Reporting Guidelines for Veterinary Research: CONSORT, STROBE, and Beyond

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• Reporting guidelines such as CONSORT (randomised trials), STROBE (observational studies), and ARRIVE 2.0 (in vivo animal experiments) are crucial for enhancing the transparency, reproducibility, and interpretability of veterinary research by specifying essential methodological details for manuscript publication.
• The EQUATOR Network serves as a central repository for these guidelines, offering a searchable library and tools to help veterinary researchers select the most appropriate standard based on their study design, thereby ensuring accurate assessment of internal validity and potential biases.
• ARRIVE 2.0 distinguishes between an "Essential 10" minimum reporting requirement and a "Recommended Set," prioritizing critical items like study design, sample size determination, randomisation, blinding, and outcome assessment for in vivo experiments to improve scientific rigor.
• Veterinary researchers must adapt human-focused guidelines like CONSORT and STROBE to account for species-specific characteristics, diverse production systems, and practical challenges such as owner compliance and ethical constraints, with specific extensions like REFLECT addressing livestock trials.
• Adherence to reporting guidelines is a prospective process, requiring study design to anticipate checklist items from the outset rather than retrofitting documentation during manuscript preparation, thereby facilitating accurate reporting of crucial elements like allocation concealment, blinding procedures, and participant flow.
• Common reporting failures include superficial checklist compliance without substantive adherence, guideline substitution (e.g., applying CONSORT to observational studies), and selective reporting of outcomes or harms, necessitating careful review against raw data and study protocols.

Veterinary research spans interventional trials, observational studies, systematic reviews, and in vivo experiments. Each design carries distinct sources of bias, and each requires a corresponding reporting standard so that readers can judge internal validity, replicate methods, and compare findings across populations. This article examines the reporting guidelines most relevant to veterinary investigators, with emphasis on CONSORT for randomised trials, STROBE for observational studies, and ARRIVE for animal experiments, and it describes how the EQUATOR Network functions as the central repository for these instruments. The intended reader is a veterinary researcher preparing a manuscript, reviewing a peer's work, or designing a study with prospective reporting in mind. The article does not cover statistical analysis methods in detail.

Reporting guidelines are not study conduct standards. They specify what must appear in a manuscript, not how the study must be performed. A poorly designed study reported completely will still be a poor study, but readers will be able to see why it is poor. That transparency serves the literature better than a polished report that conceals methodological weaknesses. For veterinary medicine, where sample sizes are often small, outcome measures are heterogeneous, and blinding is frequently difficult, complete reporting is the precondition for any meaningful synthesis of evidence across studies.

At a Glance

The Scientific Rationale for Reporting Standards

The argument for reporting guidelines rests on a simple premise: a research report must contain enough information for an independent reader to reconstruct what was done and to assess whether the conclusions follow from the data. In animal research, the original ARRIVE guidelines were developed in 2010 to address persistent deficiencies in the reporting of in vivo experiments, including missing information on animal characteriztics, housing conditions, randomisation, and blinding Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. The guidelines were published simultaneously across multiple journals, reflecting a coordinated editorial commitment to raise reporting standards Guidelines for reporting experiments involving animals: the ARRIVE guidelines.

The underlying logic is cumulative. Individual studies feed into systematic reviews, which inform clinical guidelines and regulatory decisions. If primary reports omit key methodological details, reviewers cannot assess risk of bias, cannot perform subgroup analyzes, and may inadvertently combine studies that are not comparable. In veterinary medicine, this problem is amplified by the diversity of species, breeds, production systems, and regulatory environments. A trial in dairy cattle under intensive management may not be generalizable to a pasture-based system, but that judgment is impossible if the manuscript does not describe the management context.

ARRIVE and Its Revision

The ARRIVE guidelines (Animal Research: Reporting of In Vivo Experiments) were first published in 2010 as a checklist of information to include in publications describing in vivo experiments Animal research: reporting in vivo experiments: the ARRIVE guidelines. The purpose was to enable others to scrutinise the work adequately, evaluate its methodological rigour, and reproduce the methods and results. Despite widespread endorsement by funders and journals, adherence remained inconsistent, and the anticipated improvements in reporting quality were not achieved.

The revision process leading to ARRIVE 2.0 used a Delphi exercise to prioritize checklist items and divided them into two sets The ARRIVE guidelines 2.0: Updated guidelines for reporting animal research. The Essential 10 constitutes the minimum requirement and includes items such as study design, sample size, inclusion and exclusion criteria, randomisation, blinding, and outcome assessment. The Recommended Set describes the research context, including ethical statements, housing and husbandry, and interpretation of results. This division supports a stepwise implementation approach, allowing journal editors and reviewers to verify that the most important items are reported before considering the fuller context.

An accompanying Explanation and Elaboration document provides the rationale and supporting evidence for each of the 21 items in ARRIVE 2.0, along with examples of good reporting from the published literature Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0. The document also covers advice on study design and conduct, supporting researchers from the experimental design stage through to publication. The current version of the guidelines is maintained by the NC3Rs and specifies the minimum information required for transparent and reproducible animal research publications ARRIVE Guidelines 2.0 for Reporting Animal Research.

CONSORT and STROBE in Veterinary Contexts

The CONSORT statement for randomised trials and the STROBE statement for observational studies were developed primarily for human clinical research, but their principles transfer directly to veterinary studies. The EQUATOR Network maintains the full library of these guidelines and their extensions, including REFLECT for livestock trials EQUATOR Network Reporting Guidelines. Veterinary researchers should consult the EQUATOR database early in the manuscript preparation process to identify whether a design-specific extension exists.

CONSORT centers on a 25-item checklist and a participant flow diagram. For veterinary trials, the flow diagram must account for animals that fail screening, are withdrawn for welfare reasons, or are lost to follow-up. The checklist requires explicit reporting of how the sample size was determined, which is particularly important in veterinary studies where financial and ethical constraints often limit recruitment. STROBE provides separate checklists for cohort, case-control, and cross-sectional designs, with attention to how participants or animals were selected, how variables were defined, and how missing data were handled.

The EQUATOR Network as an Access Point

The EQUATOR Network functions as the primary international repository for reporting guidelines across health research EQUATOR Network Reporting Guidelines. Its library includes CONSORT, STROBE, PRISMA, ARRIVE, and REFLECT, along with extensions for specific designs such as cluster randomised trials and diagnostic accuracy studies. The database is searchable by study type, and each entry links to the original guideline publication, the checklist, and any accompanying explanation documents. Veterinary researchers should treat EQUATOR as the first stop when determining which guideline applies to a planned manuscript.

Selecting the Appropriate Guideline for a Study Design

The first decision in reporting a veterinary study is not which checklist to complete, but which guideline matches the study architecture. Applying the wrong standard produces a misleading compliance record and distracts reviewers from the substantive weaknesses in a manuscript. The EQUATOR Network reporting guideline library functions as the definitive index for this selection step, cataloguing more than 500 reporting standards across study types.

For a randomised controlled trial, CONSORT applies. For an observational study using cohort, case-control, or cross-sectional designs, STROBE applies. For a diagnostic accuracy study, STARD applies. For any in vivo experiment, whether interventional or observational, ARRIVE applies as a baseline expectation. The overlap is intentional. A randomised trial in dogs that measures a diagnostic test's accuracy should report against CONSORT, STARD, and ARRIVE simultaneously, with each checklist contributing distinct items.

The selection logic changes when the study is not a clinical trial in the conventional sense. Field vaccine efficacy trials, production medicine interventions, and herd-level studies often randomise at the group or pen level instead of the individual animal level. The REFLECT statement, available through the EQUATOR Network, addresses this cluster-randomised structure and should replace or supplement CONSORT where unit of allocation differs from unit of analysis.

Species and production system alter the practical application of these checklists. A laboratory study using purpose-bred rodents under controlled conditions can report environmental parameters with precision. A clinical trial in client-owned horses must accommodate variable management, owner compliance, and out-of-hours events. The reporting burden does not decrease in the clinical setting. It shifts toward documenting what could not be controlled and how that was handled analytically.

The ARRIVE Essential 10 as a Working Minimum

The ARRIVE 2.0 revision restructured the original 20-item checklist into two tiers. The Essential 10 constitutes the minimum requirement for any manuscript describing in vivo experiments, while the Recommended Set provides context that strengthens interpretation. The ARRIVE 2.0 guidelines describe this division as a stepwise approach that helps editors and reviewers verify that the most important items are present before deeper evaluation.

The Essential 10 items are: study design, sample size, inclusion and exclusion criteria, randomisation, blinding, outcome measurement, statistical methods, experimental animals, experimental procedures, and results. Each item demands specific content. The sample size item, for example, requires a statement of how the number of animals was determined, including the effect size of interest and the variability estimate used. A sentence stating that sample size was calculated using standard methods does not satisfy the item.

The explanation and elaboration document for ARRIVE 2.0 provides the rationale and worked examples for each item. It also covers design and conduct best practice, which matters because reporting guidelines cannot rescue a study whose design flaws are structural. The document explicitly connects reporting quality to the reproducibility of biomedical research, arguing that transparent reporting allows readers to assess reliability and repeat the work.

For veterinary researchers, the Essential 10 items that generate the most frequent deficiencies are randomisation and blinding. Clinical caseloads, owner preferences, and ethical constraints often make true random allocation difficult. The reporting obligation is to describe exactly what was done, including the method of allocation concealment and the personnel responsible. If blinding was impossible, the manuscript must say so and explain the risk of bias this introduces.

CONSORT in Veterinary Clinical Trials

The CONSORT statement, originally developed for human medicine, transfers to veterinary randomised trials with specific adaptations. The EQUATOR Network hosts the current CONSORT checklist and flow diagram templates. The 25-item checklist covers title and abstract, introduction, methods, randomisation, blinding, participant flow, recruitment, baseline data, numbers analyzed, outcomes and estimation, ancillary analyzes, harms, limitations, generalizability, and interpretation.

The participant flow diagram requires particular attention in veterinary trials. It must track animals from enrollment through allocation, treatment, follow-up, and analysis. Losses occur at distinct points in veterinary studies: owner withdrawal, adverse events requiring rescue medication, protocol violations, and death or euthanasia unrelated to the study condition. Each loss category must be quantified separately, because they carry different implications for bias.

Veterinary-specific reporting obligations under CONSORT include species, breed, age, sex, neuter status, body condition score, and relevant comorbidities. The MSD Veterinary Manual provides standardized definitions for these baseline characteriztics, which supports consistent reporting across studies. Breed predispositions and species differences in drug metabolism mean that baseline comparability between arms is not a formality but a substantive scientific claim.

The harms item in CONSORT is frequently underreported in veterinary trials. Adverse events in animals are often managed by the attending clinician instead of documented as study outcomes. The checklist requires reporting of all important harms in each group, including the frequency and severity of events. This includes events that led to withdrawal from the study, because selective removal of animals with poor outcomes distorts the treatment effect estimate.

STROBE and Observational Veterinary Research

Observational studies dominate veterinary clinical research because randomised trials are expensive, ethically constrained, and often impossible for rare conditions or client-owned populations. The STROBE statement provides a 22-item checklist covering cohort, case-control, and cross-sectional designs. The EQUATOR Network lists STROBE extensions for specific designs, including molecular epidemiology and genetic association studies.

The STROBE items that veterinary researchers most often miss are the description of the study setting, the eligibility criteria, and the sources and methods of participant ascertainment. In veterinary medicine, the study setting is rarely a single homogeneous population. Referral hospital caseloads differ from primary care populations. Production animal studies draw from herds with distinct management systems, biosecurity status, and endemic disease profiles. Each of these factors modifies the generalizability of the findings.

The World Organization for Animal Health terrestrial animal health standards provide a framework for describing surveillance populations and disease status that aligns with STROBE reporting. For studies involving notifiable diseases or international movement of animals, the WOAH standards also define the regulatory context that may influence case ascertainment and follow-up completeness.

The handling of missing data is a STROBE item that requires explicit attention. Veterinary medical records are incomplete by nature. Owners decline diagnostic tests, animals are lost to follow-up, and necropsy consent is refused. The manuscript must state the amount of missing data for each variable of interest, the reasons for missingness where known, and the analytical approach used. Complete-case analysis is acceptable only when missingness is unrelated to the outcome, an assumption that must be defended instead of asserted.

A Working Reference Table for Guideline Selection

The following table consolidates the primary reporting guidelines applicable to veterinary research, their target designs, and practical compliance considerations. The EQUATOR Network should be consulted for the current versions and any design-specific extensions before submission.

The choice between CONSORT and STROBE is determined by the assignment mechanism, not by the clinical question. A study that randomly allocates animals to treatment groups is a trial and must report against CONSORT. A study that observes outcomes in animals that received different treatments as part of clinical care is observational and must report against STROBE. The distinction holds even when the clinical question is identical.

Practical Implementation in the Manuscript Workflow

Reporting guidelines function best when they shape the study from the design phase instead of serving as a post hoc checklist before submission. The ARRIVE 2.0 explanation and elaboration document explicitly advises that the guidelines support researchers from the start of experimental design through to publication. A protocol that anticipates each checklist item produces a manuscript that reports completely without retrofitting.

The practical sequence is as follows. Select the primary guideline based on the study design before data collection begins. Draft the checklist items that describe methods at the protocol stage. Complete the results items as data are analyzed, not at the writing stage. Use the checklist as a review instrument before submission, verifying that each item has a corresponding statement in the manuscript.

Journal instructions vary in their reporting requirements. Some journals mandate checklist submission with the manuscript, others require a completed checklist as supplementary material, and others reference guidelines in their author instructions without enforcing compliance. The AVMA professional practice resources and the MSD Veterinary Manual both provide guidance on manuscript preparation standards that complement the reporting guidelines. Checking the target journal's specific requirements before writing avoids the need to restructure a manuscript at the submission stage.

The final manuscript should state explicitly which guideline version was followed and where the completed checklist can be found. This statement belongs in the methods section or as a footnote on the title page, depending on journal preference. A manuscript that reports against ARRIVE 2.0 but cites the 2010 version creates confusion about which items were addressed. The original ARRIVE publication and the 2010 companion articles remain historically important, but the ARRIVE 2.0 revision is the current standard for new submissions.

Recognized Complications and Failure Modes

Reporting guidelines fail in predictable ways, and early detection depends on knowing where the failure originates. The most common complication is checklist compliance without substantive adherence. A manuscript may list every ARRIVE item yet omit the actual number of animals allocated to each group, the method of randomisation, or the fate of animals that died before the primary endpoint. The ARRIVE 2.0 guidelines were reorganised specifically to counter this pattern by separating the Essential 10 from the Recommended Set, but the distinction only works if reviewers verify content instead of presence of headings.

A second failure mode is guideline substitution. Authors of observational studies sometimes apply CONSORT because they are familiar with it, producing a flow diagram that implies random allocation where none occurred. The reverse error, applying STROBE to a randomised trial, obscures the allocation concealment and blinding details that readers need. The EQUATOR Network library provides the definitive mapping of study design to guideline, and consulting it before drafting prevents this mismatch.

A third complication is selective application across a multi-part study. A trial may report the primary outcome fully but describe secondary outcomes, subgroup analyzes, or adverse events in abbreviated form. This creates the appearance of completeness while preserving the opportunity for selective reporting. Detection requires comparing the registered protocol or the methods section against every result reported in the results and discussion sections.

Common Errors and Corrective Actions

Less experienced researchers frequently confuse reporting guidelines with study design standards. A guideline cannot rescue a poorly designed experiment, it can only make the design transparent. The explanation and elaboration document for ARRIVE 2.0 addresses this by covering best practice in design and conduct from the start of the experimental process, also at the writing stage. The corrective action is to treat the guideline as a planning tool, consulting it before ethics approval and data collection instead of during manuscript preparation.

A second recurring error is the misuse of the word "random." Authors may describe animals as randomly assigned when assignment was by cage, by order of arrival, or by alternation. These are not equivalent to true randomisation, and the distinction matters for bias assessment. The corrective action is to state the randomisation method explicitly, including the unit of randomisation, and to report whether allocation was concealed from the person enrolling animals.

A third error is omitting the denominator for every percentage. Reporting that 80% of animals survived conveys little without the numerator and the total. This error is detected during peer review but should be caught earlier by a co-author who checks each result against the raw data. The ARRIVE Essential 10 requires this level of specificity as a minimum, and journals that enforce the checklist catch these omissions before publication.

A fourth error involves the description of blinding. Authors may state that the study was blinded without specifying who was blinded, when blinding was broken, or how blinding was verified. The corrective action is to describe blinding separately for animal caregivers, outcome assessors, and data analysts, and to state whether blinding was maintained until the analysis was complete.

Limitations of the Current Evidence

The evidence base for reporting guidelines in veterinary research has genuine gaps. The ARRIVE guidelines were developed primarily with laboratory animal research in mind, and their transfer to clinical veterinary trials, field studies, and production animal research is not always straightforward. The original ARRIVE publication and its 2010 companion papers established the framework for in vivo experiments, but the guidelines do not address species-specific considerations such as client consent, owner-reported outcomes, or the logistics of randomising herds instead of individual animals.

Expert opinion still differs on how strictly guidelines should be enforced. Some journal editors require completion of checklists at submission, while others rely on peer reviewers to assess reporting quality. The British Journal of Pharmacology editorial accompanying the original ARRIVE guidelines signalled early commitment to enforcement, but uptake across veterinary journals has been uneven. The ARRIVE 2.0 revision acknowledges that adherence has been inconsistent despite widespread endorsement, and the stepwise implementation approach was designed to address this reality.

A further limitation is the absence of veterinary-specific extensions for many guidelines. The REFLECT statement for livestock trials exists, but it is not yet as widely adopted as CONSORT or STROBE. Researchers working in production animal medicine may find that neither the human-focused guidelines nor the laboratory-focused ARRIVE items fully capture the cluster-randomised, herd-level designs common in their field.

When to Seek Escalation

Referral to specialist consultation is warranted when the study design does not map cleanly onto an existing guideline. A veterinary epidemiologist or biostatistician should be consulted before data collection if the trial involves cluster randomisation, adaptive designs, or non-inferiority hypotheses, because these designs require reporting elements that the standard checklists address incompletely.

Laboratory involvement becomes necessary when the study involves diagnostic test validation, where reporting standards for sensitivity, specificity, and predictive values differ from those for therapeutic trials. The MSD Veterinary Manual provides species-specific reference material on diagnostic testing that can inform the methods section, but the reporting framework itself should be selected with a clinical pathologist or diagnostic laboratory scientist.

Regulatory reporting obligations arise independently of journal requirements. Studies involving notifiable diseases, adverse drug events, or animals used in regulated research may carry reporting duties to bodies such as the World Organization for Animal Health under its terrestrial animal health standards or national competent authorities. The American Veterinary Medical Association practice resources can help identify relevant obligations in the United States, but researchers must verify the requirements of their own jurisdiction. These obligations exist whether or not the manuscript is published, and they cannot be satisfied by a reporting guideline checklist.

Frequently Asked Questions

How Do I Choose Between ARRIVE and CONSORT When My Study Has Both Randomised and Descriptive Components?

Use the guideline that matches your primary research question. If the main objective is hypothesis testing through random allocation, apply the CONSORT reporting standard for the trial elements and supplement with the ARRIVE Essential 10 for animal-specific details such as housing, welfare, and species characteriztics. If the study is primarily descriptive with a randomised element embedded, reverse the priority. Journals increasingly request both checklists for hybrid designs. Submit both completed checklists with your manuscript and state in the cover letter which items overlap. This prevents reviewer confusion and demonstrates that you have addressed the distinct reporting demands of clinical trial methodology and in vivo experimental transparency.

What Should I Do When My Study Design Does Not Fit Any Published Guideline Cleanly?

Report using the closest matching guideline and declare the mismatch explicitly in the methods section. For example, a field effectiveness study with non-random allocation may align partially with STROBE guidance for observational research and partially with CONSORT. State which items you could not satisfy and why. Describe the deviation in the limitations section with a concrete rationale, such as owner consent constraints or herd-level intervention requirements. Reviewers accept justified deviations when the author has demonstrated awareness. The ARRIVE 2.0 explanation and elaboration document provides worked examples of how to adapt items to specific experimental contexts, which can serve as a model for your own adaptations.

How Much Time and Resource Investment Do These Guidelines Realistically Require?

A completed checklist typically requires two to four hours of focused work for an experienced researcher. The larger investment is prospective: designing the study so that the required information can be recorded. Retrospective completion is slower and often impossible for items such as allocation concealment or blinding procedures that were not documented at the time. Budget for checklist preparation in your grant timeline. The ARRIVE 2.0 revision process used a Delphi exercise to prioritize items, which means the Essential 10 represents the minimum burden for compliance. Start with those ten items for every manuscript, then add the Recommended Set when the study design supports it. Institutional research offices and biostatisticians can review checklists before submission.

How Do Reporting Requirements Differ Between Companion Animals and Production Animals?

The core reporting items are identical, but the contextual detail differs. Production animal studies must report herd-level variables, such as group housing density, feed batch, and production stage, because these affect treatment response and disease transmission. Companion animal studies require more detail on individual variation, including breed predispositions, body condition, and concurrent medications. The WOAH terrestrial animal health standards add reporting expectations for studies involving notifiable diseases or international movement of animals. Welfare reporting also differs: production animal protocols should describe humane endpoints in the context of commercial culling decisions, while companion animal protocols address owner involvement in endpoint decisions. State these contextual factors explicitly instead of assuming the reader will infer them from the species.

What Records Must I Keep During the Study to Satisfy These Guidelines Later?

Maintain a prospective study log that captures each checklist item as the study proceeds. For randomised trials, record the randomisation sequence generation method, allocation concealment mechanism, and implementation steps. For observational studies, document the sampling frame, eligibility criteria, and data collection instruments. Keep versioned copies of all case report forms and any amendments. Record animal-level data on housing, health status, and exclusions with dates. The ARRIVE guidelines for reporting animal research specify that the number of animals in each experimental group at each stage must be traceable, so maintain a flow diagram as you go. Store these records for at least the journal's data retention period, typically five to seven years, and be prepared to share them with editors on request.

How Should I Respond When a Journal Requests a Guideline I Have Not Used Before?

Treat the request as an opportunity to strengthen the manuscript instead of a bureaucratic hurdle. Download the checklist from the EQUATOR Network library and work through each item systematically. If an item does not apply, state that explicitly with a one-sentence justification instead of omitting it silently. For example, a retrospective record review cannot report prospective allocation concealment, so write "not applicable, retrospective design." If the journal requests ARRIVE 2.0, use the Essential 10 as the minimum standard and add Recommended Set items where your data allow. Contact the journal office if the request is ambiguous about which version of a guideline they expect. Most journals accept a completed checklist as a supplementary file, which also signals to reviewers that you have engaged with reporting standards seriously.

Related Clinical & Scientific Guides

• Conducting Systematic Reviews of Veterinary Diagnostic Test Accuracy
• Bias in Veterinary Research: Types, Sources, and Mitigation
• Cluster Randomized Trials in Veterinary Research: Design and Analysis

References and Further Reading

• Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research.. 2010.
• The ARRIVE guidelines 2.0: Updated guidelines for reporting animal research.. 2020.
• Animal research: reporting in vivo experiments: the ARRIVE guidelines.. 2010.
• Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0.. 2020.
• Guidelines for reporting experiments involving animals: the ARRIVE guidelines.. 2010.
• The ARRIVE guidelines 2.0: Updated guidelines for reporting animal research.. 2020.
• ARRIVE Guidelines 2.0 for Reporting Animal Research. PLOS Biology, 2020.
• EQUATOR Network Reporting Guidelines. EQUATOR Network.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.

Related Articles

• The ARRIVE Guidelines for Animal Research: Implementation and Compliance
• Reporting Qualitative Research in Veterinary Science
• Using Mixed Methods in Veterinary Research
• Conducting Qualitative Research in Veterinary Settings
• How to Write a Research Protocol for Veterinary Studies

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.