“Research is something that everyone can do, and everyone ought to do. It is simply collecting information and thinking systematically about it.” (Raewyn Connell, 2021)
The research process is made up of several steps, all of which are important to effectively answer a question. Whether you are carrying out a literature search or a systematic review, the research process can seem overwhelming, even confusing. You may feel confident about your search, but then wonder how to properly evaluate what you have found. This post will help you decide on important factors for screening result sets for inclusion or exclusion.
Screening your search results is the process of weeding out any inadequate articles that you obtain after running your searches (Dalhousie Libraries, 2021). Screening is an evaluation tool that can be applied not only to literature searches and formal reviews, but also when quickly looking for best evidence on a topic. Screening can help evaluate “in the moment” search results, such as those required for patient care.
The screening process can be performed by one reviewer or a team of reviewers, depending on the information need or type of study being conducted. For example, “in the moment” research results may only have one reviewer, while a scoping or systematic review may have a team of reviewers. It is important that all reviewers on a team adhere to the same screening criteria.
When screening your results, whether for a formal research project or to gather evidence "in the moment" to inform patient care, keep these tips in mind:
- Establish inclusion and exclusion criteria – If you are working with a team of reviewers, all members must be clear on which criteria are going to be included and which are going to be excluded. For example, inclusion criteria could be all participants in a study must be 50 years old and over. This means that any study that discusses participants under the age of 50 will automatically be excluded from the search results.
If you are the only person screening search results to inform a patient care decision, it is useful to know the type of study that will provide evidence-based results to support a therapeutic choice, such as an adult woman deciding what blood thinners will best manage her atrial fibrillation.
- Review results using established inclusion and exclusion criteria – Remove items that clearly do not fit inclusion criteria and those not from a trustworthy or credible source. In the informal, patient care example above, you can screen in systematic reviews and/or randomized controlled trials.
- Read the full-text – After determining which articles will be included, access and read the full-text. Read the articles using critical appraisal tools, screening for credible evidence, bias, and peer reviews. In more formal reviews, full-text screening is often categorized as the “second level of screening,” as it follows a more in-depth, rigorous process (Dalhousie Libraries, 2021).
- Check for and remove any duplicates – If you are searching multiple sources, make sure you haven’t included the same item more than once. Many citation managers, such as Zotero, have a feature that will check your search results for duplicates. Be sure to check out Library Services’ Citation Management guide for further instruction.
Tools that will guide your screening work
The PRISMA Flow Diagram (Moher et al., 2009) is a great tool to help more formal research projects record and summarize the screening process. There are also several other tools to help you manage evidence synthesis and aid in the screening process. The University of Toronto Libraries provide excellent step-by-step instruction on screening search results for those researchers looking to dive deeper into the process.
Screening your results is a vital step in the research process—whether you are gathering evidence for a formal publication or using it to inform a local project or patient care. Our tips and tools will make it easier for you to put the knowledge you find into action.
Reach out to us at AskLibrary@nshealth.ca with any questions you have about screening search results, or the research process in general. You can also book a one-on-one consultation with a Library Services team member. We are always here to help!
References
1. Dalhousie Libraries. (2021). Knowledge Syntheses: A How-To Guide. https://dal.ca.libguides.com/systematicreviews/selectionscreening
2. Fisher, S. (2021). 17 research quotes to inspire and amuse you. Qualtrics. https://www.qualtrics.com/blog/research-quotes/
3. Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRISMA Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 6(6): e1000097. doi:10.1371/journal.pmed1000097
Amanda Andrews
Librarian Educator, Education & Training Lead
Cape Breton Regional Hospital, Eastern Zone
Current, peer-reviewed info about COVID is vital as we move into the later stages of the pandemic. Thankfully, scientists and clinicians around the world continue to answer the call, publishing record numbers of articles and sharing even more pre-prints.
You may have noticed that the majority of the results for your recent searches have a COVID focus. But what if you don’t need COVID-focused evidence? When you do not want a term included in your search and that term is clouding your ability to screen in relevant items, consider using the NOT Boolean operator.
Using NOT to exclude results
If you’ve participated in any of the Library’s education sessions about database searching, you may be familiar with how Boolean operators work:
Use caution when using NOT, as it is easy to unintentionally exclude potentially valuable results. For example, an item’s metadata (the info that describes the item) may use the term you are excluding alongside other terms that are relevant to your search.
Using NOT to exclude COVID in PubMed
Let’s say you’ve heard a lot about wastewater monitoring for COVID in Canada and are wondering what else is being monitored this way. If you run a quick search in PubMed for wastewater AND (monitor OR surveillance) AND Canada, you’ll notice that the top results have a strong COVID focus. To remove COVID results, you must do two things:
Adding brackets around your first search, then adding NOT and any additional term(s) will make sure the database processes your original search first.
Try: (wastewater AND (monitor OR surveillance) AND Canada) NOT COVID
Step 2 is where things can get complex because we have choices about how comprehensive we want to be with our exclusion. Simply adding “NOT COVID” to the original search in brackets will exclude some COVID articles but may miss excluding some since there are many variants of the virus. If you want to be sure you are being as comprehensive as possible, what should you do?
Using COVID-19 PubMed filters: Inclusion & Exclusion
The National Library of Medicine (NLM) maintains PubMed and has created highly accurate search strategies to isolate results about COVID. These search strategies range from General to Treatment-focused and take the variable terminology inherent in COVID into account. Originally developed to let you easily include COVID content in a more comprehensive way, you can also use these strategies to exclude COVID content more accurately.
Thinking about Step 2 in the process above, you can grab the shorthand for a specific filter, Filter Name (1), and add it to your search with NOT using the Filter command (i.e. [Filter]).
Let’s add the General COVID-19 Filter: (wastewater AND (monitor OR surveillance) AND Canada) NOT LitCGeneral[Filter]
This extra step isn’t always necessary, as your own terms may be just as effective at excluding from a current search result set. However, when you are setting up a search strategy for monitoring new results over time, using these tested filters can help ensure you account for evolving terminology.
Using NOT in other databases
NOT, the other Boolean operators, and brackets are core mechanics of how any database works. You can use the steps outlined here to exclude a term or search strategy in most biomedical databases supported by Library Services (such as CINAHL, Embase and more).
You can also look for COVID-19 search strategies, or other search strategies developed by each database. You might notice these filters in pop-ups or alerts in a given database, or labelled as “evidence-based” or “clinical” filters within the Advanced search options.
If you are interested in learning more about using advanced searching techniques or tested search filters, please book an online consultation with a librarian at a time that works for you. If you’re interested in learning more on your own, visit the Search Filters page on our Searching subject guide.
Katie McLean
Librarian Educator, Outreach Lead
Dickson Building, Central Zone
Related posts