ICH E8(R1) Critical-to-Quality Endpoints: Protocol Design Rules

Every clinical development team faces an operational crossroad during study design: which endpoint attributes need quality-by-design protection, and which remain ordinary protocol detail? Attempting to classify every secondary assessment, exploratory biomarker, and procedural schedule as critical to quality dilutes operational focus, overburdens investigator sites, and obscures genuine risks to study integrity. Conversely, leaving truly critical endpoint features unspecified until after the protocol is frozen exposes the study to design or conduct errors that, if they undermine those attributes, would also undermine the reliability or ethics of decisions based on the results.

This guide answers one specific operational question: which endpoint design choices does ICH E8(R1) treat as critical-to-quality (CtQ) factors during study planning, and how should they be specified and locked before procedures and data capture are frozen? Answering this question correctly allows clinical development leaders, biostatisticians, and clinical operations heads to build quality directly into the scientific architecture of the study, rather than relying on retrospective audits or extensive on-site monitoring to clean flawed data after the fact.

To maintain strict methodological boundaries, this guide focuses entirely on the prospective planning phase governed by ICH E8(R1). It does not retell the conduct-phase verification routines detailed in our ICH E6(R3) Endpoint Implementation Checklist, nor does it replace the mathematical specifications covered in ICH E9(R1) Estimands and Intercurrent Events or Statistical Analysis Plans for Endpoints. Furthermore, patient consultation under E8(R1) informs study feasibility; it does not substitute for empirical content validity research under FDA PFDD Guidance 2 or fit-for-purpose evidence under FDA PFDD Guidance 3. Crucially, classifying an endpoint as critical to quality is an internal governance priority—it does not constitute regulatory acceptance, qualification, device clearance, or approval of an endpoint for labeling claims.

The International Council for Harmonisation (ICH) adopted ICH E8(R1) General Considerations for Clinical Studies at Step 4 on 6 October 2021, formally replacing the legacy 17 July 1997 E8 guideline. The revision represents the foundational cornerstone of the ICH GCP Renovation initiative launched in 2017, designed to modernize clinical trial design, accommodate diverse data sources, and instill contemporary concepts of fit-for-purpose data quality across the clinical development lifecycle.

On 4 February 2022, the ICH Management Committee approved a minor editorial correction to Section 5.6 (fourth paragraph, page 16), which added the word plan to clarify statistical analysis plan terminology. This editorial correction was purely administrative; it did not alter any substantive quality-by-design principles, endpoint rules, or regulatory expectations.

Regional regulatory authorities implemented ICH E8(R1) as formal guidance:

• Food and Drug Administration (FDA): Issued E8(R1) as final guidance for industry in April 2022 through CDER and CBER (Docket No. FDA-2019-D-3049). The Federal Register notice of availability on 11 April 2022 (87 FR 21127) finalized the 1 August 2019 draft (84 FR 37649). FDA states that the guidance represents current thinking, does not have the force and effect of law, and is not binding on FDA or the public unless a cited statute or regulation applies. In Agency guidance, "should" means suggested or recommended, not required.

• European Medicines Agency (EMA): Published E8(R1) as Step 5 scientific guideline EMA/CHMP/ICH/544570/1998 on 14 October 2021, with date for coming into effect 14 April 2022. EMA lists ICH E8 (R1) as the current effective version of General considerations for clinical studies. That is regional implementation of the ICH text, not a finding that any named endpoint is accepted.

• Interplay with ICH E6(R3): The consolidated Step 4 guideline of ICH E6(R3) (16 June 2026) builds on key concepts outlined in E8(R1). Principle 6 states that quality should be built into scientific and operational design and conduct. Principle 6.1 treats quality as fitness for purpose. Principle 6.2 and Section 3.10 (Sponsor Quality Management) state that factors critical to trial quality should be identified prospectively as described in ICH E8(R1), then identify risks that may have a meaningful impact on those factors before initiation and throughout conduct. E8(R1) is the planning method for what to protect; E6(R3) carries those factors into conduct. Neither document, by itself, accepts a named endpoint.

Section 3.2 of ICH E8(R1) defines critical-to-quality factors with exact precision: they are attributes of a study whose integrity is fundamental to the protection of study participants, the reliability and interpretability of the study results, and the decisions made based on the study results. These factors are critical because, if their integrity were compromised by errors in design or operational conduct, the safety of participants or the validity of scientific and regulatory decisions drawn from the trial would be undermined.

A common error among clinical trial teams is treating every variable in the protocol as critical to quality. Section 3.2 explicitly cautions against this tendency:

The critical to quality factors should be clear and should not be cluttered with minor issues (e.g., due to extensive secondary objectives or processes/data collection not linked to the proper protection of the study participants and/or primary study objectives).

Quality by design asks sponsors to accept that perfection across every operational parameter is rarely achievable and often can be achieved only with resources out of proportion to the benefit. Instead, sponsors should prioritize the factors that stand out as genuinely critical for the specific trial at hand. Different factors will stand out depending on the clinical phase, patient population, intervention novelty, and disease setting.

Furthermore, clinical teams should avoid rigid vendor toolkits that advertise fixed catalogs of CtQ factors. For example, industry publications and software vendors frequently cite a rigid catalog of 18 mandatory ICH CtQ factors. This framing misrepresents the guideline. Section 7 of ICH E8(R1) provides an illustrative, non-exhaustive list of considerations that may be evaluated where applicable. Similarly, the Clinical Trials Transformation Initiative (CTTI) Quality by Design Principles Document provides an inquiry framework for cross-functional teams, not a regulatory checklist. In Section 3.3.1, ICH E8(R1) explicitly instructs sponsors to go beyond sole reliance on tools and checklists, discouraging one-size-fits-all templates that stifle critical thinking.

ICH E8(R1) Section 5 describes design elements of clinical studies. Section 5.4 defines study endpoints as the response variables chosen to assess drug effects; the response variable may relate to pharmacokinetics, pharmacodynamics, efficacy, or safety, or to use of the drug, including adherence to post-approval risk-minimisation measures. Among the Section 3 design elements that support quality are endpoints that are well-defined, measurable, clinically meaningful, and relevant to patients:

1. Well-Defined: The response variable needs an unambiguous definition. Section 5.4 states that the definition of each study endpoint should be specific. E8(R1) does not itself prescribe composite algorithms, laboratory diagnostic criteria, or baseline-reconciliation rules; those details belong in the protocol when they are how the chosen endpoint is ascertained.

2. Measurable: Section 7 asks whether the choice of response variables and the methods to assess them are well-defined and support evaluation of the effects of the drug. E8(R1) does not set numeric reliability thresholds, declare a named instrument validated, or treat a bioanalytical method as qualified by being listed as CtQ.

3. Clinically Meaningful: The primary endpoint should be capable of providing clinically relevant and convincing evidence related to the primary objective of the study (ICH E9). E8(R1) does not equate "clinically meaningful" with a qualified surrogate, a Drug Development Tool qualification, or any named clinical outcome assessment.

4. Relevant to Patients: The choice of endpoints should be meaningful for the intended population and may take into account the views of patients. That patient input informs design; it is not PFDD fit-for-purpose evidence for a named instrument.

Those four attributes are how E8(R1) expects endpoints to be specified. They are not a pass/fail test that automatically makes every well-specified endpoint a CtQ factor. Section 3.2 still requires a study-specific judgment: only attributes whose integrity is fundamental to participant protection, to the reliability and interpretability of results, and to the decisions based on those results should stand out as critical. Different factors stand out as critical for different types of studies.

Section 5.4 states that the protocol definition for each study endpoint should be specific and include how and at what time points in a participant's treatment course and follow-up it is ascertained. Naming the instrument without saying how and at what time points the endpoint is ascertained leaves the definition incomplete under Section 5.4. E8(R1) does not publish a universal visit-window width; windows belong in the protocol only when they are part of how that study ascertains the endpoint.

Not all study endpoints share the same evidentiary burden or CtQ status. Section 5.4 and the ICH E9 primary-variable rule it incorporates establish distinct roles; Section 3.2 still requires a study-specific judgment about which of those roles, if undermined, would also undermine participant protection or the decisions made from the results:

• Primary Endpoint: Should be capable of providing clinically relevant and convincing evidence related to the primary objective (E8(R1) Section 5.4, pointing to ICH E9). ICH E9 states that there should generally be only one primary variable, specified in the protocol with the rationale for selection. Redefinition after unblinding will almost always be unacceptable. That primary-variable rule is a statistical-principles recommendation; it is not a finding that the chosen primary is accepted, and it does not make every primary a CtQ factor in every study type.

• Secondary Endpoints: Are either supportive measurements related to the primary objective or measurements of effects related to the secondary objectives. A key secondary that is itself fundamental to the decisions the study will support may stand out as critical. Extensive minor secondaries should not clutter the CtQ list (Section 3.2). Formal multiplicity and labeling-claim hierarchy are a separate job; see Multiple Endpoints and Error Control. They are not automatically inherited from the primary endpoint's CtQ status.

• Exploratory Endpoints: Used to further explain or support findings or to explore new hypotheses for later research. They should not automatically inherit primary-endpoint CtQ status. If they are collected, Section 5.4 still expects a specific how-and-when definition; E8(R1) does not authorize leaving them unspecified.

• Pharmacodynamic vs Confirmatory Outcomes: Section 5.4 distinguishes a short proof-of-concept study that may employ a pharmacodynamic outcome rather than the outcome of primary interest (ICH E9) from a larger, longer study used to confirm a clinically meaningful effect on that outcome of primary interest. The pharmacodynamic outcome does not automatically become the confirmatory clinical endpoint.

Section 3.3 of ICH E8(R1) outlines a method for identifying CtQ factors during study design. It begins by asking whether the research questions are clearly articulated, scientifically valid, and meaningful to patients, and whether those objectives can be met well and most efficiently by the chosen design and data sources. Identification is enhanced by five supporting elements:

1. 3.3.1 Fostering a Culture of Critical Thinking: Sponsors are encouraged to value critical thinking and open dialogue about what is critical for that study, going beyond sole reliance on tools and checklists. Inflexible one-size-fits-all approaches should be discouraged. Evidence used to inform design should be reviewed transparently, including known gaps and conflicting data.

2. 3.3.2 Focusing on Essential Activities: Efforts should focus on activities essential to reliable, meaningful outcomes and to safe, ethical conduct. Consideration should be given to eliminating nonessential activities and data collection so resources target errors that matter.

3. 3.3.3 Engaging Diverse Stakeholders Early: Design is best informed by patients, healthcare providers, investigators, coordinators, and site staff, and should be open to challenge from inside and outside the sponsor. Site staff and participants can speak to eligibility feasibility, visit and procedure burden, dropouts, and the general relevance of study endpoints and settings. Early engagement with regulatory authorities is encouraged particularly when novel populations, procedures, or endpoints are themselves considered critical to quality; that sentence is in Section 3.3.3, not Section 3.3.5.

4. 3.3.4 Periodic Review and Dynamic Adaptation: Accumulated experience, together with periodic review of CtQ factors, should be used to determine whether adjustments to risk control are needed, because new issues may arise after start. Studies with adaptive features or interim decision points need specific attention during planning and ongoing review (ICH E9).

5. 3.3.5 Ensuring Operational Feasibility: The foundation is a protocol that is both scientifically sound and operationally feasible. Feasibility includes regional practice and population differences, qualified investigators and sites (ICH E6), equipment and facilities, target-population availability, and ability to enrol enough participants. E8(R1) states that retention and follow-up of study participants are also key critical-to-quality factors.

flowchart TD
  A["Study Primary Objective & Scientific Question"] --> B["Identify Candidate Endpoints"]
  B --> C{"Is integrity fundamental to protection, interpretable results, and decisions?"}
  C -- "Yes" --> D["Designate as Critical-to-Quality Factor"]
  C -- "No (exploratory / minor secondary)" --> E["Keep as protocol detail; do not clutter CtQ"]
  D --> F["Specify how and at what time points it is ascertained"]
  F --> G{"Is the endpoint operationally feasible and acceptable to patients?"}
  G -- "Feasibility gaps / excessive burden" --> H["Refine schedule / remove nonessential procedures"]
  H --> F
  G -- "Feasible" --> I["Lock estimand, key analyses, and sensitivity analyses in the protocol"]
  I --> J["Detail operational analysis specifications in the SAP before results are known"]
  J --> K["Manage risks to those factors during conduct under ICH E6(R3)"]

A frequent planning error is postponing critical endpoint decisions from the protocol to the Statistical Analysis Plan (SAP). ICH E8(R1) Section 5.6 draws the design-stage boundary using "should," not a statute.

Section 5.6 states that the principal statistical features of the study should be planned during design and specified in a protocol written before the study begins (ICH E9). In accordance with ICH E9 and ICH E9(R1), the protocol should:

• Estimand Definitions: The protocol should define the estimand(s) following the ICH E9(R1) framework. E9(R1) describes an estimand through the treatment, population, variable, intercurrent-event strategy, and population-level summary. A trial protocol should specify explicitly a primary estimand corresponding to the primary objective; estimands for secondary objectives that are likely to support regulatory decisions should also be specified. E9(R1) is explicit that it is not a regulatory requirement to document an estimand for each exploratory objective.

• Key Efficacy and Safety Analyses: Analyses of primary and secondary endpoints that address key efficacy and safety objectives should be described in the protocol, including any interim analyses and planned design adaptations, together with the analytical methods for planned estimation and tests of hypotheses.

• Sample-Size Justification: A justification of the sample size should be described in the protocol. E8(R1) does not prescribe which variance, withdrawal, or effect-size inputs to use.

• Interim Analyses and Adaptations: Interim analyses and planned design adaptations should be described in the protocol. E8(R1) does not publish a required futility boundary, efficacy stopping rule, or sample-size re-estimation method.

• Prespecified Sensitivity Analyses: The statistical analysis should include pre-specified sensitivity analyses for the assumptions made for the primary and important secondary analyses (E9(R1)). If the analysis relies on a particular assumption about the reasons for missing data, sensitivity analyses should be planned to assess the impact of that assumption. E8(R1) does not require a missing-at-random versus departure pair by name.

Full details of the planned statistical analysis should be specified and documented before knowledge of the study results that may reveal drug effects, which may be done in a separate SAP. For double-blind studies, E8(R1) states that the SAP should be finalised before treatment assignments are revealed; if a study includes unblinded interim analyses, the planned statistical analysis should not be changed after that look. For open-label and single-blind studies, details of the primary and important secondary analyses would ideally be finalised before the first participant is randomised. The published Statistical Analysis Plans for Endpoints article covers that unblinding-timing job. This page only records the E8(R1) design-stage split: do not use the SAP to invent an endpoint or estimand that the protocol never defined.

Section 3.3.3 of ICH E8(R1) encourages sponsors to engage with regulatory authorities early, particularly when a study has novel elements considered critical to quality, for example defining patient populations, procedures, or endpoints. Section 2 separately encourages early engagement to understand local and regional requirements. Section 3.3.5 is the operational-feasibility discussion, including retention and follow-up; it is not the novel-endpoint advice sentence. Section 7 is a where-applicable consideration list, not a meeting-type catalog.

Regional advice procedures, such as an FDA Type C meeting or EMA scientific advice, are vehicles some sponsors use for that dialogue. E8(R1) does not name those meeting types and does not treat the conversation as a qualification consultation. Early engagement is an advice-seeking opportunity, not a definitive ruling. Reviewers evaluate the plausibility of the sponsor's rationale, provide guidance on potential evidentiary pitfalls, and highlight operational risks.

Quality by design under ICH E8(R1) is an internal methodological discipline that focuses trial resources on what matters most. Once the protocol is frozen and clinical conduct begins, E6(R3) is the handoff for protecting those factors during execution; see ICH E6(R3) Good Clinical Practice controls. Under E6(R3), risks to those prospectively identified CtQ factors are managed during conduct, including computerized systems that capture endpoints (see Validating Computerized Systems for Endpoint Capture). Implementing those conduct controls does not substitute for the E8(R1) design-stage choice.

By specifying patient-relevant endpoint definitions, locking estimand and sensitivity analyses in the protocol, and keeping the CtQ list uncluttered by nonessential secondaries, clinical development leaders focus resources on the attributes whose integrity E8(R1) treats as fundamental. That design-stage classification still does not qualify, clear, or label the measure.

This guidance is grounded exclusively in public regulatory guidelines, international harmonisation standards, and established scientific frameworks:

• ICH Harmonised Guideline: General Considerations for Clinical Studies E8(R1) — International Council for Harmonisation, Step 4 final guideline adopted 6 October 2021, including the 4 February 2022 editorial correction.

• FDA Guidance for Industry: E8(R1) General Considerations for Clinical Studies — U.S. Food and Drug Administration (CDER/CBER), April 2022 (Docket No. FDA-2019-D-3049).

• Federal Register Notice of Guidance Availability (87 FR 21127) — Announcement of finalized E8(R1) guidance for industry, 11 April 2022.

• EMA Scientific Guideline: ICH Guideline E8 (R1) on General Considerations for Clinical Studies — European Medicines Agency (EMA/CHMP/ICH/544570/1998), Step 5 adoption 14 October 2021, effective 14 April 2022.

• ICH Harmonised Guideline: Guideline for Good Clinical Practice E6(R3) — Consolidated Step 4 guideline, 16 June 2026, pointing CtQ identification back to E8(R1) during planning and risk management during conduct.

• ICH Harmonised Guideline: Statistical Principles for Clinical Trials E9 — Step 4 guideline, 5 February 1998, establishing primary variable selection standards.

• ICH Harmonised Guideline: Addendum on Estimands and Sensitivity Analysis E9(R1) — Step 4 addendum, 20 November 2019, governing estimand specification and sensitivity analyses.

• Clinical Trials Transformation Initiative (CTTI): Quality by Design Project — Multi-stakeholder project and inquiry principles informing quality-by-design implementation.