PRISMA Flow Diagram: Step-by-Step Guide for Systematic Reviews

A PRISMA flow diagram is a visual summary of the stages of a systematic review process, tracking how many records were identified, screened, assessed for eligibility, and finally included. It’s the single most important reporting tool in systematic review methodology — and understanding how to create one correctly is essential for graduate students in health sciences, education, social sciences, and engineering.

This guide walks you through every stage of creating an accurate PRISMA 2020 flow diagram, explains the critical difference between the 2009 and 2020 versions, and covers the most common mistakes students make.

• PRISMA stands for “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” — it’s a reporting guideline, not a research methodology
• PRISMA 2020 uses four template variants (not one), depending on whether you search only databases or add grey literature, and whether it’s a new or updated review
• The flow diagram has four main stages: Identification, Screening, Eligibility, and Inclusion
• Reports vs. Studies is the most confusing element — a single study can generate multiple published reports; PRISMA 2020 separates them to prevent counting errors
• Only 30% of systematic reviews cite the PRISMA 2020 guideline, and of those, average adherence is just 42.64% — journals increasingly enforce compliance as a condition of publication
• Google Scholar has a hard 1,000-record ceiling — students must cap their reported search results at this number

What Is a PRISMA Flow Diagram and Why Does It Matter?

A PRISMA flow diagram maps your systematic review process visually. It tells reviewers, readers, and examiners exactly how many papers you found, screened, and ultimately included — and why you excluded others.

In practical terms, it’s the visual evidence of your review’s rigor. A well-constructed diagram answers three critical questions:

1. How did you generate your search? (Identification stage)
2. How did you reduce thousands of results to your final sample? (Screening and Eligibility stages)
3. How many studies did you include and why? (Inclusion stage)

A note on terminology: PRISMA is not a methodology framework like Cochrane or JBI. It’s a reporting guideline — it tells you how to report what you already did. You conduct a systematic review using Cochrane’s methods, then document the process with a PRISMA diagram.

PRISMA 2009 vs PRISMA 2020: What Changed?

Students frequently create diagrams using the outdated PRISMA 2009 template. If your institution or journal requires PRISMA 2020 compliance, you’ll need to understand the differences.

The single biggest change: PRISMA 2020 introduced the “Reports vs. Studies” distinction. Before 2020, students counted “reports assessed” without distinguishing whether one study produced multiple published reports. For example: if you assessed 50 reports, excluded 40, and included 8 studies — and 2 of those 8 studies had 2 reports each — you included 10 reports (not 8). The 2020 template makes this explicit.

Recommendation: Always use the PRISMA 2020 template unless your supervisor explicitly requires the 2009 version. The 2020 updates address real counting errors and are now the standard in all major journals.

The Four Stages of a PRISMA Flow Diagram (Step-by-Step)

Stage 1: Identification

This is your starting point — where you found every potential record.

Step 1: Search all relevant databases (e.g., PubMed, Scopus, Web of Science, EconLit). Record the number of records found in each database.

Step 2: Search other sources — this is your “other sources” arm, which includes:

• Grey literature (theses, reports, conference proceedings)
• Trial registries
• Citation chasing (forward and backward)
• Contacting authors
• Expert searchers

Step 3: Remove duplicates. Record the total number of duplicates across all sources.

⚠️ Common mistake: Students forget the “other sources” arm entirely. If you searched grey literature or ran citation chasing, you must include it. Journals will flag this as a methodological gap.

Step 4: Record the total number of unique records remaining after deduplication. This feeds into the screening stage.

Stage 2: Screening

This is the filter — where you exclude obviously irrelevant records.

Step 1: Take the unique records from the identification stage and begin screening. Most students use software like Rayyan or Covidence for this, but manual screening is acceptable for smaller projects.

Step 2: Screen by title and abstract (and sometimes full text, depending on your resources). Exclude records that clearly don’t meet your inclusion criteria.

Step 3: Record the number excluded at this stage with reasons. This is critical — you must explain why you excluded records. Common reasons include:

• Wrong study design (e.g., qualitative studies when you only wanted RCTs)
• Wrong population (e.g., adults when your review only covered children)
• Wrong outcome (e.g., secondary outcomes instead of primary)
• Wrong language or publication type
• Wrong setting (e.g., high-income countries when your review focuses on low-income)

📌 Pro tip: Use the PRISMA-S extension if your review is a scoping review (the flow diagram structure differs slightly). Use PRISMA-AB if your review is an abstrac review.

Stage 3: Eligibility

This is the deep-dive — where you assess full-text papers for eligibility.

Step 1: Retrieve full-text versions of all papers that passed your screening.

Step 2: Assess each full text against your inclusion criteria. Exclude records that don’t meet all criteria and provide specific reasons for each exclusion.

Step 3: Record the number of full texts assessed and the number excluded, with reasons. You can group exclusion reasons into categories (e.g., “Wrong population: 12”, “Wrong study design: 8”).

Step 4: If one study produces multiple published reports (e.g., a primary trial with separate papers for different outcomes), track this as the “Reports vs. Studies” distinction. Record the number of reports assessed, the number excluded, and the number of studies finally included.

Stage 4: Inclusion

This is the final count — how many studies made it into your synthesis.

Step 1: Record the final number of studies included in the review.

Step 2: Ensure your inclusion stage accurately reflects the “Reports vs. Studies” distinction. If you included 8 studies that produced 10 reports, your diagram shows 8 studies included and 10 reports included.

Step 3: You can include a flow diagram showing the final included studies broken down by subgroup, country, or study design — this is optional but helpful for transparency.

Tools for Creating PRISMA Flow Diagrams

Several tools automate or simplify the PRISMA diagram creation process. Here’s a comparison:

My recommendation: For most student projects, the PRISMA Word Template (CC BY 4.0) combined with manual tracking is sufficient and requires no special software. For larger projects with multiple reviewers, Covidence is the gold standard because it automatically generates the flow diagram and handles deduplication.

Common Mistakes Students Make (And How to Avoid Them)

1. Using the PRISMA 2009 Template

If you’re using the 2009 version, you’re likely excluding “other sources” as a separate arm and not tracking reports vs. studies. This is a methodological gap. Use the 2020 template.

2. Miscounting Duplicates

Students often remove duplicates manually and then lose track of the count. Use deduplication software (Covidence, EndNote, or Zotero) and record the exact number removed before you start screening.

3. Forgetting the “Other Sources” Arm

If you searched grey literature, ran citation chasing, or contacted authors, this must appear as a separate arm in the identification stage. Omission = incomplete search documentation.

4. Google Scholar Over-Reporting

Google Scholar has a hard ceiling of 1,000 records per search. If you retrieved 5,000 results and reported them all, that’s technically impossible. Cap your number at 1,000 and note: “Records shown capped at 1,000 (Google Scholar limit).”

5. Missing Exclusion Reasons

You must provide reasons for exclusions at both the screening and eligibility stages. Generic labels like “not relevant” are insufficient — specify the exact criterion that led to exclusion.

6. Putting the Diagram in the Wrong Section

PRISMA diagrams go in the Methods or Results section of a systematic review. They are part of the reporting structure, not a standalone element.

How to Draw a PRISMA Flow Diagram: Practical Walkthrough

Here’s a concrete example from a hypothetical systematic review on “effectiveness of telehealth interventions for chronic disease management”:

Identification:

• PubMed: 800 records
• Scopus: 650 records
• Web of Science: 420 records
• Grey literature (thesis repository): 150 records
• Total records identified: 2,020

After deduplication: 400 duplicates removed → 1,620 unique records remain

Screening (title/abstract):

• Records screened: 1,620
• Records excluded: 1,450 (wrong population: 320, wrong outcome: 480, wrong study design: 350, wrong language: 100, other: 200)
• Remaining: 170 full-text papers assessed

Eligibility:

• Full-text articles assessed: 170
• Full-text articles excluded: 145 (wrong population: 60, wrong intervention: 45, wrong comparator: 25, wrong outcome: 15)
• Reports vs. Studies: 25 studies produced 32 reports → 7 excluded studies had 14 reports; 8 included studies had 10 reports

Inclusion:

• Studies included in qualitative synthesis: 8
• Reports included in qualitative synthesis: 10

Where Does the PRISMA Diagram Go in Your Paper?

The PRISMA flow diagram typically appears in the Methods section of a systematic review or meta-analysis. It serves as the visual complement to the text describing your search strategy and eligibility criteria.

If your paper is structured with separate “Search Strategy” and “Study Selection” subsections within Methods, place the diagram after “Study Selection” — this mirrors the flow of your reader’s understanding: you describe the process in text, then show it visually.

Final Checklist Before Submitting

• [ ] Used the PRISMA 2020 template (not 2009)
• [ ] Included both database and “other sources” arms in identification
• [ ] Tracked duplicates and recorded their exact count
• [ ] Provided specific exclusion reasons at screening and eligibility stages
• [ ] Separated “reports” from “studies” where applicable
• [ ] Capped Google Scholar records at 1,000 if applicable
• [ ] Placed diagram in Methods or Results section
• [ ] Ensured numbers are consistent across all figures and tables in the paper

What To Know Next

Creating a PRISMA flow diagram is only one part of a rigorous systematic review. To build your understanding further:

• Literature Review Writing: If you’re new to systematic reviews, start with our guide on how to write a literature review for foundational context. The PRISMA diagram is specifically for systematic reviews, not narrative or scoping reviews.
• Systematic Review Process: If you need a complete step-by-step overview of conducting a systematic review before drawing the diagram, our systematic literature review guide walks through the full process, including protocol development, search strategy design, and data extraction.
• Citation Management: If you’re tracking thousands of records, make sure you’re using a citation management tool (like Zotero or EndNote) to keep your library organized. Learn more about citation management tools.

The PRISMA flow diagram is the single most important visual tool in systematic review reporting. It’s not just a figure — it’s your methodological roadmap. Understanding how to create one accurately (using the 2020 template, with proper reports vs. studies tracking, with complete exclusion reasons) is the difference between a review that gets desk-rejected and one that gets published.

If you need help conducting the systematic review itself — from search strategy to data synthesis — our professional writers with subject-matter expertise can support your project. Order your custom academic writing today and get a paper drafted by an expert with a relevant graduate degree.