The Institute of Medicine, the  Cochrane Collaboration, the Centre for Reviews and Dissemination (University of York), and other organizations each have their own recommendations on conducting a systematic review or practice guideline.  But by and large they have in common the following steps:

• Establish an expert team
• Manage conflicts-of-interest of team members
• Formulate the question
• Identify inclusion and exclusion criteria
• Identify relevant studies from different sources, published and unpublished
• Select studies for inclusion and evaluation
• Systematically collect and manage data
• Synthesize the evidence
• Create the report

Research librarians and other information specialists can help with the 4th bullet point - "identify relevant studies"  - through the professional literature search service.  You can also do your own search with the recommendations on the search process below.

Before beginning a systematic, scoping, or other evidence synthesis review, develop a detailed protocol to ensure clarity, consistency, and integrity in the review process. A protocol serves as a roadmap for the project that outlines the objectives and methodology and guides the research team through pre-established steps.  This increases accountability and makes the process more transparent and and reproducible for future studies.

• PRISMA for Systematic Review Protocols (PRISMA-P) is a checklist to guide the preparation and reporting of a systematic review protocol.
• PROSPERO is a systematic review registry that promotes transparency, reduces reporting bias, and help prevent unintended duplication and research waste.  Search the PROSPERO protocol database and/or register your protocol.
• Open Science Framework (OSF) provides forms and and registration for systematic reviews, scoping reviews and other research.
• Some journals publish protocols.  For example, Systematic Reviews is an open access journal that publishes protocols. Learn about open access fees here.

Librarians can assist you with your search, but if you conduct your own comprehensive search, here are some recommendations.

• Formulate your question to identify the major concepts.  PICO - Patient, Intervention, Comparison, Outcome -  is a useful tool. 
• Search in as many relevant databases as possible.  For clinical topics, we recommend you search at a minimum Ovid Medline (or PubMed), Embase and Cochrane CENTRAL via the Cochrane Library.  See the Where To Search tab in this Resource Guide.
• Consider searching the unpublished grey literature which includes clinical trial registries, government and industry reports, dissertations, unpublished practice guidelines, press releases, and conference abstracts. See the Where To Search tab in this Resource Guide.
• Apply as few limits as possible. Common limits often applied but generally not recommended include: year limits (e.g., last 10 years) and language limits (e.g., English Only).
• Use both subject headings where available (e.g., MeSH) and free-text terms (textwords)
• Expand your textword searching (not subject headings) by:
  • using similar terms for each concept and combining them with “OR” (e.g., braces OR supports OR orthoses)
  • using spelling variations, e.g., orthoses OR orthosis; randomised OR randomized
  • searching abbreviations, acronyms as well as their spelled-out versions: EHR OR electronic health records
  • truncating or using wild card symbols. For example in PubMed, back pack* includes back pack, back packs, back packer, back packing, and more.
• Document your search process, noting:
  • search concepts (e.g., Heart Failure) and terms (heart failure OR congestive heart failure OR CHF).  Try using a table or spreadsheet to keep track.
  • names of databases and other resources where you searched
  • the range of years of the databases (PubMed includes 1946-current)
  • dates you conducted the searches (e.g., May 10, 2017 - May 30, 2017) or when you stopped searching ("searches are complete through May 30, 2017")
  • the search strategy, including the terms and their combinations (using OR, AND, NOT), and all the sets
  • any limits (e.g., date ranges, languages, study designs, age groups) that you applied with justifications
  • any citation or systematic review management software  utilized (e.g., Endnote, Covidence, RevMan)
• Consider a peer-review of your search strategy.  CADTH (Canadian Agency for Drugs and Technologies in Health) developed the PRESS checklist. 

Guidelines for Conducting Systematic Reviews

• Cochrane Collaboration
  • Cochrane Handbook v6.0
• Campbell Collaboration
  • Methodological Expectations of Campbell Collaboration Intervention Reviews (MECCIR) - Conduct Standards

Guidelines for Reporting Systematic Reviews.  Utilize one of these guidelines when you're ready to write-up and report your review.

• PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)  Provides recommendations on reporting systematic reviews and meta-analyses.
  • Statement
  • Checklist
  • Flow Diagram
  • Extensions (for protocols, scoping reviews, etc.)
• Systematic reviews registered with JBI
• Equator Network Reporting Guidelines for Main Study Types includes a list of reporting guidelines for different publications, e.g., systematic reviews, quality improvement, randomized trials, etc.