# Selecting and Implementing Blinding Methods in Veterinary Trials


## Key Takeaways

- Blinding, or masking, is crucial in veterinary trials to prevent conscious or unconscious bias in outcome assessment, animal care, and data collection, thereby safeguarding study validity. Key parties to blind include outcome assessors, animal caregivers, data analysts, and potentially owners or investigators, depending on the trial design and species.
- The feasibility of blinding varies significantly by species and setting; owner blinding is often achievable in companion animals with identical placebo formulations, while caregiver blinding in production animals may be limited by group housing, necessitating pen-level or herd-level allocation.
- Interventions that are inherently visible, such as surgery or radiation therapy, necessitate a "blinded endpoint" design where the intervention itself is open-label, but the outcome assessor remains masked to treatment assignment to mitigate bias.
- Maintaining blinding integrity requires active monitoring throughout the trial, including documenting any accidental disclosures or therapeutic necessity code breaks, and adhering to reporting standards like ARRIVE and CONSORT, which mandate detailed descriptions of blinding methods.
- Formulation matching is paramount for effective blinding, requiring identical appearance, taste, odour, and packaging for active and placebo treatments across all relevant sensory modalities for the target species, with special considerations for injectables, topicals, and oral medications.
- Blinding failures can manifest as differential handling of treatment groups, contamination of control conditions, or observer drift; these are best addressed through structured monitoring, formulation audits, inter-rater reliability checks, and transparent reporting of any compromises.

---

Blinding, also termed masking, is the practice of concealing treatment assignment from individuals involved in a clinical trial. In veterinary research, blinding serves the same purpose as in human medicine: it prevents conscious or unconscious bias from influencing outcome assessment, animal care, and data collection. This article provides practical guidance for veterinary researchers on choosing which parties to blind, implementing blinding procedures across species, and maintaining blinding throughout a study. It addresses the design decisions that precede randomisation and the operational measures that preserve masking until database lock. Statistical analysis of blinded data is outside the scope of this article.

The reader is assumed to be a veterinary researcher designing or reviewing clinical trials, whether in companion animals, livestock, or laboratory species. The content applies across species, with attention to the constraints that each setting imposes. The central question this article answers is: who should be blinded in a veterinary trial, and how can that blinding be achieved and protected in practice?

## At a Glance

| Parameter | Decision or Fact |
|---|---|
| Primary purpose of blinding | Prevent differential treatment or assessment of animals based on knowledge of group assignment |
| Parties typically blinded | Outcome assessors, animal caregivers, data analysts, and in some designs, owners or investigators |
| Parties that cannot be blinded | Personnel delivering inherently visible interventions, such as surgery or radiation |
| Blinding of owners | Feasible in most companion animal trials, requires identical-appearing placebo and active treatments |
| Blinding of animal caregivers | Often feasible in laboratory and production animal settings, may be impossible when treatment requires restraint or injection that differs between groups |
| Blinding of outcome assessors | Usually the most critical and most achievable form of blinding |
| Unblinding triggers | Serious adverse events, suspected unexpected serious adverse reactions, or emergency veterinary care decisions |
| Reporting requirement | Blinding methods must be described in publications per reporting standards such as the ARRIVE guidelines and CONSORT extensions |

## Rationale for Blinding in Veterinary Trials

Bias in animal research is not hypothetical. A systematic survey of published biomedical research using laboratory animals found that 86% of papers did not report blinding, and the authors identified this omission as a threat to the validity of conclusions drawn from those studies. The same survey found that 87% of papers did not report randomisation, indicating that basic safeguards against bias are frequently absent in animal research. These findings, published by Kilkenny and colleagues, underscore that blinding is not an optional refinement but a core component of experimental design.

The mechanisms by which blinding protects validity are several. An investigator who knows a dog is receiving an experimental analgesic may unconsciously score lameness more favourably. A caregiver who knows a calf is in the control group may handle it differently or report adverse events with different thresholds. An owner who knows their cat received a placebo may interpret behavior changes through that expectation. Blinding interrupts these pathways by removing the knowledge that enables them.

Blinding also protects the trial from the effects of differential co-intervention. If caregivers know which animals receive active treatment, they may provide additional supportive care to control animals, consciously or not. This confounds the treatment effect with the effect of differential care. Masking caregivers prevents this form of contamination.

## Who Can Be Blinded

### Outcome Assessors

The outcome assessor is the party whose blinding most directly protects the primary endpoint. In veterinary trials, the assessor may be a clinician scoring lameness, a pathologist grading histologic lesions, a technician measuring joint angles, or an owner completing a quality-of-life questionnaire. Blinding the assessor requires that all animals present identically with respect to treatment. This is straightforward when treatments are identical capsules or injections. It becomes difficult when one treatment produces visible effects, such as sedation, injection-site swelling, or coat staining.

When visible treatment effects cannot be avoided, several strategies exist. The assessor can be separated from the animal's history and treatment record. Video recordings can be scored by an assessor who never sees the animal in person. Objective instruments, such as force plates or activity monitors, reduce but do not eliminate the need for assessor blinding, because the person operating the instrument may still influence the measurement.

### Animal Caregivers

Blinding caregivers is feasible when treatments are administered identically. In laboratory settings, this often requires that a person not involved in animal care prepares and administers the treatment. In production animal trials, group housing may make individual blinding impractical, and pen-level allocation with a single treatment per pen may be the only workable design. In such cases, the caregiver knows the treatment for the pen, but the outcome assessor can remain blinded if assessment is performed by a separate individual.

### Owners and Clients

Owner blinding is a distinctive feature of companion animal trials. Owners are often the primary outcome assessors for subjective endpoints such as pruritus, activity level, or appetite. They are also the individuals who decide whether to seek veterinary care during the trial, which can influence adverse event reporting. Blinding owners requires that active and placebo treatments be indistinguishable in appearance, taste, and packaging. This is achievable for oral medications and many topical products but may be impossible for injectable treatments with different viscosities or for devices.

### Investigators and Study Personnel

The principal investigator and study coordinators may be blinded or unblinded depending on the trial structure. In single-center trials, the investigator often serves as both clinician and assessor, making investigator blinding essential. In multicentre trials, a local unblinded pharmacist or coordinator can manage treatment allocation while the investigator remains masked. The decision to blind the investigator must be made before randomisation and documented in the protocol.

## Constraints Imposed by the Intervention

Some interventions cannot be delivered in a blinded fashion. Surgical procedures, radiation therapy, and certain behavioral interventions are inherently visible. In these cases, the trial may use a blinded endpoint design, in which the intervention is open-label but the outcome assessor is masked to treatment assignment. This design was used in the MISTIE III trial of minimally invasive surgery for intracerebral hemorrhage, which was described as an open-label, blinded endpoint phase 3 trial. The same logic applies to veterinary surgical trials: the surgeon cannot be blinded, but the assessor of postoperative outcomes can be.

Blinded endpoint designs require careful separation of the clinical team from the assessment team. The surgeon must not be present during outcome assessments. The assessment protocol must specify which measurements are taken, by whom, and under what conditions. The protocol must also specify how the assessor will be prevented from learning the treatment assignment through conversation with the owner or review of the medical record.

## Species-Specific Considerations

### Companion Animals

Dogs and cats present the full range of blinding opportunities and challenges. Owner blinding is usually feasible and desirable. The owner administers most treatments at home, so the owner must be blinded to ensure consistent administration and unbiased reporting. The veterinarian assessing outcomes must also be blinded, which requires that the clinic staff manage treatment allocation separately from clinical care.

One challenge in companion animal trials is the detection of treatment effects by owners. If an owner perceives improvement, they may be tempted to guess the treatment assignment. This is not a failure of blinding but a natural consequence of an effective treatment. The protocol should instruct owners not to discuss their guesses with clinic staff.

### Production Animals

In cattle, pigs, sheep, and poultry, blinding is complicated by group housing and the scale of operations. Individual animal blinding is possible when treatments are administered at handling, but group-level treatments such as feed additives or water medications cannot be blinded at the individual level. Pen-level or herd-level allocation with blinded outcome assessment is often the most practical design.

Production animal trials also involve multiple stakeholders with different interests. The producer may know the treatment assignment because they administer it. The veterinarian assessing outcomes may be the herd veterinarian who also prescribed the treatment. In these settings, the trial must identify an independent assessor who has no role in treatment administration.

### Laboratory Animals

Laboratory animal trials offer the greatest control over blinding. The research team can be divided into those who handle treatment preparation and those who handle animals. Cage labels can be coded so that caregivers do not know which animals receive which treatment. The main challenge is the small size of many laboratory animals, which makes blood sampling and other procedures difficult to standardize across groups.

## Maintaining Blinding During the Trial

Blinding is not established at randomisation and then forgotten. It must be actively maintained through monitoring and documentation. The protocol should specify how treatment codes are stored, who has access to them, and under what circumstances they may be broken. Emergency unblinding procedures must be defined for cases in which an animal's clinical condition requires knowledge of treatment for appropriate care.

The protocol should also specify how the success of blinding will be assessed. Some trials ask assessors to guess treatment assignments at the end of the study and compare the proportion of correct guesses to chance. This practice is controversial, because a correct guess may reflect genuine treatment effects instead of a breach of blinding. A more useful approach is to document any instances in which blinding was compromised and to report these in the publication.

Reporting standards require that blinding methods be described in sufficient detail for readers to judge their adequacy. The ARRIVE guidelines 2.0 specify that publications should state which parties were blinded and how blinding was achieved. The CONSORT statement, originally developed for human trials, includes items on blinding that have been shown to improve reporting completeness in journals that endorse them. Veterinary researchers should consult these standards when designing and reporting trials.

## Selecting the Blinding Level: A Decision Framework

The extent of blinding should be determined by the risk of bias introduced by each role, the feasibility of masking that role given the intervention, and the resources available. A tiered approach allows investigators to allocate effort where it matters most.

| Blinding Level | Roles Masked | Appropriate When | Typical Challenges |
|---|---|---|---|
| Full double-blind | Owners, caregivers, outcome assessors, investigators | Identical placebo available, intervention has no perceptible effects | Formulation matching, accidental code breaks |
| Assessor-blind only | Outcome assessors | Intervention cannot be masked from owners or caregivers (e.g., surgery, radiation) | Assessor contamination from owners, subjective outcomes |
| Open-label with blinded endpoint adjudication | Endpoint committee | Objective or centrally adjudicated outcomes, surgical or device trials | Defining endpoints before unblinding, adjudicator training |
| Unblinded | All roles | Feasibility pilot studies, certain production animal trials with objective production outcomes | Justification required in protocol, higher risk of bias |

Select the minimum blinding level that adequately protects the primary outcome from bias. If the primary outcome is subjective, such as pain scoring or owner-reported quality of life, assessor and owner blinding carry the greatest weight. If the primary outcome is objective, such as survival time or a laboratory measurement, assessor blinding may suffice, though owner blinding still prevents differential reporting of adverse events.

The [CONSORT statement reporting standards](https://pubmed.ncbi.nlm.nih.gov/23194585/) require explicit description of who was blinded and how, which serves as a useful checklist when designing the blinding protocol. Trials that omit this information cannot be properly evaluated for bias.

## Blinding Materials and Formulation Matching

The placebo must be indistinguishable from the active treatment across all sensory modalities relevant to the species. For oral formulations, match appearance, size, shape, odour, and taste. Flavour masking agents may be needed for compounded preparations. For injectables, match volume, viscosity, color, and the appearance of the vial or syringe. For topical preparations, match texture, fragrance, and application vehicle.

When the active drug has a distinctive odour or taste that cannot be replicated, consider using a strong masking flavour in both preparations. When the active drug causes local tissue irritation that the placebo does not, the difference may be detectable by the animal even if not by the owner. This scenario requires careful pilot testing and documentation.

For food-producing animals, the placebo must be approved for use in the target species and must not carry a withdrawal period that differs from the active product. This can complicate blinding when the active product has a labelled withdrawal period and the placebo does not. Consult current [MSD Veterinary Manual pharmacology references](https://www.msdvetmanual.com/) and the relevant regulatory authority for guidance on extemporaneous placebo preparation.

## Packaging, Allocation Concealment, and Code Management

Allocation concealment precedes blinding and must be maintained until the moment of assignment. Use sequentially numbered, opaque, sealed containers prepared by a party not involved in outcome assessment. The randomisation list should be held by an independent statistician or pharmacy, with access restricted to those who must prepare or administer the intervention.

The blinding code should be broken only under predefined circumstances. These include suspected serious adverse reactions where knowing the treatment is necessary for clinical management, and regulatory requirements for expedited reporting. The protocol must specify who may request a code break, how the request is evaluated, and who authorises it.

A common failure mode is the unblinded administrator who inadvertently reveals group assignment through behavior or comments. When the intervention must be prepared by an unblinded person, such as a compounding pharmacist or a technician who draws up the dose, ensure that person has no contact with owners, assessors, or data entry personnel. Use opaque syringes, covered infusion lines, or a two-person system where one prepares and the other administers without seeing the preparation.

## Monitoring and Documentation of Blinding Integrity

Blinding integrity should be monitored throughout the trial, not assumed at the end. At each follow-up visit, ask owners and assessors to guess the treatment assignment and record the reason for the guess. Compare the proportion of correct guesses to chance using a binomial test. A correct guess rate significantly above 50% indicates that blinding has failed, even if the code was never formally broken.

Document all code breaks, including the date, the individual requesting the break, the reason, and the outcome. The [ARRIVE guidelines for reporting animal research](https://arriveguidelines.org/) require reporting of blinding procedures and any deviations, which supports transparent evaluation of trial quality.

## Troubleshooting Unblinding Events

Unblinding events fall into several categories, each requiring a distinct response.

| Event Type | Example | Immediate Response | Protocol Action |
|---|---|---|---|
| Accidental disclosure | Owner sees the drug label | Record the event, do not reveal assignment to assessor | Analyze as per-protocol or sensitivity analysis, report in publications |
| Therapeutic necessity | Suspected adverse reaction requiring knowledge of treatment | Authorised code break, document reason | Treat as unblinded for that participant, consider replacement if protocol specifies |
| Formulation failure | Placebo visibly different after storage | Halt enrollment, investigate stability and appearance | Revise formulation, consider whether enrolled participants remain valid |
| Assessor contamination | Assessor overhears owner comment about response | Reassign assessor, record the event | Include in sensitivity analysis, report as a limitation |

When a code break occurs, the participant may continue in the trial, but the outcome assessment for that participant should be handled with awareness of the unblinding. If the primary outcome is subjective and the assessor becomes unblinded, consider using a blinded endpoint committee to re-evaluate the outcome from recorded data where possible.

## Species-Specific Implementation Notes

In companion animal trials, owners are often present during assessments and may inadvertently reveal information about the animal's response. Instruct owners before each visit not to discuss their observations with the assessor, and structure the visit so that the assessor examines the animal before interviewing the owner.

In production animal trials, group housing creates unique challenges. Animals in the same pen or flock may receive different treatments, and caregivers may observe which animal receives which preparation. Consider pen-level or flock-level randomisation when the intervention cannot be masked at the individual level. This changes the unit of analysis and must be reflected in the statistical plan.

In laboratory animal trials, the [systematic survey of animal research reporting](https://pubmed.ncbi.nlm.nih.gov/19956596/) found that blinding was absent in 86% of published studies, indicating substantial room for improvement. Cage-level blinding is often feasible and should be standard practice. When the intervention causes visible effects, such as changes in coat condition or behavior, consider using a separate cohort for welfare assessments that is not involved in data collection.

Equipment availability changes the correct choice. Video recording of assessments allows blinded central review when on-site assessors cannot be masked. Electronic case report forms can enforce blinding by hiding treatment assignment fields from assessor interfaces. When such infrastructure is unavailable, simpler measures, such as using coded labels and separate data collection forms, remain effective.

## Recognized Complications and Failure Modes

Blinding failures in veterinary trials typically follow recognizable patterns. The most consequential is inadvertent unblinding through differential handling of treatment groups. An animal that receives an injection with different viscosity, volume, or color cues can prompt caregiver speculation that propagates through the ward. Detection depends on structured monitoring: scheduled checks of allocation code integrity, incident logs for suspected breaches, and exit interviews with study personnel. A breach is confirmed when a caregiver can state the assigned treatment with confidence exceeding chance and can articulate the basis for that belief.

A second failure mode is contamination of the control condition. Placebo formulations that are not truly inert, or that differ in palatability, can introduce systematic differences unrelated to the active intervention. The discriminating check is a formulation audit comparing appearance, texture, odour, and administration characteriztics between active and control products before trial commencement.

Third, observer drift degrades blinding over time. Assessors who become familiar with treatment-associated clinical signs may unconsciously adjust their scoring. This is detected by periodic inter-rater reliability checks and by reviewing whether outcome distributions shift across the trial duration. Blinded endpoint adjudication, as used in the MISTIE III trial where an open-label intervention was paired with blinded outcome assessment, offers a model for protecting outcome validity when full blinding is impossible ([blinded endpoint assessment in the MISTIE III trial](https://pubmed.ncbi.nlm.nih.gov/30739747/)).

## Common Errors and Corrective Actions

Less experienced trialists often blind outcome assessors while leaving allocation visible to caregivers who interact with those assessors. The corrective action is to map all points of contact between personnel and animals, then assign blinding status at each node. A related error is failing to test whether blinding was successful. Asking caregivers and assessors to guess treatment allocation at trial close provides empirical evidence of blinding integrity, though the interpretation of such tests requires care because correct guesses can reflect genuine treatment effects instead of protocol breaches.

Another frequent mistake is treating blinding as a single binary decision instead of a graduated design. Trials can blind allocation, administration, assessment, or analysis independently. Selecting the level of blinding should follow from the intervention's constraints, not from convention. When complete blinding is impossible, partial blinding with explicit documentation of what remained open is preferable to abandoning blinding altogether.

A third error involves code management. Investigators sometimes store allocation codes with the case report forms, allowing accidental exposure during data entry. Codes should reside with an independent statistician or pharmacy service, with access logged and restricted. The [CONSORT statement's effect on reporting completeness](https://pubmed.ncbi.nlm.nih.gov/23194585/) includes blinding-related items, and journals increasingly expect explicit description of code management procedures.

## Limitations of Current Evidence

The veterinary blinding literature remains sparse. Systematic survey data from laboratory animal research show that blinding was absent in 86% of published studies, and randomisation was absent in 87%, indicating that basic bias-reduction practices are not yet normative ([survey of experimental design quality in animal research](https://pubmed.ncbi.nlm.nih.gov/19956596/)). Extrapolating from human trial evidence is reasonable for general principles but unreliable for species-specific challenges such as masking drug taste in cats or managing group housing in pigs.

Expert opinion diverges on several points. Whether owner blinding is essential when outcome measures are objective, such as survival time or laboratory values, remains contested. Some argue that owner blinding primarily protects against differential reporting of subjective signs and is unnecessary for hard endpoints. Others maintain that unblinded owners influence animal behavior and caregiver expectations in ways that contaminate even objective measures. A second area of disagreement concerns the acceptability of unblinded administration with blinded assessment. This design is increasingly accepted for surgical and device trials, following the model of the MISTIE III trial, but some reviewers still regard it as inferior to full blinding regardless of the intervention ([blinded endpoint methodology in surgical trials](https://pubmed.ncbi.nlm.nih.gov/30739747/)).

Reporting standards continue to evolve. The ARRIVE guidelines specify blinding as an essential item for transparent animal research reporting, and the EQUATOR Network maintains a library of reporting checklists that includes veterinary-specific extensions ([ARRIVE guidelines for animal research reporting](https://arriveguidelines.org/), [EQUATOR Network reporting guideline library](https://www.equator-network.org/)). Adherence to these standards remains variable, and reviewers should treat their absence as a risk-of-bias signal instead of a stylistic omission.

## Escalation and Referral

Most blinding problems can be resolved within the trial team. Referral to a biostatistician or methodologist is warranted when a suspected breach affects a substantial proportion of enrollments, when the primary outcome is subjective and assessor blinding has failed, or when the decision to continue or stop the trial hinges on the severity of unblinding. Laboratory involvement is indicated when blinding failures arise from formulation issues, such as instability of the active product or unexpected interactions with the placebo matrix, where pharmaceutical analysis is required to determine whether the products remain indistinguishable.

Regulatory reporting obligations vary by jurisdiction and species. For studies conducted under veterinary authorisation frameworks, unblinding that affects animal welfare decisions, such as rescue medication or early withdrawal, may require notification to the relevant authority. Production animal trials may trigger reporting under animal health or food safety regulations if blinding failures coincide with adverse events. The [WOAH terrestrial animal health standards](https://www.woah.org/en/what-we-do/standards/codes-and-manuals/terrestrial-code-online-access/) provide a reference point for international expectations regarding trial conduct in animals that may enter trade. When in doubt, consult the institutional animal ethics committee and the relevant national authority before proceeding.

| Observation | Likely cause | Discriminating check |
|---|---|---|
| Caregivers guess allocation above chance | Formulation differences or procedural cues | Formulation audit, exit interview |
| Outcome scores shift mid-trial | Observer drift or unblinding event | Inter-rater reliability check, timeline review |
| Placebo group shows unexpected effects | Contaminated control product | Pharmaceutical analysis of placebo matrix |
| Code access logged without authorisation | Inadequate code security | Audit access logs, review storage procedures |
| Blinding status undocumented in publications | Reporting omission | Compare manuscript against ARRIVE checklist |

## Frequently Asked Questions

**What is the minimum blinding level acceptable when resources are limited?**

When full blinding is impossible, protect the outcome assessment first. Blinded outcome assessors using prespecified, objective criteria reduce bias more than blinded caregivers when endpoints are subjective. For objective outcomes such as survival or laboratory values, assessor blinding may matter less. If you cannot blind assessors, consider central review of recorded outcomes, as used in the MISTIE III trial where an open-label design still achieved blinded endpoint evaluation. Document any deviation from the ideal blinding scheme in the protocol and report it transparently. Reporting guidelines such as the [CONSORT statement](https://pubmed.ncbi.nlm.nih.gov/23194585/) require explicit description of blinding methods, and incomplete reporting of blinding was common in animal research surveyed by [Kilkenny and colleagues](https://pubmed.ncbi.nlm.nih.gov/19956596/).

**How do I blind when the intervention has a distinctive odour or appearance?**

Formulation matching is the standard solution. Use identical vehicles, flavourings, and delivery devices for active and control preparations. For oral medications with strong taste, use a masking flavour in both arms or administer via a route that bypasses sensory detection, such as a feeding tube. For topical or injectable products, match color, viscosity, and container type. When perfect matching is impossible, separate the personnel who prepare and administer the intervention from those who assess outcomes. This keeps the assessor blinded even if the animal or caregiver cannot be. Validate the matching in a small pilot group before trial initiation, asking naive staff to identify which preparation is active.

**Can I blind a field trial in production animals when treatments are given by farm staff?**

Yes, with careful protocol design. Farm staff can remain blinded if the treatment is dispensed in identical, pre-labelled containers by an off-site pharmacist or investigator. For group-level treatments such as feed additives, use coded batches and have a third party perform allocation. Water-soluble treatments require attention to mixing equipment and residual contamination between pens. Blinding is easier when the control is a placebo instead of a different active treatment. If farm staff must know the treatment for safety reasons, restrict their role to administration and use independent assessors for all outcome measurements. The [ARRIVE guidelines](https://arriveguidelines.org/) specify that blinding procedures should be reported in enough detail to allow replication.

**What records must I keep to demonstrate blinding was maintained?**

Maintain three separate documents: the allocation list held by an independent custodian, the blinding integrity log recording any suspected or confirmed unblinding events with dates and circumstances, and the final blinding assessment completed before data analysis. The blinding integrity log should include who became unblinded, how it occurred, and whether outcome assessments were affected. At trial closeout, ask each blinded individual to record whether they remained blinded and, if not, when they became aware of allocation. This information belongs in the trial report. The [EQUATOR Network](https://www.equator-network.org/) library provides reporting checklists that specify which blinding items must be documented, and incomplete reporting of blinding was identified as a common deficiency in animal research publications.

**How do I explain the need for blinding to an owner who wants to know which treatment their pet received?**

Frame blinding as a scientific safeguard that protects the validity of the study and, by extension, the value of the information gained. Explain that neither you nor they can reliably judge improvement without bias, and that knowing the treatment could unconsciously influence observations. Reassure them that the allocation code can be broken immediately if their animal's health requires it, and that emergency unblinding procedures are in place. Emphasize that the same uncertainty applies to the veterinarian, which is a feature, not a flaw. Owners are generally receptive when the explanation focuses on fairness to their animal and the integrity of the research question.

**Does blinding work differently in wildlife or exotic species compared with domestic animals?**

The principles are identical, but practical constraints differ. Wildlife studies often rely on remote monitoring, telemetry, or capture-based sampling, which limits who can be blinded and when. Anesthetised animals may be assessed by personnel who were not involved in drug administration, preserving assessor blinding. For exotic species in zoological settings, keeper staff who know individual animals may be necessary for behavioral assessments, making caregiver blinding impractical. In these cases, use video recordings scored by blinded observers or automated sensors for outcome measurement. The [WOAH terrestrial animal health standards](https://www.woah.org/en/what-we-do/standards/codes-and-manuals/terrestrial-code-online-access/) emphasize transparent reporting of methods, which applies equally to wildlife research.

## Related Clinical & Scientific Guides

* [Conducting Systematic Reviews of Veterinary Diagnostic Test Accuracy](/knowledge/veterinary-medicine/veterinary-research-methods/conducting-systematic-reviews-veterinary-diagnostic-test-accuracy)
* [Bias in Veterinary Research: Types, Sources, and Mitigation](/knowledge/veterinary-medicine/veterinary-research-methods/bias-veterinary-research-types-sources-mitigation)
* [Cluster Randomized Trials in Veterinary Research: Design and Analysis](/knowledge/veterinary-medicine/veterinary-research-methods/cluster-randomized-trials-veterinary-research-design-analysis)


## References and Further Reading

- [Does use of the CONSORT Statement impact the completeness of reporting of randomised controlled trials published in medical journals? A Cochrane review.](https://pubmed.ncbi.nlm.nih.gov/23194585/). 2012.
- [Survey of the quality of experimental design, statistical analysis and reporting of research using animals.](https://pubmed.ncbi.nlm.nih.gov/19956596/). 2009.
- [Sugar consumption, metabolic disease and obesity: The state of the controversy.](https://pubmed.ncbi.nlm.nih.gov/26376619/). 2016.
- [Human rhinoviruses.](https://pubmed.ncbi.nlm.nih.gov/23297263/). 2013.
- [Trends in prevalence of blindness and distance and near vision impairment over 30 years: an analysis for the Global Burden of Disease Study.](https://pubmed.ncbi.nlm.nih.gov/33275950/). 2021.
- [Efficacy and safety of minimally invasive surgery with thrombolysis in intracerebral hemorrhage evacuation (MISTIE III): a randomised, controlled, open-label, blinded endpoint phase 3 trial.](https://pubmed.ncbi.nlm.nih.gov/30739747/). 2019.
- [ARRIVE Guidelines 2.0 for Reporting Animal Research](https://arriveguidelines.org/). PLOS Biology, 2020.
- [EQUATOR Network Reporting Guidelines](https://www.equator-network.org/). EQUATOR Network.
- [MSD Veterinary Manual, Professional Edition](https://www.msdvetmanual.com/). MSD Veterinary Manual.

## Related Articles

- [Implementing Bayesian Methods in Veterinary Clinical Trials](/knowledge/veterinary-medicine/veterinary-research-methods/implementing-bayesian-methods-veterinary-clinical-trials)
- [Blinding in Veterinary Clinical Research: Methods and Challenges](/knowledge/veterinary-medicine/veterinary-research-methods/blinding-veterinary-clinical-research-methods-challenges)
- [Statistical Analysis of Veterinary Clinical Trials: Common Methods and Misconceptions](/knowledge/veterinary-medicine/veterinary-research-methods/statistical-analysis-veterinary-clinical-trials-common-methods-misconceptions)
- [Using Mixed Methods in Veterinary Research](/knowledge/veterinary-medicine/veterinary-research-methods/using-mixed-methods-veterinary-research)
- [Assessing Risk of Bias in Veterinary Randomized Trials](/knowledge/veterinary-medicine/veterinary-research-methods/assessing-risk-bias-veterinary-randomized-trials)

> 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.