Welcome to the course page for HSCI 4380. This guide offers you resources and support for designing your Evidence Synthesis proposal.
As part of your assignment, you will be meeting with the Health Science Librarian to review and discuss your proposal.
Please email Celia Brinkerhoff (celia.brinkerhoff@kpu.ca) if you have any questions.
In general, an evidence synthesis is a research methodology designed to answer a well-defined question by gathering all relevant information on that topic. A systematic review is one type of synthesis.
The Cochrane Handbook of Systematic Reviews of Interventions, 6.1 defines a systematic review as an attempt to:
"collate all empirical evidence that fits pre-specified eligibility criteria in order to answer a specific research question."
The key characteristics of an evidence synthesis are:
a reproducible methodology;
an assessment of the quality of the data and an interpretation of the findings
There are several types of evidence syntheses, and the nature of the research question will determine which type is appropriate. For a description of the various types, see the Guide to Evidence Synthesis from Cornell University Library.
Concept of systematic review from Cochrane Infographics.
Here are some additional resources for understanding what an evidence synthesis is:
A research protocol is a plan for how you intend to carry out your evidence synthesis. It includes the rationale, a clearly stated research question (see PICO sheet next box), definitions, eligibility criteria, and so on. It also indicates the databases you will search and, where appropriate, whether you will be considering the grey literature and hand-searching journals.
The main purpose of a detailed protocol is to provide as much transparency and reproducibility as possible
For formatting your protocol, see the helpful Evidence Synthesis Protocol Template on the Open Science Framework.
Before you get too far in the review process, it's important to ensure that no systematic review has already been done on your question. It can also be helpful to note where reviews have been done so that you might consider adjusting your question slightly by identifying a gap in the research, or including a different population, comparative treatment, outcome, etc.
Use one of these databases to determine what reviews already exist on your topic.
The first step in in your synthesis is to frame a good question. Recall that one of the defining characteristics of a systematic review is a clearly stated objective, along with eligibility criteria.
Health researchers and librarians use a tool called PICO to help them develop well-defined, searchable clinical questions and to structure their review process.
Use this PICO worksheet to help you plan your question.
Define the major concepts of your question. Consider synonyms or related terms, as well as broader and narrower terms. This page from the UToronto Gerstein Science Library provides good details on developing your search terrms.
Gather synonyms and think broadly about your terms
Use the MeSH database to check your terminology.
MeSH (Medical Subject Headings) is a controlled vocabulary, much like a thesaurus, that is used to index the biomedical literature in a systematic way. All variations and similar terms are linked to one authoritative heading which then gets applied across the journal articles to richly describe what each is about. The vocabulary has a tree structure, so that every term is part of a branch with broader and narrower terms. Most articles will have 10-15 MeSH terms applied, some indicating a major focus of the article.
Check out the quick tutorial Medical Subject Headings in MedLine and PubMed, from the National Library of Medicine.
Now is a good time to also consider your inclusion and exclusion criteria. These may need to be revised as you proceed with searching and view initial results. This will include things like:
Chapter 3 of the Cochrane Handbook has detailed questions to consider when coming up with your eligibility criteria.
Several databases will have an ability to filter by study type or design. What kinds of articles you will include in your review will depend in part on the question you are asking;
Image from Evidence Based Practice, UVA Health Sciences.
Remember that search is an iterative process, and will involve revising and altering your terms and strategies until you eventually get a set of results that looks relevant. Using several databases and search tools will be an important part of your strategy.
Start running searches in your selected databases (see the list below). Identify a KPU Library subscription database to start your searching, and supplement the results with a search of PubMed.
Why search more than one database?
Image adapted from PubMed Quick Tour- What's in PubMed?
Most comprehensive systematic reviews will also search the grey literature; that is, research that is published outside of the academic and commercial publishers, but still of very high quality. Think of research organizations such as the NIH, CIHR and so on as possible sources to include in your strategy.
See the Grey Literature and Statistics box on this guide.
Add Google Scholar to your suite of research tools. Google Scholar is a search engine that covers much of the scholarly, peer reviewed literature from academic publishers, plus conferences, theses and dissertations, abstracts and more. Some of the content is available full text, and some will be citation-only.
You can configure your settings to include links to articles available from Kwantlen Polytechnic Library subscription journals to avoid hitting publishers' paywalls. See the screenshot below.
For articles NOT available in full text in your database searches:
1. Verify the library does not have access through any database, using the Journal title search on the library's homepage.
2. Use Interlibrary Loan to obtain a free copy.
Plan a tool for organizing your search results. It is highly recommended that you use a bibliographic manager for this, such as one of these free tools:
For importing files into COVIDENCE, it is recommended that you first export your articles from a database search to a citation manager such as Mendeley or Zotero, using the .ris file format. You may also import articles from a PubMed search (.nbib format).
The following guide from the Health Sciences Library at the University of North Carolina has full details for importing studies into your COVIDENCE review.
Select articles.
Apply the inclusion or eligibility criteria as defined earlier and conduct an initial assessment of articles based on their titles and abstracts only. Use a system to indicate which studies will be considered eligible, as well as why you have eliminated others. This should be collected in a notes field in your data collection file, for later reference. Recall that the validity of your systematic review depends on being transparent about your reasons for including or rejecting the articles.
Analyze articles.
Conduct a thorough second reading of the full text of those studies you initially accepted, paying careful attention to methodologies, populations, etc.
PubMed tutorial (approx 30 mins)
Here is a very useful 4 part tutorial from the National Library of Medicine on how to use PubMed in evidence-based practice. It offers an introduction to searching on the new interface, making use of MeSH, and applying filters to search results.
Systematic Searches from Yale University Library (60 minutes)
A suite of videos introducing the planning of a systematic review, utilizing PubMed and MeSH and OVID Medline. Video #2 Systematic searches is especially useful for providing an overview of the process.
An Introduction to Systematic Reviews by 