Pragmatism, Proportionality and Progress: What Veterinary Clinical Research Can Learn from ICH E6(R3) and E8(R1)

Veterinary clinical research continues to operate under VICH GCP GL9 1 , a guideline that has remained unchanged since 2000. During the same period, human clinical research has undergone major modernisation, culminating in ICH E8(R1) 2 (2021) and ICH E6(R3) 3 (2025). These revisions reflect a transformed clinical study environment with digital technologies, decentralised models, risk-based approaches, and a renewed emphasis on Quality by Design (QbD).

This article explores how the principles embedded in E6(R3) and E8(R1) can be applied meaningfully to veterinary clinical studies. It examines opportunities for simplification, risk proportionate processes, improved sponsor oversight, and modern data governance. It argues that pragmatism is not only relevant to veterinary research, it is essential for ensuring scientific robustness, operational feasibility, and
the credibility of study outcomes.

Introduction: Two Decades of Divergence
When VICH GCP GL9 was finalised in 2000, it represented a major step forward for veterinary clinical research. Yet the world around it has changed dramatically. Digital tools, electronic data capture, remote monitoring, and increasingly complex study designs have become routine.

Meanwhile, human clinical research has evolved through multiple revisions of ICH GCP, culminating in two landmark updates:

ICH E8(R1), 2021 introduced foundational concepts such as stakeholder engagement, operational feasibility, and the prioritisation of activities essential to study quality.

ICH E6(R3), 2025 represents a complete overhaul, emphasising principles-based flexibility, proportionality, risk-based quality management, and
modern expectations for data governance and technology.

The contrast is stark. Human GCP has modernised to reflect the realities of contemporary research. Veterinary GCP has not.

Yet nothing prevents the veterinary sector from learning from the principles of E6(R3) and E8(R1) and applying them thoughtfully within existing frameworks. These concepts are not regulatory burdens, they are practical ways of improving study quality, reducing unnecessary complexity, and strengthening reliable decision making. The central question is simple: Aren’t pragmatic and risk proportionate approaches equally relevant to veterinary clinical studies? This article argues that they are not only relevant: they are overdue.

Quality by Design: A Mindset Veterinary Studies Need
ICH E6(R3) places Quality by Design (QbD) at the centre of modern clinical research. Although not a new concept, its prominence in R3 marks a shift away from compliance driven thinking toward purposeful, fit-for-purpose study design. At its core, Quality by Design requires that protocols be built around clear, decision-driven objectives and a deliberate focus on the factors that are truly critical to quality. It encourages early identification of risks, multidisciplinary input during protocol development, and a willingness to simplify procedures and data collection wherever simplification does not compromise scientific integrity. When applied well, Quality by Design supports reliable decision making while ensuring that study procedures remain operationally feasible for investigators, site staff, and service providers.

This mindset is particularly relevant in veterinary research, where protocols can easily become overengineered. It is not uncommon to see studies with numerous secondary endpoints, complex inclusion and exclusion criteria, or burdensome assessments collected “just in case”. Each additional requirement increases workload for veterinarians, technicians, and owners, and with that comes a greater risk of errors, inconsistencies, protocol deviations, missing data, and noncompliance. Quality by Design challenges this tendency by asking whether each element of the protocol genuinely contributes to the study’s objectives. A companion animal dermatology study illustrates the point. The original protocol included eight secondary endpoints, weekly photographic assessments, owner diaries, and two laboratory parameters that were ultimately never used in the analysis. During monitoring, sites struggled with the sheer volume of assessments, and approximately 40% of photographs were unusable. A Quality by Design review, grounded in scientific rationale rather than habit, would likely have identified early
that only a small number of endpoints were critical to decision making, that photographic assessments added little value, and that owner diaries increased burden without improving data quality.

Similar challenges arise in livestock settings. In a dairy mastitis study, for example, the protocol initially required daily quarter level milk sampling for both primary and exploratory endpoints. Farm staff quickly reported that this level of sampling was unworkable during peak milking hours. A Quality by Design informed review showed that composite samples were sufficient for the primary endpoint, whilst quarter level sampling added burden without improving interpretability. Simplifying the sampling schedule improved compliance, reduced animal handling stress, and strengthened the consistency of the dataset. Quality by Design also applies in preclinical laboratory environments. In a tolerance
study conducted at a contract laboratory, early discussions revealed that bodyweight measurements were critical to dose accuracy. However, two different scales were being used interchangeably across rooms, introducing avoidable variability.

Recognising this as a critical to quality factor allowed the team to assign a single calibrated scale preventing dose miscalculations and improving data integrity. Importantly, Quality by Design is not about ignoring scientific advice or lowering standards. Rather, it is about stepping away from “we have always done it this way” thinking and ensuring that scientific decisions are purposeful, proportionate, and aligned with study objectives. Identifying critical to quality factors early is especially important in veterinary studies, where site resources, staffing, and owner
involvement vary widely and operational feasibility can make or break a study. A simplified, risk informed protocol not only reduces workload and improves compliance but also strengthens the resulting dataset, demonstrating how thoughtful design upfront can prevent avoidable issues later.

Engaging Stakeholders and Ensuring Operational Feasibility
Stakeholder engagement and operational feasibility are central themes in ICH E8(R1), and they are equally relevant, arguably essential, in veterinary clinical research. Engaging stakeholders early in study design provides insight into the practical realities of conducting studies across diverse veterinary environments. In veterinary studies, stakeholders extend far beyond the investigator alone; they include technicians, dispensers, laboratory staff, monitors, data managers, CROs, service providers, and, in some cases, animal owners. Each group brings a different perspective on feasibility, workload, and potential sources of error, and their input can reveal risks or inefficiencies that may not be visible from the sponsor’s vantage point.

This early engagement is particularly important given the variability inherent in veterinary practice. Regional differences in clinical workflows, diagnostic capabilities, staffing levels, and client expectations can significantly influence whether a protocol is workable. Stakeholders are often the first to identify procedures that are overly burdensome, require equipment not routinely available, or introduce unnecessary complexity. For example, during the development of a multi country field study, technicians highlighted that a proposed assessment required specialised equipment rarely found in general practice. Addressing this issue early, by modifying the assessment or providing alternative methods, prevented delays, reduced the risk of inconsistent data collection, and ensured that the study remained feasible across all
participating sites.

Stakeholder engagement also supports recruitment and retention, two areas where veterinary studies can face unique challenges. Technicians and dispensers, who often have the closest contact with owners, can provide valuable insight into which procedures may deter participation or lead to dropouts. Laboratory staff may flag logistical issues such as sample transport times or storage requirements that could compromise sample integrity if not addressed proactively. By incorporating these perspectives into the protocol, sponsors can design studies that are not only scientifically robust but also operationally realistic.

Farm-based studies illustrate the importance of this approach. In a lameness study in dairy cattle, herd managers highlighted that locomotion scoring during morning milking was impractical due to cow flow and staff workload. Adjusting the scoring window to afternoon turnout improved feasibility and data consistency, an example of how operational insight from those closest to the animals can materially strengthen study execution. Companion animal studies face their own challenges. In a dermatology study, veterinary nurses noted that the proposed skin scoring system required around 20 minutes per animal, an unrealistic expectation during busy consulting hours. Simplifying the scoring tool improved adherence, reduced missing data, and made the protocol workable for first opinion practices.

The advantages of this collaborative approach are clear: it builds trust, improves adherence, and enhances the likelihood of study success. However, it also requires time, openness to feedback, and a willingness to adjust longstanding practices. Sponsors must balance scientific objectives with operational realities, ensuring that stakeholder input informs decision making without compromising the integrity of the study. When done well, stakeholder engagement becomes a powerful tool for identifying risks early, simplifying study design, and ensuring that veterinary clinical studies are both feasible and fit for purpose.

Risk Proportionality: Moving Beyond One Size Fits All
One of the most significant shifts in ICH E6(R3) is the expectation that study processes should be proportionate to the risks involved. MHRA inspectors have emphasised in recent GCP forums that proportionality is no longer simply encouraged but is now an expected principle underpinning modern study design and oversight. This reflects the broader move within E6(R3) toward flexibility, critical thinking, and a principles based approach to quality.

In veterinary research, VICH GCP GL9 makes no reference to risk proportionality, and this absence has shaped longstanding practice. Many studies still apply identical scrutiny to primary and secondary data, rely on uniform monitoring plans, or limit risk-based thinking to narrow activities such as 10% data checks or audit site selection. This uniformity is not proportionate; it is simply habitual, and it often results in unnecessary burden without improving scientific robustness.

The value of proportionality becomes clear when considering the realities of veterinary practice. In a respiratory study conducted in first opinion clinics, for example, radiographs were initially scheduled at fixed timepoints. Practices highlighted that weekend diagnostic capacity was limited, leading to missed windows and avoidable deviations. A proportional approach allowed a ±24hour window for noncritical imaging, maintaining scientific integrity while accommodating real world workflow constraints. Similarly, in a parasitology study in beef cattle, faecal egg
counts formed the primary endpoint, yet environmental conditions such as mud, weather, and pasture rotation introduced variability in sample quality. Applying proportionality meant focusing oversight on sample collection and storage, processes that directly influenced the primary endpoint, while reviewing secondary behavioural observations only periodically.

These examples illustrate the broader principle: a proportionate approach directs attention and resources to the aspects of a study that genuinely influence scientific validity, animal welfare, and regulatory decision making. In a multisite field study where the primary endpoint depends on a laboratory assay, for instance, processes such as sample collection, labelling, storage, and analysis warrant intensive oversight, whereas owner diaries or noncritical secondary data may require only light touch review or sampling. The advantages are clear: reduced burden, improved operational feasibility, and more meaningful oversight.

However, proportionality also requires careful judgement, strong scientific rationale, and a willingness to move away from the comfort of “treat everything the same”. If applied superficially, there is a risk of under monitoring or overlooking emerging issues. When applied thoughtfully, proportionality aligns veterinary studies with modern expectations for efficiency and scientific robustness, ensuring that oversight
is both meaningful and manageable.

Risk-Based Quality Management: Preventing Problems Before They Occur
Risk-based quality management is one of the most substantive shifts introduced by ICH E6(R3), embedding a structured, proactive approach to quality throughout the entire study lifecycle. Rather than treating quality as something verified retrospectively through monitoring or audits, risk-based quality management encourages teams to identify potential risks early, evaluate their likely impact, implement proportionate controls, and continuously review whether those controls remain effective. Quality tolerance limits add an additional layer of protection by detecting emerging issues before they compromise study integrity. This represents a fundamental shift from reactive problem solving to anticipatory risk prevention. The value of risk-based quality management becomes particularly clear in veterinary studies, where variability is inherent. Veterinary studies often involve diverse site types, variable equipment, differing levels of research experience, and a high degree of owner involvement. These factors introduce operational and scientific variability that cannot be managed effectively through uniform oversight. For example, in a
pharmacokinetic study where sample timing is critical, a systemic issue with centrifuge calibration at one site could invalidate the entire dataset. A risk-based quality management approach would require clear instructions for sample handling, verification of equipment settings, monitoring visits timed to observe sample processing, and predefined quality tolerance limits for deviations in sample integrity. Early escalation mechanisms ensure that emerging problems are addressed before they affect the study’s primary endpoint.

Proactive risk prevention can take many forms. In a multisite field study evaluating a new injectable product, early risk assessment revealed that improper needle gauge selection could affect dose accuracy in smaller breeds. Addressing this risk through targeted training, simplified instructions, and a brief competency check prevented inconsistent dosing and reduced protocol deviations. Similarly, in a study relying on
owner reported outcomes, early identification of the risk of incomplete or inconsistent diary entries led to the introduction of a simplified diary format, automated reminders, and a short onboarding call to reinforce expectations, small interventions that significantly improved data completeness and reliability.

Livestock studies present their own distinct challenges. In a sheep vaccine study, early risk assessment identified that inconsistent restraint techniques during blood sampling were contributing to haemolysis and sample rejection. A simple mitigation, brief refresher training for stockpersons and a standardised restraint method, substantially reduced haemolysis rates and improved the reliability of laboratory
results. This example illustrates how risk based quality management can strengthen data quality by addressing practical, operational risks that might otherwise go unnoticed.

The advantages of risk-based quality management are clear: it directs attention to what truly matters, reduces unnecessary burden, and strengthens the scientific robustness of veterinary studies. However, it also requires disciplined judgement, a willingness to challenge established habits, and a cultural shift away from treating all data and processes as equally important. If applied superficially, risk-based quality
management risks becoming a box ticking exercise or, worse, an excuse for under monitoring. When applied thoughtfully, it provides a structured, defensible, and efficient approach to ensuring that veterinary clinical studies generate reliable, meaningful results while remaining operationally feasible for investigators and site staff.

Sponsor Oversight: Fit for Purpose, Not Formulaic
Sponsor oversight is another area where ICH E6(R3) introduces a meaningful shift, reframing oversight as a flexible, risk tailored responsibility rather than a formulaic set of activities. The sponsor remains accountable for ensuring reliable results, protecting study animals, and supporting sound decision making, but the way this responsibility is exercised should be proportionate to the complexity and risks of the study. This represents a departure from traditional oversight models that rely heavily on routine site monitoring and standardised documentation checks, regardless of the study’s actual risk profile.

In veterinary studies, this more nuanced approach has clear implications. Oversight must extend beyond the animal sites themselves to include laboratories, data management functions, technology vendors, and any subcontracted service providers involved in study-related activities. Written agreements that clearly define roles, responsibilities, and delegated tasks are essential, particularly in studies involving multiple third-party contributors. Targeted QA audits can then be used strategically, focusing on high-risk processes or critical service providers rather than
applying the same audit intensity across all parties. Training, too, should be tailored to the protocol and the specific tasks individuals will perform, rather than relying on generic GCP refreshers that may not address the real operational risks of the study. Importantly, service providers may operate under their own quality systems, provided these systems are demonstrably fit for purpose, a pragmatic approach that reflects the diversity of veterinary research environments.

A modern, R3 aligned oversight model therefore looks quite different from the traditional one. Instead of centring oversight almost exclusively on site monitoring, sponsors are encouraged to adopt a broader, risk-based perspective that includes review of data flow, system performance, and the integrity of digital tools used for data capture or sample tracking. In a study using an electronic owner diary platform, for example, oversight may involve verifying that the system timestamps entries correctly, that reminders function as intended, and that data transmission failures are promptly flagged and resolved. In another scenario, where a central laboratory performs a complex immunoassay underpinning the primary endpoint, oversight may focus on assay validation, sample logistics, and turnaround times rather than on routine site visits, ensuring that attention is directed to the processes that most influence data quality.

Contract Research Organisations (CROs) also play a central role in many veterinary studies, and proportional oversight is equally important in these partnerships. In a multi-country companion animal study outsourced to a CRO, the sponsor identified that the CRO’s monitoring plan applied identical visit frequency to all sites, regardless of enrolment volume or complexity. By applying a risk proportionate approach, the sponsor and CRO jointly redesigned the plan so that high enrolling sites received more frequent visits, while low enrolling sites were supported through remote review. This adjustment improved efficiency, reduced travel burden, and ensured that oversight was targeted where it added the most value. Similarly, laboratory vendors require oversight that reflects the criticality of their contribution. In a biologics study, a central laboratory performed a complex immunoassay underpinning the primary endpoint. Rather than relying on routine site visits, the sponsor focused oversight on assay validation, sample handling logistics, and turnaround times; an example of fit-for-purpose oversight aligned with R3
principles.

The advantages of this approach are substantial: it is more efficient, more scientifically defensible, and better aligned with the realities of contemporary veterinary research. However, it also requires a cultural shift. Sponsors must be willing to move away from “we have always done it this way” oversight models and instead apply critical thinking to determine where oversight truly adds value. When implemented thoughtfully, fit-for-purpose oversight strengthens study quality while reducing unnecessary burden on investigators, service providers, and sponsor
teams alike.

Technology and Data Governance: Computers Are Computers, Data Is Data
Technology and data governance have become central considerations in contemporary veterinary clinical research, as studies increasingly rely on electronic CRFs, digital owner diaries, wearable devices, electronic study files, and remote data capture. ICH E6(R3) provides a comprehensive framework for managing these technologies, and the principles of data integrity apply regardless of whether the data originate from human or veterinary studies. At its core, the guideline requires sponsors to prospectively define data sources and methods of capture, avoid unnecessary transcription, ensure continuous access to data for review, protect confidentiality, validate systems based on risk, and maintain audit trails that cannot be disabled or modified. These expectations reflect a simple truth: computers are computers, and data are data, irrespective of species.

The relevance of these principles to veterinary research is clear. Veterinary studies often operate across a patchwork of mixed paper and electronic systems, variable site infrastructure, owner-generated data, third-party laboratories, and multiple data streams that must ultimately converge into a coherent dataset. Without strong data governance, the risk of data loss, inconsistency, or integrity issues increases substantially. For example, in a study using wearable activity monitors to assess mobility, several sites experienced data transmission failures. Because no dataflow diagram or contingency plan had been established, the issue went undetected until late in the study, resulting in missing primary endpoint data. Under an R3 aligned approach, this risk would have been identified early, and mitigation strategies, such as automated alerts, periodic data reconciliation, or a fallback manual recording method, would have been implemented.

Similar challenges arise in studies that rely on electronic owner diaries. If the system does not timestamp entries correctly or allows retrospective editing without an audit trail, the reliability of owner reported outcomes may be compromised. Proactive data governance would require verification of system functionality, clear user access controls, and periodic review of metadata to ensure that entries are
contemporaneous and attributable. These steps not only protect data integrity but also build confidence in the robustness of outcomes derived from owner-generated data.

Farm-based technologies introduce additional complexities. In a beef cattle study using automated water intake sensors, data gaps emerged when animals moved between paddocks with inconsistent WiFi coverage. A predefined dataflow diagram and risk-based mitigation plan, such as local data buffering and periodic manual downloads, would have prevented these losses and ensured continuity of primary endpoint data. Companion animal studies face their own challenges. In a canine osteoarthritis study using an electronic owner diary, retrospective editing was
possible without an audit trail. Strengthening user access controls and verifying metadata early ensured that entries were contemporaneous and attributable, protecting the reliability of owner reported outcomes.

The advantages of applying E6(R3) data governance principles in veterinary research are significant: improved reliability, reduced risk of data loss, and greater regulatory confidence in study outcomes. However, these benefits require thoughtful implementation. Sponsors must be prepared to assess the fitness for purpose of digital tools, support sites with variable technological capability, and ensure that data governance processes are proportionate to the study’s complexity and risk. When applied well, these principles provide a modern, pragmatic framework that strengthens the scientific credibility of veterinary clinical studies while supporting operational feasibility across diverse research environments.

Conclusion: Pragmatism Is Not Just for Humans
Veterinary clinical research stands at a pivotal moment. The principles embedded in ICH E6(R3) and E8(R1) offer a clear and contemporary roadmap for designing studies that are both scientifically robust and operationally feasible. These guidelines emphasise pragmatism, proportionality, and critical thinking, qualities that are just as relevant to veterinary studies as they are to human clinical studies. Importantly,
applying these principles does not require a formal revision of VICH GCP. What it requires is a willingness within the veterinary sector to learn from these modernised frameworks and to integrate their underlying concepts into everyday thinking and practice.

Quality by Design encourages teams to focus on what truly matters for decision making and to identify critical to quality factors early. Early stakeholder engagement ensures that protocols are grounded in the realities of veterinary practice, improving feasibility, adherence, and overall study success. Risk proportionate processes ensure that oversight and effort are directed where they add the most value, rather than being spread uniformly across all activities. Risk-based quality management shifts the emphasis from detecting problems to preventing them, strengthening the reliability of study outcomes. Fit for purpose sponsor oversight broadens the lens beyond traditional site monitoring
to include laboratories, digital tools, and service providers. And modern expectations for data governance help safeguard the integrity of increasingly complex and technology enabled datasets.

Pragmatism is not a compromise. It is a commitment to designing studies that work in the real world, respect the time and expertise of investigators and site staff, and generate data that regulators and sponsors can trust. By embracing these principles now, without waiting for formal changes to GL9, the veterinary sector can enhance study quality, reduce unnecessary burden, and strengthen the scientific credibility of
the studies that shape the future of animal health.

References:
1. International Cooperation on Harmonisation of Technical Requirements for Registration of Veterinary Medicinal Products (VICH) (2000) VICH GL9: Good clinical practices.
2. International Council for Harmonisation. (2021). ICH guideline E8 (R1) on general considerations for clinical studies.
3. International Council for Harmonisation. (2025). Good clinical practice: ICH E6(R3) guideline (EMA/CHMP/ICH/135/1995).

Helen Harlow has 25+ years’ experience working in the animal health industry, in addition to 14 years’ experience in human clinical research. She is Global Quality & Training Director at GXP Engaged Auditing Services, and provides both veterinary and human GCP consultancy support to clients. Helen is a founding and current member of the RQA Animal and Veterinary Products Committee.

Suzanne Metcalfe is an independent global animal health consultant with over 20 years’ experience in VICH GCP, regulatory quality assurance, monitoring, and clinical study operations. She supports sponsors and investigator sites internationally with audit and inspection preparation, monitoring oversight, documentation and data review, and clear, compliant report writing for veterinary clinical studies.