SANRA Scale Self-Assessment: Score Your Narrative Review on the Six SANRA Items Before Submission

SANRA, the Scale for the Assessment of Narrative Review Articles, is a six-item instrument for judging the scientific quality of a non-systematic review [1]. Three editors at the general medical journal Deutsches Ärzteblatt developed it between 2010 and 2017, through two earlier seven-item versions, and published the final scale in 2019 [1]. Each of its six items is rated 0, 1 or 2, giving a total SANRA score from 0 to 12.

In the authors’ test on 30 consecutive narrative reviews, the intraclass correlation between three raters was 0.77 (95% CI 0.57 to 0.88) and Cronbach’s alpha was 0.68 [1]. The raters were the scale’s own authors, so reliability may be lower in less familiar hands. SANRA was built to help editors, reviewers and readers, and the developers also encourage authors to use it when preparing a narrative review [1]. That makes it a natural self-assessment tool: running your manuscript through the six items before submission turns a vague worry about narrative review quality into a score you can act on.

The scale assesses six items [1]:

1. Justification of the article’s importance for the readership. Does the review explain why the topic matters now?
2. Statement of concrete aims or formulation of questions. Is the purpose explicit rather than implied?
3. Description of the literature search. Are the sources, search terms and approach described, even briefly?
4. Referencing. Are key statements supported by appropriate, verifiable citations?
5. Scientific reasoning. Is the evidence, and its level of evidence, weighed rather than opinion asserted?
6. Appropriate presentation of data. Are relevant outcome data reported, with numbers, tables and figures where they help?

Together these items separate a sourced synthesis from an unsupported essay, which is what review appraisal needs to capture. They also mirror long-standing advice on writing narrative reviews: state the purpose, describe how the literature was found, and support claims with sources [2].

Each item is scored 0 for a low standard, 2 for a high standard and 1 for anything in between [1]. The developers publish separate explanations and examples for each item so that scoring is as consistent as possible. Summing the six items gives the SANRA score. When you self-assess, score against those anchors, not against how you hope the manuscript reads. The value of the exercise comes from catching the items you would rather not admit are a 0 or a 1.

SANRA has no established cut-offs for grades of quality, and its authors advise against comparing scores across different settings. They do give one marker from experience: a score of 4 or below indicates very poor quality [1]. Beyond that, read the profile as well as the sum. A respectable total can hide a single 0, and a 0 on referencing or the literature search is the kind of weakness reviewers notice first. A useful working target is a high total with no item below 1.

The point of self-scoring is repair. For a low score on justification, open with why the topic matters to the journal’s readers now; for aims, turn an implied purpose into one or two explicit questions. Fix a weak literature-search item with a short paragraph naming the databases, search terms and date range you used [2]. Address thin referencing by sourcing every substantive claim, weak scientific reasoning by stating the level of evidence behind key statements, and poor data presentation by reporting the outcome data that matter, moving comparisons into a table or figure [2]. Then score the manuscript again and check that no zeros remain.

Print the worksheet, score your draft, and note one fix for every item below 2. If you want a second opinion, ask a co-author to score the manuscript independently and compare: the items where you disagree are usually the ones a peer reviewer will question. For the next steps after scoring, see our guide on how to publish a literature review.

Used before submission, the SANRA scale turns the goal of “a good narrative review” into six measurable items and one 0–12 score. The workflow is simple and repeatable: score the six items honestly, read the profile as well as the sum, repair every item rated 0 or 1, and score again. That loop raises your narrative review quality on the same criteria an editor using the SANRA checklist will apply.