Skip to main content

Systematic Reviews: Define Topic & Register Protocol

A guide to conducting systematic reviews.

Need Help?

 

        

Visit the Reference Desk.

(Mon. - Thurs., 9am - 6pm, and Friday, 9am - 5pm.)

Email us your question.
   Call us: 718.430.3104 
(Mon. - Thurs., 9am - 6pm, and Friday, 9am - 5pm.)
Text us

Text us: 718.407.0761

(Mon. - Thurs., 10am - 6pm, and Friday, 10am - 5pm.)

Virtual Consultation

Call us at 718.430.3104. We'll show you how it's done -- on your own computer.

(Mon. - Fri., 10am - 4pm)

Search our FAQs.

 

Clearly Define the Question

Meet with the research team to clarify:

  • objectives
  • key questions

With a librarian's assistance, find any existing or in-process systematic reviews on your topic.

 

Use the PICO framework to identify the key concepts of the question:

  • Patient or Population or Problem 
  • Intervention
  • Comparison
  • Outcomes


See the AHRQ report from January 2013: The Refinement of Topics for Systematic Reviews: Lessons and Recommendations From the Effective Health Care Program: This report details guiding principles and methodological recommendations that may help investigators to better refine topics for systematic reviews. The primary goal of topic refinement is to formulate research questions that can be addressed by a systematic review;

Develop & Register Protocol

A protocol should include:

  • context and rationale for the review
  • primary outcomes of interest
  • preliminary search strategy, specifying databases, search terms, etc.
  • inclusion/exclusion criteria, stating which “subjects” or studies will be included into the review. The topic or question addressed will influence the study design(s) you search for. See What To Look for in the Evidence-Based Medicine research guide.
  • data synthesis strategy
  • other aspects of the research plan, as appropriate.

Publish the protocol in a public registry, such as Prospero.

The protocol can be altered as necessary during the project, but the inclusion/exclusion criteria should be set before the actual search is conducted in order to reduce bias.