The goal of this session is to explore the different elements of systematic searching in comparison to less standardized searching. Important note: non-systematic searching is not wrong! There are plenty of instances where a quick, simple search is the ideal method. But for an advanced approach, it's important to employ new strategies for comprehensive search results.
The workspace provided for this course contains four tabs within an Excel document. You can download the workspace from the home page of the research guide; it is also included at the bottom of these instructions for convenience.
The first tab is a term-harvest worksheet with space to record key concepts and potential keywords for your research topic, as well as inclusion and exclusion criteria.
The second tab provides space to paste your PubMed search strings and record the total number of results retrieved. This helps you track changes as you refine your search.
The third tab offers a similar workspace, but with additional tables to help you keep track of your Embase and Scopus searches as you work through the nuances of translating your PubMed search into other relevant databases.
The fourth and final tab provides an example of how to use the document.
The type of search you use directly impacts the quality, transparency, and reproducibility of the evidence you gather. While non-systematic searches may be sufficient for quick information needs or exploratory purposes, systematic searches are required when the goal is to comprehensively capture all relevant literature on a topic, such as in systematic reviews or evidence-based practice.
A non-systematic search is typically organic and meandering, often ending once the searcher feels they have found "enough" information. It usually involves one or two databases selected for convenience rather than based on a rigorous plan. The search terms used are often simple and may not be consistent across different databases. In contrast, a systematic search is structured and carefully planned, continuing until all available literature on the topic has been located. It includes as many relevant databases as possible, and the search terms are comprehensive, accounting for all possible ways a topic might be described. When multiple databases are used, the search strategy is translated as consistently as possible across them.
| Non-Systematic Search | Systematic Search |
| Organic, meandering, ends when you've found "enough" | Structured, planned searches that conclude only when all available literature on the topic has been found |
| One or two databases, chosen without rigor or planning, priority on convenience of search | Uses as many databases as are available and relevant for the search topic |
| Search terms are simple and not necessarily identical across multiple databases | Search terms are exhaustive, ideally accounting for all ways a topic may be described. When using multiple databases, the systematic search is translated as identically as possible. |
Controlled vocabularies are a set of premade terms that describe specific concepts. Many databases create their own thesaurus of controlled vocabulary terms; this includes Medical Subject Headings (MeSH) in PubMed and Emtree in Embase, to name a few.
Controlled vocabularies are created by subject experts that review citations as they are added to the database, then they map controlled vocabulary terms that describe all of the concepts covered in the article. Think of them almost like #hashtags: using a hashtag is a way to connect your information to an index of conversations happening around the same topic. That's how controlled vocabulary works!
Visit the MeSH Database (also accessible from the PubMed homepage under “More Resources”).
In the search box, type a keyword or concept (e.g., "heart attack") and click Search.
Browse the results to find the MeSH term that most closely matches your concept (e.g., "Myocardial Infarction" is the MeSH term for "heart attack").
Click on the term to view its definition, entry terms (synonyms), and subheadings (which let you focus on specific aspects like prevention or diagnosis).
If you're using the MeSH Database, you can check boxes for subheadings and click “Add to search builder.”
Click Search PubMed to run the search with the selected MeSH term.
Alternatively, copy and paste the term directly into your search using the MeSH tag, like this:
"Myocardial Infarction"[MeSH Terms]
Keywords, also called "entry terms", are the natural language you use when searching browsers ("brunch spots near me", etc.) When using keywords in research databases, you're asking for hits on a term used in any field of a record: title, abstract, journal title, author name... This has pros and cons. A pro of this kind of searching include capturing every time a term appears in the database, not just the time a subject expert assigned a controlled vocabulary term. However, this also means you need to consider all the way an author might describe your topic, including multiple spellings, singulars/plurals, and any synonyms.
Keyword searching is also useful when attempting to identify literature that may not have been indexed with controlled vocabulary terms; for example, there can be a backlog of records not yet assigned MeSH terms. If you're only using MeSH terms, you can miss the most recent literature!
It's recommended to combine controlled vocabulary and keywords for the most thorough search. Here's why!
To ensure your search is covering all your bases, use both controlled vocabulary terms AND keywords!
Navigate to PubMed.
Click on “Advanced” just below the main search bar. This opens the Advanced Search Builder.
In the Add terms to the query box section:
Choose “Title/Abstract” from the drop-down menu on the left.
Type your keyword or phrase into the search box (e.g., opioid use disorder).
Click “Add” to add it to the query box.
Repeat for any additional synonyms for your search concept, combining them with OR.
You can also manually type OR between your synonyms in the search box to save time.
Once your full query is built in the Query box, click the “Search” button. This search will then be added to your history, which will be helpful when we combine all our subject-searches later. You may also copy and paste the search into your documentation file.
Boolean operators—AND, OR, and NOT—are used in database and online searching to help narrow or expand search results.
Quotation marks are used in searching to keep words together as an exact phrase. Without quotation marks, most databases and search engines will treat each word separately, potentially returning results where the words appear anywhere in the document—not necessarily together or in the desired order.
Truncation is a search technique used to find different word endings or variations by using a symbol—usually an asterisk ()—at the root of a word. This helps you retrieve more results without having to type each variation.
For example, searching for nurs will find nurse, nurses, nursing, and nursed.
Parentheses are used in searching to group terms and control the order in which Boolean operators are applied—much like in math. This is especially important when combining AND, OR, and NOT in a complex search. For example, in the search (diabetes OR "blood sugar") AND diet, the parentheses ensure that the database finds results that mention either diabetes or blood sugar, and also include diet.
When searching in PubMed, you can improve the precision and coverage of your search by using field tags, which tell the database where to look for your terms. The two most commonly used tags in advanced searching are:
[MeSH Terms]: This tells PubMed to search for articles indexed with a specific Medical Subject Heading (MeSH). MeSH terms are standardized vocabulary that help group together articles on the same topic.
[tiab]: This stands for title/abstract, and it tells PubMed to search for the words in either the article’s title or abstract. This is useful for capturing newer articles that haven't been indexed with MeSH terms yet, or to include natural language used by authors. For instance, "high blood pressure"[tiab] would find that exact phrase in titles or abstracts, regardless of MeSH indexing. Using this tag instead of all fields helps ensure you only get relevant articles.
The session recording explains how you can utilize advance searching filters to automatically apply these tags instead of typing them out. However, it is important to know these will be added to your search when you copy it out of PubMed for your search documentation.
In the Term harvesting tab of your spreadsheet, complete each concept with appropriate MeSH Terms and keywords. Below you'll see walkthrough steps for finding MeSH and keywords, as well as nesting them together. If you need extra help in the nesting/search string construction, scroll to the bottom of this page for some essential tips for searching databases.
Remember that there are examples in the final tab of your spreadsheet!