGPT Write a Good Boolean Query for Systematic Review Literature Search?. 1, 1 (February 2023), 19 pages. https://doi.org/10.1145/nnnnnnn.nnnnnnn.) 

Look at the search strategy that ChatGPT generated: 

(((differentiated thyroid cancer[MeSH] OR "differentiated thyroid"[All Fields] OR "thyroid carcinoma"[All Fields] OR "papillary microcarcinoma"[All Fields]) AND (prevalence[All Fields] OR incidence[MeSH] OR "etiology of"[All Fields] OR "risk factors"[All Fields] OR gender[All Fields] OR hormonal[All Fields] OR "nodular goiter"[All Fields] OR "Hashimoto’s thyroiditis"[MeSH] OR malignancy[MeSH] OR "concomitant lesion"[All Fields] OR tumor[All Fields] OR infiltrate[All Fields] OR fibrosis[All Fields] OR "early stages of development"[All Fields] OR frequency[All Fields])) AND (autopsy[MeSH] OR surgical[All Fields] OR material[All Fields] OR series[All Fields] OR specimens[All Fields] OR cases[All Fields]))

Ask yourself the following questions:

1. Can you identify the errors in this search?
2. Can you explain what the search is doing in PubMed?
3. Would you trust the results of this query to find all the relevant studies for your question (ie. recall/sensitivity) while limiting the number of irrelevant studies (precision?)
4. Would you be able to translate this query into additional databases?
5. Do you trust ChatGPT to do your data collection?

If your answer is not 'yes' to all these five questions, we would advise against relying solely on a generative AI tool to create your systematic review search strategy.

These tools will help you improve the quality of your search strategy:

• Pubmed ReMiner (link opens in new window): a tool to help you build your search by showing you the most commonly used freetext and MeSH terms used in articles about your topic
• Visualizing Pubmed (link opens in new window): provides various formats of data visualisation of the results you find in Pubmed
• InterTASC Search Filters Resource (link opens in new window): a collaborative venture to identify, assess and test search filters designed to retrieve research by study design or focus
• Library of Search Strategy Resources (link opens in new window): a constantly updated set of resources such as databases, search filters, and tutorials to help systematic reviewers build their search strategies
• SIGN methodology filters (link opens in new window): pre-tested search filters designed to narrow results to particular study design types in a range of databases
• Database Syntax (link opens in new window): librarian Janneke Staaks has designed a resource which translates search syntax across the most commonly used database platforms
• OECD countries search filter (link opens in new window): search terms validated by NICE to limit your search to specific geographic regions
• Country URLs (link opens in new window): list of domains to limit search results to websites from specific countries

Systematic review search strategies are more complex than those you are perhaps used to — they frequently require tens or even hundreds of lines of search terms, a mixture of MeSH and freetext terms, translating search syntax from one database to another, and use of advanced search syntax such as truncation (using * to truncat* words), proximity (using code to search for words being within a certain number of words adjacent to each other), and wildcards (using ? to replace one letter with another).

This guide by the University of Exeter Library (link opens in new window) gives a detailed explanation of how to 'decode' search strategies in systematic reviews.