Skip to main content
v2026.11,986 entries · CC-BY 4.0

Melnyk Level of Evidence: The Seven Levels (I-VII) and How to Cite Them

Melnyk’s seven levels of evidence, I to VII, with 2011, 2015 and 2023 wording compared and a worked appraisal-table example.

Written and maintained by CASRAI Editorial Board

Published

On this page

The Melnyk levels of evidence are a seven-level scale (Level I to Level VII) that nursing and healthcare evidence-based practice texts use to rank a study by its design: Level I is a systematic review or meta-analysis of randomized controlled trials, Level II a randomized controlled trial, and Level VII the opinion of authorities or expert-committee reports. The scale comes from Melnyk and Fineout-Overholt’s textbook, and the wording of some levels differs between editions, so a paper should name the edition it follows.

Purpose of the section: what the Melnyk level tells a reader

A level of evidence labels the design of a single study or review. It answers one question: what kind of study produced this finding? An evidence-based practice (EBP) paper, a critical appraisal table or a clinical-question assignment uses the level to show the reader, at a glance, how each cited source was generated before the paper weighs what the source found. The scale is presented in the library guides that nursing programs use as a ranking from systematic reviews of randomized trials at the top to expert opinion at the bottom, and EBSCO’s research overview describes seven-level hierarchies of this kind as among the most widely adopted versions of the hierarchy of evidence.

Melnyk levels of evidence

The plural form in searches refers to the whole seven-level scale. Every source read for this page presents it as Level I through Level VII, attributed to Bernadette Melnyk and Ellen Fineout-Overholt, Evidence-Based Practice in Nursing and Healthcare: A Guide to Best Practice, published by Wolters Kluwer (earlier editions by Lippincott Williams and Wilkins).

Melnyk’s level of evidence

The possessive form usually means the level assigned to one particular study. When your appraisal table says a trial is “Level II”, you are applying the scale to that study; you are not describing its quality. Design and quality are separate judgments, as the mistakes section below explains.

The seven levels in order

The list below follows the 3rd edition (2015) wording as reproduced by the University of Colorado Anschutz Medical Campus library guide, which cites Box 1.3, page 11 of that edition. Where the 5th edition (2023) words a level differently, the difference is given in parentheses, taken from the Winona State University library guide to the 5th edition.

  1. Level I: systematic reviews and meta-analyses of randomized controlled trials. (5th edition: evidence from a systematic review or meta-analysis of all relevant randomized controlled trials.)
  2. Level II: a randomized controlled trial. (5th edition: at least one well-designed randomized controlled trial, for example a large multi-site trial.)
  3. Level III: a non-randomized controlled trial, also called a quasi-experiment. (5th edition: single well-designed controlled trials without randomization, or a systematic review of a complete body of evidence that integrates higher and lower levels, or mixed-methods intervention studies.)
  4. Level IV: case-control or cohort studies. (5th edition: well-designed case-control or cohort studies.)
  5. Level V: systematic reviews of qualitative or descriptive studies. (5th edition: systematic reviews of descriptive and qualitative studies, described as meta-synthesis.)
  6. Level VI: qualitative or descriptive studies. (5th edition: a single descriptive or qualitative study, and also evidence-based practice, evidence-based quality improvement and quality improvement projects.)
  7. Level VII: opinion of authorities and/or reports of expert committees. (5th edition: reports from committees of experts, and narrative and literature reviews.)

The 5th edition guide labels the levels with Arabic numerals (1 to 7) while the earlier-edition guides use Roman numerals. The numbering is the same scale; copy the style your assignment, journal or institution uses and keep it consistent in one document.

How the wording differs by edition

Three sources were compared for this page: the Colorado Anschutz guide (3rd edition, 2015), the Arizona State University nursing guide (which cites a 2011 Lippincott Williams and Wilkins edition without an edition number) and the Winona State guide (5th edition, 2023). The data table below sets the wording of each level side by side. Four points matter when you cite a level.

Levels I, II, IV and V keep the same design logic. Systematic reviews of randomized trials, randomized trials, cohort and case-control studies, and reviews of descriptive and qualitative studies appear in the same positions in all three sources. Only small wording changes appear, such as “well-designed” being attached to the trial and cohort levels.

Level III is broader in the 5th edition. The 2011 and 2015 wording gives only controlled trials without randomization at this level. The 5th-edition guide adds two further descriptions to the same level: a systematic review of a complete body of evidence that integrates higher and lower levels (an integrative review), and mixed-methods intervention studies. If your source study is an integrative review or a mixed-methods intervention study, the level you assign depends on which edition you cite.

Level VI is broader in the 5th edition. Earlier wording is descriptive and qualitative studies. The 5th-edition guide also places evidence-based practice, evidence-based quality improvement and quality improvement projects at Level VI, which matters for nurses appraising practice-change reports from clinical units.

Level VII also adds narrative reviews in the 5th edition. The Winona State guide lists “narrative and literature reviews” alongside committee reports at Level VII. The earlier guides name only authority opinion and expert-committee reports.

Because the library guides adapt the book’s wording and the book itself is the primary source, check any level you are unsure about against the edition your course or journal requires. The sources section lists the guides read and the date they were read.

How to cite and report a Melnyk level in an EBP paper

Treat the level as a field of your evidence table, next to the study design and the finding. A workable pattern, which is editorial guidance and not a rule stated by the sources, is:

  1. State once, in the methods or the table notes, which scale you used and which edition: “Levels of evidence follow Melnyk and Fineout-Overholt (5th ed., 2023).” Add the book to your reference list in the citation style your program uses (see the CASRAI guide on levels of evidence, OCEBM and GRADE for how other scales are reported).
  2. Give each study its design in plain words, then its level: “randomized controlled trial, Level II”. Naming the design lets a reader who uses another scale translate your label.
  3. Assign the level from the study’s design as the authors describe it, not from the journal or from how impressive the result is. A retrospective cohort study is a cohort study whatever its sample size.
  4. Where a review is the source, say what it reviewed. A systematic review of randomized trials is Level I; a systematic review of qualitative studies is Level V. The word “systematic” alone does not fix the level (see the dictionary entry for systematic review).

A short annotated example

Illustrative composite: the studies below are invented for demonstration and are not real publications. A nurse writing a clinical question about early mobility after surgery collects four sources and records them in an appraisal table.

Source (illustrative) Design as the authors describe it Level (Melnyk, 5th ed.)
Review A Systematic review and meta-analysis of all relevant randomized controlled trials 1
Trial B Single multi-site randomized controlled trial 2
Cohort study C Prospective cohort study with a comparison group 4
Unit report D Quality improvement project on one clinical unit 6

Annotations: Review A earns the top level because it pools randomized trials, not because it is a review. Trial B sits one level lower because it is a single trial. Cohort study C is placed at Level 4 because it observes groups without randomizing them. Unit report D is at Level 6 under the 5th-edition wording that places quality improvement projects there; under the 2015 wording an author would need to justify the level some other way, which is why the edition statement in the table notes matters. The table records design only: it says nothing about whether any of the four was well conducted.

Common mistakes

  • Treating the level as a quality score. The levels rank design. The scales’ own descriptions attach “well-designed” to several levels, which signals that a separate judgment of conduct is still needed. Appraise the study’s validity separately from assigning the level.
  • Citing the wrong edition or none. Level III, VI and VII wording changed in the 5th edition. A paper that borrows a level table from a library guide without naming the edition can place a mixed-methods study or a quality improvement project inconsistently.
  • Mixing scales in one table. A column that uses Melnyk Roman numerals for some rows and OCEBM levels for others cannot be read. Pick one scale per table and, if a reviewer asks for the other, compare them using the CASRAI guides on the hierarchy of evidence pyramid and OCEBM and GRADE.
  • Calling every review Level I. Only reviews of randomized trials sit at Level I. Reviews of descriptive and qualitative studies are Level V, and narrative reviews appear at Level VII in the 5th-edition wording.
  • Mixing Roman and Arabic numerals. The numbering is the same, but one document should use one style.
  • Leaving out the design. A table with only “Level IV” and no design forces the reader to trust the label.

Checklist before you submit

  • The scale and the edition are named once, and the book is in the reference list.
  • Every cited study has its design in words and its level in the same table.
  • Reviews are leveled by what they review (randomized trials, or descriptive and qualitative studies).
  • Mixed-methods studies, integrative reviews and quality improvement projects are leveled according to the edition you cite.
  • One numeral style is used throughout.
  • Level and study quality are reported as separate judgments.
  • If your course also uses a competency framework for evidence-based practice, see the guides on EBP competencies for nurses and nurse leaders and the ARCC model for EBP mentorship, which come from the same research group.

Sources

  • University of Colorado Anschutz Medical Campus, Strauss Health Sciences Library, “Evidence: Levels of Evidence” (cites Melnyk and Fineout-Overholt, Evidence-based practice in nursing and healthcare: A guide to best practice, 3rd ed., 2015, Box 1.3, p. 11): https://library-cuanschutz.libguides.com/Evidence/Levels, checked 2026-11-15.
  • Winona State University Library, EBP Toolkit, “Levels of Evidence” (cites Melnyk and Fineout-Overholt, 5th ed., Wolters Kluwer, 2023): https://libguides.winona.edu/ebptoolkit/Levels-Evidence, checked 2026-11-15.
  • Arizona State University Library, Nursing guide, “Evidence-Based Practice” (cites Melnyk and Fineout-Overholt, 2011, Lippincott Williams and Wilkins): https://libguides.asu.edu/nursing/EBP, checked 2026-11-15.
  • EBSCO Research Starters, “Hierarchy of Evidence”: https://www.ebsco.com/research-starters/health-and-medicine/hierarchy-evidence, checked 2026-11-15.

Data: Melnyk and Fineout-Overholt levels of evidence: wording by source edition

Side-by-side wording of the seven levels of evidence as reproduced by three university library guides citing the Melnyk and Fineout-Overholt textbook (a 2011 edition, the 3rd edition of 2015 and the 5th edition of 2023). Compiled by aligning each guide's level descriptions row by row.

Period covered: 2011-2023.

Melnyk and Fineout-Overholt levels of evidence: wording by source edition
Level2011 wording (ASU guide)3rd edition 2015 wording (CU Anschutz guide)5th edition 2023 wording (Winona State guide)
Level ISystematic reviews or meta-analyses of RCTsSystematic reviews and meta-analysis of randomized controlled trialEvidence from a systematic review or meta-analysis of all relevant RCTs
Level IIWell-designed RCTsRandomized controlled trialAt least one well-designed RCT (e.g. large multi-site RCT)
Level IIIWell-designed, controlled trials without randomizationNon-randomized controlled trial (quasi-experiment)Single well-designed controlled trials without randomization (quasi-experimental); or systematic review of a complete body of evidence (integrative review of higher and lower evidence); or mixed methods intervention studies
Level IVWell-designed cohort and case-control studiesCase-control or cohort studiesWell-designed case-control or cohort studies
Level VSystematic reviews of descriptive and qualitative studiesSystematic reviews of qualitative or descriptive studiesSystematic reviews of descriptive and qualitative studies (meta-synthesis)
Level VIDescriptive and qualitative studiesQualitative or descriptive studiesSingle descriptive or qualitative study, EBP, EBQI and QI projects
Level VIIExpert consensus reportsOpinion of authorities and/or reports of expert committeesOpinion of authorities and/or reports from committees of experts and narrative and literature reviews

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Melnyk Level of Evidence: The Seven Levels (I-VII) and How to Cite Them

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →