PRISMA 2020 Systematic Review: Flow Diagram & Checklist (2026)

Nov 11, 2026

PRISMA 2020 Systematic Review: Flow Diagram, Checklist & Student Guide (2026)

Direct Answer & Quick Summary: What is the PRISMA 2020 Statement for Systematic Reviews?

The PRISMA 2020 Statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), published by Page et al. (2021) in the BMJ, is the internationally mandated methodological reporting framework governing health sciences, biomedical, social science, and postgraduate systematic reviews. PRISMA 2020 replaces the outdated 2009 iteration, introducing an expanded 27-item reporting checklist and an upgraded 4-phase flow diagram that transparently maps study attrition: 1. Identification (database queries, grey literature, and deduplication), 2. Screening (title/abstract filtering and retrieval seeking), 3. Eligibility (full-text evaluation against explicit inclusion/exclusion criteria with documented exclusion justifications), and 4. Included (final corpus synthesized qualitatively or meta-analyzed quantitatively). To achieve a Distinction (First-Class / 80%+) grade in university dissertations, students must integrate pre-registration on PROSPERO, formal risk-of-bias evaluation using validated instruments (such as CASP, Cochrane RoB 2, or ROBINS-I), and transparent multi-database Boolean search strings. Academic researchers at AssignPro Solutions Dissertation Help provide 1-on-1 methodology mentoring, synthesis matrices, and Turnitin-compliant systematic review guidance.

Standard Authority: Page et al. (BMJ 2021; 372:n71) Core Components: 27-Item Checklist & 4-Phase Flow Diagram Target Academic Levels: MSc, MRes, DNP & PhD Dissertations Primary Databases: PubMed, CINAHL, Scopus, Web of Science

Table of Contents (Jump to Section)

1. Methodological Rationale: Narrative Review vs. Systematic PRISMA Review
2. The 4 Stages of the PRISMA 2020 Flow Diagram Deconstructed
3. Protocol Formulation: PICO/PEO Criteria & PROSPERO Registration
4. Database Search Architecture: Boolean Strings, MeSH & Deduplication
5. Quality Appraisal & Risk of Bias: CASP, Cochrane RoB 2 & ROBINS-I
6. Worked Chapter 3 Methodology Walkthrough: Telehealth in Chronic Disease
7. University Dissertation Marking Rubric: Distinction (80%+) vs. Pass
8. 5 Fatal Methodological Errors in Student Systematic Reviews
9. Frequently Asked Questions (FAQs)

1. Methodological Rationale: Narrative Review vs. Systematic PRISMA Review

In postgraduate academic research, markers penalize traditional narrative literature reviews that rely on cherry-picked papers to support preconceived arguments. A narrative review is inherently vulnerable to publication bias, confirmation bias, and subjective selection. In contrast, a systematic literature review structured under the PRISMA 2020 framework is a rigorous, reproducible scientific experiment whose laboratory materials are published peer-reviewed studies.

By defining transparent, prospective search parameters, searching multiple indexed bibliographic databases, tracking study attrition, and objectively scoring methodological quality, PRISMA provides empirical synthesis that stands up to doctoral viva scrutiny and academic peer review. For students exploring methodological foundations, PRISMA aligns directly with the external operational layers of Saunders' Research Onion and the structured data extraction of the Literature Review Synthesis Matrix.

2. The 4 Stages of the PRISMA 2020 Flow Diagram Deconstructed

The centerpiece of every systematic review chapter is the official PRISMA Flow Diagram. The 2020 edition separates records identified from databases/registers from records identified via other methods (e.g. citation chaining, backward searching, grey literature).

PRISMA 2020 flow diagram framework for systematic literature reviews detailing 4 phases

Figure 1: Standardized PRISMA 2020 4-phase study attrition framework detailing screening, eligibility, and exclusion protocols.

Phase	Core Procedural Actions	Documentation & Software Requirements
1. Identification	Run exact Boolean search strings across ≥3 major electronic databases. Export complete citations into reference management software. Execute automated and manual deduplication.	Record exact date of search, total hits per database (e.g. PubMed n=842, CINAHL n=615). Document software used for deduplication (EndNote, Covidence, Rayyan).
2. Screening	Screen remaining unique citations based solely on Title and Abstract against primary inclusion/exclusion criteria. Filter out blatantly irrelevant papers. Seek full texts for remaining corpus.	Dual independent screening is gold-standard. Record total screened (n=S), total excluded (n=E), and number of full-text reports successfully sought vs. unretrievable (n=NR).
3. Eligibility	Conduct thorough, deep reading of all retrieved full-text articles. Assess every paper against granular inclusion/exclusion criteria. Explicitly categorize reasons for every single exclusion.	Crucial for Distinction: Every excluded full-text paper must have a recorded reason: e.g. Wrong study population (n=14), Ineligible intervention (n=9), Lacks control group (n=6).
4. Included	Final synthesis corpus. Extract quantitative or qualitative data into structured extraction forms. Perform methodological quality appraisal and risk-of-bias scoring.	Typically 8 to 25 studies for a Master's dissertation. Detail studies synthesized narratively vs studies pooled statistically into meta-analysis forest plots.

3. Protocol Formulation: PICO/PEO Criteria & PROSPERO Registration

A fatal error in postgraduate research is searching databases prior to locking down the formal review protocol. Under PRISMA 2020 Item 5, authors must indicate whether a review protocol was registered, providing the registration registry and ID number (such as PROSPERO - International Prospective Register of Systematic Reviews, or Open Science Framework [OSF]).

Your review question must be operationalized using an established clinical or academic structuring tool:

PICO Framework (Quantitative / Interventional Reviews): Population (e.g. Type 2 diabetic adults aged ≥65), Intervention (e.g. Nurse-led mobile telehealth coaching), Comparator (e.g. Routine outpatient care), Outcome (e.g. Glycemic control HbA1c reduction & hospital readmission rates). See our guide on PICOT Question Formulation.
PEO / SPIDER Framework (Qualitative / Experiential Reviews): Population (e.g. Intensive care staff nurses), Exposure/Phenomenon of Interest (e.g. Moral distress during triage), Outcome/Theme (e.g. Coping strategies and burnout perception).

4. Database Search Architecture: Boolean Strings, MeSH & Deduplication

PRISMA 2020 Item 7 requires authors to present the full search strategy for all databases, registers, and websites, including filters and limits. University examiners test whether your search string can be copied, pasted into PubMed or CINAHL, and return the exact same citation volume.

Exemplar Advanced Search String (PubMed / MEDLINE Syntax):

("Diabetes Mellitus, Type 2"[Mesh] OR "Type 2 Diabetes"[Title/Abstract] OR "T2DM"[Title/Abstract])
AND
("Telemedicine"[Mesh] OR "Telehealth"[Title/Abstract] OR "mHealth"[Title/Abstract] OR "Digital Health"[Title/Abstract] OR "Mobile Application"[Title/Abstract])
AND
("Glycated Hemoglobin A"[Mesh] OR "HbA1c"[Title/Abstract] OR "Patient Readmission"[Mesh] OR "Self-Management"[Title/Abstract])
NOT
("Editorial"[Publication Type] OR "Letter"[Publication Type] OR "Animal"[Mesh])

Best Practice Note: Include truncation asterisks (e.g. tele*, diabet*) to capture morphological word variants. Restrict publication date ranges (e.g., past 5–10 years) only when supported by technological or policy shifts.

5. Quality Appraisal & Risk of Bias: CASP, Cochrane RoB 2 & ROBINS-I

A literature review that accepts all published findings as equally valid is unscientific. PRISMA 2020 Item 11 mandates assessment of the risk of bias in included studies. You must select an appraisal tool validated for the specific study designs in your review:

Cochrane RoB 2 (For RCTs)

Gold standard for randomized controlled trials. Evaluates 5 domains: randomization bias, deviations from intended interventions, missing outcome data, outcome measurement bias, and selective reporting.

Low / Some Concerns / High Risk

ROBINS-I (Non-Randomized Studies)

Mandatory for cohort and quasi-experimental designs. Evaluates confounding, participant selection, classification of interventions, deviations, missing data, and reported results.

Low / Moderate / Serious / Critical

CASP Checklists (Mixed/Qualitative)

Critical Appraisal Skills Programme (CASP) 10-question checklists for qualitative studies, cohort studies, and case-control studies. Widely favored in nursing and health sciences dissertations.

10-Point Scoring / Yes-Can't Tell-No

6. Worked Chapter 3 Methodology Walkthrough: Telehealth in Chronic Disease

Dissertation Empirical Walkthrough: Numbers, Screening & Attrition Accounting

Research Objective: "Evaluate the clinical effectiveness of mobile app-based nurse monitoring on HbA1c reduction in adults with Type 2 Diabetes Mellitus: A Systematic Literature Review."

Attrition Data Mapping for PRISMA Flow Diagram:

1. Database Identification (Total = 1,482 records):
• PubMed / MEDLINE: 524 records
• CINAHL Complete: 486 records
• Scopus: 362 records
• Web of Science Core Collection: 110 records

2. Deduplication:
• Total duplicates removed via Covidence auto-de-duplication & manual verification: 412 duplicates
• Total unique citations advancing to title/abstract screening: 1,070 records

3. Title & Abstract Screening:
• Total records excluded based on Title/Abstract: 982 records (reasons: animal models n=45, Type 1 diabetes n=180, pediatric cohorts n=220, non-digital interventions n=537)
• Total full-text articles sought for retrieval: 88 reports
• Reports unable to retrieve (inter-library loan unfulfilled): 3 reports

4. Full-Text Eligibility Assessment (85 articles assessed):
• Total full-text articles excluded: 71 articles with recorded, transparent justifications:
  – Ineligible intervention: SMS text-only without nurse feedback (n = 26)
  – Ineligible study design: cross-sectional or protocol paper only (n = 19)
  – Ineligible outcome: satisfaction only, no biochemical HbA1c measured (n = 15)
  – Non-English publication without verified translation (n = 7)
  – Intervention duration under 12 weeks (insufficient for HbA1c turnover) (n = 4)

5. Final Included Synthesis Corpus:
• Final peer-reviewed studies included in systematic review synthesis: 14 studies (10 RCTs, 4 prospective quasi-experimental studies). Full extraction executed via Literature Review Synthesis Matrix.

7. University Dissertation Marking Rubric: Distinction (80%+) vs. Pass

Criterion	Distinction Grade (80%+)	Pass Grade (50–59%)
1. PRISMA 2020 Compliance	Adheres fully to PRISMA 2020 27-item checklist; presents an immaculate 4-stage flow diagram with perfect mathematical attrition accounting across every branch.	Uses outdated PRISMA 2009 template or presents flow diagram with mathematical discrepancies between screening counts and inclusions.
2. Search String Reproducibility	Appends complete, reproducible search histories for every database with full MeSH/Emtree headings, Boolean operators, and dates of execution in appendix.	Lists generic keywords only (e.g. "searched diabetes and telemedicine") without syntax or reproducible Boolean strings.
3. Exclusion Transparency	Itemizes exact, categorized reasons for all full-text exclusions; includes full exclusion table in appendices linking citations to specific criteria.	States vague exclusions (e.g. "71 papers excluded because they were not relevant") without categorized counts or justifications.
4. Critical Appraisal Rigor	Applies validated critical appraisal tool (CASP, RoB 2); presents comprehensive risk-of-bias summary table and critically weaves study quality into thematic discussions.	Appraisal is superficial or checklist-only; fails to evaluate how methodological weaknesses impact the credibility of review findings.

8. 5 Fatal Methodological Errors in Student Systematic Reviews

Error 1: Searching Only One Database (e.g., PubMed Alone): PRISMA standards strictly prohibit single-database searches. To avoid indexing bias, you must search at least three major multidisciplinary and subject-specific databases (e.g. PubMed/MEDLINE, CINAHL Complete for nursing/allied health, and Scopus or Web of Science).
Error 2: Mathematical Discrepancies in the Flow Diagram: University markers audit PRISMA numbers with a calculator. If you report 1,070 records screened, 982 excluded, but list 85 full texts sought (1,070 − 982 = 88, leaving 3 missing), your review fails reproducibility checks. Numbers must balance perfectly across every box.
Error 3: Failing to Report Reasons for Full-Text Exclusions: In the Title/Abstract screening phase, reasons are not mandatory. But in the Full-Text Eligibility Phase, PRISMA 2020 Item 10 mandates reporting the exact number of articles excluded for each specific exclusion reason.
Error 4: Retrospectively Altering Inclusion Criteria to Fit Desired Studies: Modifying eligibility criteria mid-review to fit a favorite paper violates scientific rigor. Any post-hoc changes must be explicitly disclosed and justified in the methodology chapter as protocol amendments.
Error 5: Confusing Systematic Reviews with Scoping Reviews: Systematic reviews synthesize evidence to answer a focused, narrow clinical/empirical question (using PICO). Scoping reviews map broad concepts, clarify definitions, and identify knowledge gaps (using PRISMA-ScR). Using systematic review appraisal tools on scoping reviews is an instant red flag for examiners.

9. Frequently Asked Questions (FAQs)

Q1: How many studies should be included in a Master's or PhD systematic review?

There is no arbitrary minimum or maximum required study count. The final number is the natural empirical output of your search strings and eligibility criteria. However, in postgraduate dissertations, most focused systematic reviews yield between 8 and 25 included studies. If your search yields 3 studies, your criteria may be too narrow; if it yields 150 studies, your criteria are too broad for a student dissertation.

Q2: What is the difference between PRISMA 2009 and PRISMA 2020?

PRISMA 2020 reflects significant advances in systematic review methodology. Key changes include: separate flow diagram pathways for internal database searches vs external searches (citation chaining, grey literature), dedicated reporting of study registration (Item 5), reporting automation tools used for screening (e.g. AI tools in Rayyan/Covidence), and reporting full search strategies for all databases rather than just one.

Q3: Can a student conduct a systematic review independently without a second reviewer?

While Cochrane and international publishing standards mandate dual independent screening to minimize selection bias, university dissertations are individual academic works. For Master's dissertations, universities typically allow single-student screening provided you explicitly discuss single-reviewer screening as a methodological limitation in Chapter 5, or pilot a 10% random sample with your supervisor or a peer to calculate inter-rater reliability (Cohen's Kappa).

Q4: Does every systematic review require a meta-analysis?

No. A meta-analysis is a statistical technique that pools numerical data from homogeneous studies. If your included studies exhibit high clinical or methodological heterogeneity (e.g. differing intervention dosages, follow-up times, or outcome metrics), statistical pooling is inappropriate. In such cases, PRISMA 2020 mandates a narrative synthesis guided by the SWiM (Synthesis Without Meta-analysis) reporting guideline.

Need Expert Guidance on Your Systematic Review Dissertation?

Overwhelmed by database search strings, PRISMA flowcharts, or CASP critical appraisal? Our PhD-level academic research specialists at AssignPro Solutions provide personalized protocol design, synthesis matrix building, and dissertation methodology coaching.

Explore Dissertation & Thesis Consulting Services