Skip to Main Content

All Roseman University campuses will be closed on Monday, December 23 for Winter Break. Campuses will reopen for regular building
hours on Saturday, January 4.

Perform an Evidence Synthesis

Steps in an Evidence Synthesis


There are many steps involved in an evidence synthesis. A librarian can be involved throughout the process, providing support and subject expertise. It is recommended that as well as a librarian the team should include subject area experts as well as a team member with quantitative methods expertise.

Develop Your Protocol


An evidence synthesis protocol states your rationale, hypothesis, and planned methodology. It acts as a guide as you complete the steps of your evidence synthesis. Be sure to develop your protocol before getting started on your evidence synthesis.

Below are some resources you can use to help you develop you evidence synthesis protocol.

Step One - Develop Your Research Question


Developing your research question is key to any literature review. It will be the basis of your entire evidence synthesis. To get started developing your research question you will need to identify a knowledge gap in your field and aim to answer a specific question.

You can learn more using our Search the Literature Library Guide. This guide has a section specifically covering how to Formulate a Question.

Inclusion & Exclusion Criteria


Once you've finalized your question, but before you begin searching, you should establish your inclusion and exclusion criteria. Inclusion and exclusion criteria establish limits for your evidence synthesis and are typically reported in the methods section of the publication.

 

 

 

 

 

 

 

 

 

 

 

 

 

Step Two - Select Databases


A librarian can help you select the databases for your review. Every database is different and will require a different search syntax. Some databases utilize a controlled vocabulary. An example of this would be the Medical Subject Headings or MeSH Terms that are used in PubMed and the Cochrane Library.

To begin browsing databases, visit the All Databases and Resources List:

Grey Literature


Grey literature is literature that is produced outside of the commercial or academic spheres. This can be a variety of different types of information such as government reports, conference proceedings, dissertations, theses, unpublished clinical trials, and much more.

Depending on the type of review you are performing you may consider including some sources of grey literature. Grey literature represents a significant body of information that could be critical to answering your research question.

Finding grey literature can be challenging. Below you can find a few resources that you can use to search for grey literature.

Step Three - Write a Search Strategy


When creating a search strategy it is recommended that you speak with a librarian. Librarians can help you ensure that you retrieve the highest number of relevant results as well as assist you in translating your search strategy into multiple databases.

You can learn more about searching the literature and building a search strategy by visiting our Search the Literature Library Guide.

Keywords

To start you should select your keywords based on your topic and research question. Based on these key words you will need to create a list. 

Your list should inlcude:

  • synonyms
  • alternate spellings (e.g. British vs. American English)
  • punctuation (hyphen or no hyphen)
  • acronyms and spelled out acronyms
  • narrower and broader terms
  • plural and singular

Do Not Use

There are some terms that are better left out when developing your search strategy. These are often referred to as "stopwords". You can find a list of all the "Stopwords" in PubMed.

Some of the more common types of "stopwords" are:

  • comparative words (effect, impact, reaction, cause)
  • prepositions (during, due to, within, without)
  • interrogative words (who, what, when, where, and why)

Boolean Operators

Translating Your Search Strategy


Evidence synthesis requires researchers to search multiple databases. Unfortunately most databases do not accept the same search syntax. This means that a search in PubMed using the syntax Tooth[tiab] will pull articles with the term "Tooth" in the title or abstract, but using the same syntax 

Tooth[tiab] in web of science will provide no results.

You can talk with a librarian and they can help you with this step. Additionally, there are a few resources that you can use to help you translate your search strategy.

Step Four - Citation Management


In an evidence synthesis you will likely be working with hundreds or even thousands of citations. Citation management tools can help by storing, organizing, and managing your citations. There are a number of different citation managers that you can use. Roseman provides access to EndNote but you are welcome to use the citation manager that works best for you.

 

 

You can learn more by visiting our Use EndNote and Use Zotero Library Guides.

Deduplication


As you retrieve citations you will likely retrieve multiple copies of the same article. You will need to deduplicate your results before moving on to the article screening. This is something that your Citation Manager can do for you.

  • In EndNote, from the menu bar select “Library” → “Find Duplicates”. You can select which of the duplicate records to keep by clicking “Keep This Record”. You can also keep both records by clicking “Skip”, which will cause the records to be added to a “Duplicate References” folder.
  • In Zotero, click on the "Duplicate Items" collection in your library. You can resolve duplicates by merging the files.

Step Five - Article Screening


Once you've retrieved all your citations and removed the duplicates you can start the screening process. During this step you review all of the articles that you pulled during your search and remove any that are not relevant to your topic. You do this by checking each article to see how it matches your Inclusion/Exclusion criteria. Any studies that don't meet your criteria should be excluded from your evidence synthesis.

It is important to note that at this stage you do not need to read the full text of every article. You can determine whether the article meets your Inclusion/Exclusion criteria by reading the title and abstract.

The screening should be performed by multiple team members each checking the others work. Any disagreements about Inclusion/Exclusion should be settled by a third party.

Step Six - Data Extraction


Now that the initial screening is complete you can move on to the data extraction. In the data extraction you will take all of the articles that were included in the article screening and read the full text and extract the data relevant to your review. It is important that your data extraction makes sense to you and is accurate. This is where you will be pulling data from as you write your evidence synthesis.

It is not uncommon to exclude articles at this stage. Often an article looks promising in the screening step and is included in the data extraction. Then, once you start reading the full text you realize it includes one or more of your exclusion criteria. It is ok to exclude, just keep a record of why.

You can do a data extraction using Excel, Google Sheets, or any other software.

Step Seven - Risk of Bias Assessment


An evidence synthesis strives for the elimination of bias. Each study that you include has the potential to introduce bias. In order to be transparent about this bias you should do a risk of bias assessment. This means that you look at each article and examine it for bias, which you then report as part of your evidence synthesis.

Scoping reviews don't typically include a risk of bias assessment. This is one of the key differences between a scoping review and a systematic review.

Step Eight - Synthesize Results


In this step you need to present the main findings of your evidence synthesis. There are a number of ways that you can synthesize the results from your included studies.

If the studies you've included are sufficiently similar (i.e all the same study design and data collection methods) you can synthesize the data in a meta-analysis. This uses statistical methods to bring together the results from multiple studies.

If the studies you've included are not similar, a meta-analysis is not possible. Instead you will examine the data from your data extraction and look for themes that you can extrapolate for your evidence synthesis.