Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Data, Analytics & AI · STARD Checklist

The STARD Checklist: Reporting Standards for Diagnostic Accuracy Studies

STARD 2015 lists 30 essential items for reporting a diagnostic accuracy study. It was developed after repeated audits found that published diagnostic studies omitted basic information needed to judge whether a test’s reported sensitivity and specificity were trustworthy.

Pubrica AcademyWritten by a diagnostic test methodology writer Reviewed bya biostatistics reviewer Published10 October 2026 Last checked10 October 2026 Reading time8 minutes Sources10 peer-reviewed

STARD (Standards for Reporting of Diagnostic Accuracy Studies) is a reporting guideline for studies that evaluate how accurately a diagnostic test identifies a target condition compared against a reference standard. The current version, STARD 2015, lists 30 essential items authors should report, covering the study's title and abstract, its rationale and objectives, methods, results, and discussion. STARD exists because repeated audits of the published diagnostic accuracy literature found that a large share of studies omitted information, such as how patients were selected, how the index test and reference standard were interpreted, and how missing or uninterpretable results were handled, that is necessary for a reader to judge whether the reported accuracy figures reflect real-world performance or an inflated estimate produced by a biased study design.

Learning objectives

  • The scope of STARD and the number of items specified in the current 2015 version
  • The documented reporting deficiencies in the diagnostic literature that motivated STARD's development
  • How STARD's flow diagram requirement addresses a specific, common source of bias in diagnostic studies

Common misconceptions

  • “Reported sensitivity and specificity estimates can be compared directly across diagnostic studies.” The developers of STARD designed the checklist because studies that select patients differently, apply the reference standard inconsistently, or exclude uninterpretable results without disclosure can report very different sensitivity and specificity estimates for the same test, even when its real-world performance is unchanged. The STARD items enable readers to examine these specific sources of distortion.

Rationale for the development of STARD

STARD was first published in 2003 after a series of methodological reviews found that published diagnostic accuracy studies frequently omitted details needed to assess their validity, including how the study population was selected, whether the people interpreting the index test and reference standard were blinded to each other's results, and how indeterminate or missing test results were handled1. These omissions matter because diagnostic accuracy studies are particularly prone to specific, well-characterized sources of bias, such as spectrum bias (testing a population that does not reflect the clinical population in which the test will be used) and differential verification bias (applying a different or less rigorous reference standard to patients who test negative on the index test), both of which can inflate reported accuracy without any change in the test’s real performance2. The guideline was substantially revised in 2015, reducing the original list to 30 items organized to improve the completeness and transparency of diagnostic accuracy reporting across medical fields3.

Content of the 30 checklist items

STARD 2015's 30 items are organized by manuscript section3. The title, abstract, and introduction items require the study to be identifiable as a diagnostic accuracy study and to state the test's clinical role and the study's objectives or hypotheses. The methods items are the most extensive, requiring a clear description of the study design, the eligibility criteria and setting used to recruit participants, the exact index test and reference standard procedures (detailed enough for replication), the rationale for choosing that reference standard, how test results were defined as positive or negative, whether and how assessors were blinded to each other's results, the planned sample size calculation if one was performed, and the statistical methods used to estimate diagnostic accuracy and to investigate variability in that accuracy across subgroups3. The results items require reporting of participant flow (how many were eligible, included, and excluded, and why), baseline clinical and demographic characteristics, the time interval and any treatment between index test and reference standard, the distribution of the severity of disease in those with the target condition, and a full cross-tabulation of index test results against reference standard results, along with any adverse events from testing. The discussion items require authors to address the study's clinical applicability and limitations explicitly3.

Rationale for the patient flow diagram requirement

Among STARD’s items, the requirement for a patient flow diagram addresses a specific, recurring problem in diagnostic accuracy reporting: studies that exclude, without disclosure, patients with indeterminate, missing or uninterpretable index test or reference standard results, which can substantially change reported sensitivity and specificity if those excluded patients were not evenly distributed between those with and without the target condition23. The flow diagram requires authors to show, step by step, how many participants were initially eligible, how many were enrolled, how many received the index test and reference standard, and how many were excluded at each stage and why, making such selective exclusion visible to readers and reviewers rather than obscured in the text or omitted entirely3. Methodological audits of the diagnostic literature conducted both before and after STARD's introduction have used the presence and completeness of this flow information as one measurable indicator of a study's overall reporting quality4.

Application of STARD in writing and appraising a study

For authors, STARD functions best as a drafting structure completed alongside the manuscript rather than as a retrospective audit, since several items (the reference standard rationale, the blinding arrangements and the planned statistical methods for subgroup accuracy) are considerably easier to report completely if planned from the design stage of the study3. For readers and appraisers, STARD compliance is also one of the inputs used by the QUADAS-2 risk-of-bias tool, since several STARD-reported items (patient selection method, blinding and the interval between tests) map directly onto the QUADAS-2 domains for assessing bias in diagnostic accuracy studies; a well-reported, STARD-compliant paper is therefore also easier to appraise reliably using QUADAS-25. Most major medical journals that publish diagnostic accuracy studies now request a completed STARD checklist as a condition of submission, mapping each of the 30 items to the page or section where it is addressed3.

Independent audits of the diagnostic accuracy literature conducted after STARD’s original 2003 publication found measurable, if incomplete, improvement in reporting completeness at journals that endorsed the checklist compared with those that had not, which constitutes some of the more direct evidence that a reporting guideline can change what authors publish rather than remaining unused guidance6. The same audits, however, also found that adherence remained uneven across items even at endorsing journals, with items requiring more interpretive judgment, such as the rationale for the chosen reference standard and the precise handling of indeterminate results, reported less completely than more mechanical items like sample size or participant counts, a pattern that informed the 2015 revision's attempt to clarify exactly these items6. STARD is also deliberately scoped to studies with a single, clearly defined index test and reference standard, and does not cover two related but distinct designs: case-control diagnostic studies, which compare clearly disease-positive and clearly disease-free groups rather than a representative clinical population and are well documented to inflate accuracy estimates relative to studies using a consecutive, representative sample, and studies developing or validating multivariable diagnostic or prognostic prediction models, for which the TRIPOD statement is the appropriate reporting guideline instead78. Authors working with a prediction-model-based test, rather than a single binary index test, should determine before drafting whether STARD, TRIPOD or, in some cases, both best fit their specific study design, since using the wrong guideline can result in reporting the wrong information entirely rather than merely reporting the correct information incompletely8. A further design-related limitation that STARD’s items are intended to reveal, rather than one that STARD itself introduces, concerns case-control diagnostic accuracy studies, in which clearly disease-positive patients are deliberately compared with clearly disease-free controls rather than a consecutive, representative clinical sample. Lijmer and colleagues’ foundational empirical study found that this design feature was one of the strongest and most consistent predictors of inflated diagnostic accuracy estimates across a large sample of published meta-analyses. For this reason, STARD’s items on study design and patient recruitment ask authors to state explicitly whether a case-control or cohort-type design was used, so that readers can apply this documented caution where appropriate rather than having to infer the design10. The practical implication for authors is that STARD’s items are not neutral formatting requirements; several of them, including this one, exist specifically to make a known, quantified source of distortion visible to the reader rather than obscured in an ambiguous methods section. The EQUATOR Network’s listing for STARD describes the checklist’s purpose in similar terms, as helping readers judge a study’s potential for bias and its applicability: two distinct questions that the 30 items are organized to allow a reader to answer independently of one another, rather than conflating them into a single undifferentiated impression of “quality”.9

Key points

Key points

STARD 2015 specifies 30 reporting items derived from documented deficiencies that allowed diagnostic accuracy studies to overstate real-world performance. It should be used as a drafting structure from the design stage, together with a complete patient flow diagram.

Practical application

Practical application

  1. The STARD items should be planned at the design stage rather than after data collection. Decisions about blinding, the choice of reference standard and planned subgroup analyses are considerably easier to report completely if they are specified before the study begins.
  2. The patient flow diagram should be constructed as data collection proceeds. Each exclusion and its reason should be recorded contemporaneously, so that the flow diagram is complete and accurate rather than reconstructed retrospectively.
  3. STARD items should be cross-checked against QUADAS-2 domains where the study is likely to be appraised or included in a systematic review. Several STARD items correspond directly to the bias domains of QUADAS-2, so complete STARD reporting also supports a more reliable subsequent risk-of-bias assessment.

Frequently asked questions

STARD 2015 contains 30 items, covering the title/abstract, introduction, methods, results, and discussion sections of a diagnostic accuracy study manuscript.

Spectrum bias occurs when a study tests a population that does not reflect the clinical population in which the test will be used, which can inflate or deflate accuracy; STARD’s items on eligibility criteria and study setting are designed to allow readers to check for this.

Yes, STARD requires a flow diagram showing participant numbers at each stage (eligible, enrolled, tested, excluded) with reasons for exclusion.

STARD is a reporting guideline for writing up a diagnostic accuracy study; QUADAS-2 is a risk-of-bias assessment tool used by readers or systematic reviewers to appraise a published study. Several items overlap, since complete STARD reporting provides the information QUADAS-2 assessors need.

STARD was first published in 2003 and substantially revised in 2015, reducing and reorganizing the checklist to its current 30 items.

References

  1. Bossuyt PM, Reitsma JB, Bruns DE, et al. STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. BMJ. 2015;351:h5527. read 10 October 2026. View on BMJ
  2. Whiting PF, Rutjes AWS, Westwood ME, et al. Sources of variation and bias in studies of diagnostic accuracy: a systematic review. Annals of Internal Medicine. 2004;140(3):189-202. read 10 October 2026. View on Annals of Internal Medicine
  3. Cohen JF, Korevaar DA, Altman DG, et al. STARD 2015 guidelines for reporting diagnostic accuracy studies: explanation and elaboration. BMJ Open. 2016;6(11):e012799. read 10 October 2026. View on BMJ Open
  4. Korevaar DA, Cohen JF, Reitsma JB, et al. Updated reporting guidance for diagnostic accuracy studies: STARD 2015. Clinical Chemistry and Laboratory Medicine. 2016;54(8):1321-1323. read 10 October 2026. View on Clinical Chemistry and Laboratory Medicine
  5. Whiting PF, Rutjes AWS, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Annals of Internal Medicine. 2011;155(8):529-536. read 10 October 2026. View on Annals of Internal Medicine
  6. Korevaar DA, van Enst WA, Spijker R, Bossuyt PM, Hooft L. Reporting quality of diagnostic accuracy studies: a systematic review and meta-analysis of investigations on adherence to STARD. Evidence-Based Medicine. 2014;19(2):47-54. read 10 October 2026. View on Evidence-Based Medicine
  7. Rutjes AWS, Reitsma JB, Di Nisio M, Smidt N, van Rijn JC, Bossuyt PM. Evidence of bias and variation in diagnostic accuracy studies. CMAJ. 2006;174(4):469-476. read 10 October 2026. View on CMAJ
  8. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015;350:g7594. read 10 October 2026. View on BMJ
  9. EQUATOR Network. STARD: Standards for Reporting of Diagnostic Accuracy Studies. read 10 October 2026. View on EQUATOR Network
  10. Lijmer JG, Mol BW, Heisterkamp S, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA. 1999;282(11):1061-1066. read 10 October 2026. View on JAMA

Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 10). The STARD Checklist: Reporting Standards for Diagnostic Accuracy Studies. Pubrica. https://pubrica.com/academy/concepts-definitions/stard-checklists/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

STARD 2015 Drafting Checklist

This checklist is intended for use during the design and writing of a diagnostic accuracy study.

[ ] Mapped all 30 STARD items against the current manuscript draft
[ ] Built the patient flow diagram with exclusion reasons tracked from the start of data collection
[ ] Documented the reference standard choice and rationale explicitly
[ ] Confirmed blinding arrangements between index test and reference standard assessors are described
[ ] Checked the statistical methods section covers subgroup accuracy analyses if performed
[ ] Prepared a completed STARD checklist table for journal submission

Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.

How Pubrica can help

Pubrica’s diagnostic research writing team assesses draft manuscripts against the full STARD 2015 checklist and prepares compliant flow diagrams before submission.

Find something else

Search the Pubrica Academy

Access is free and requires no account. Enter a topic, or open the full library to filter by category, sub-topic or resource type.

Pubrica can assess whether a diagnostic accuracy study reports all 30 STARD items.

A scoping assessment is free and takes two business days. It provides a written assessment of what the work requires and which specialist would undertake it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.