Clinical Events Committees (CEC) and Endpoint Adjudication in Medical Device Trials

Guide to CEC endpoint adjudication in device trials: when it's required, CEC vs DMC vs core lab, charter anatomy, independence rules, and cost.

A Clinical Events Committee (CEC) — also called an Endpoint Adjudication Committee (EAC) — is an independent panel of blinded clinical experts that reviews suspected endpoint events in a clinical trial, applies pre-specified definitions, and classifies each event consistently. For high-risk, operator-dependent medical device pivotal trials with subjective or composite clinical endpoints, FDA expects independent adjudication — and since March 2026, ISO 14155:2026 has placed formal CEC requirements directly in the international device GCP standard, giving EU notified bodies an explicit basis to expect adjudication too. When endpoints are objective and instrument-measured, a CEC is usually unnecessary.

Getting the independence architecture right — structurally separate from the CRO, free of sponsor financial ties, excluding study-site investigators — matters more than the adjudication platform, because FDA reviewers scrutinize CEC methodology and independence during PMA and De Novo review.

Direct answer: If your pivotal IDE or EU MDR clinical investigation uses composite clinical endpoints (MACE, device success, stroke) or any endpoint requiring medical judgment, you almost certainly need a CEC. Constitute it during protocol development, charter it with pre-specified endpoint definitions drawn from ARC-2, VARC-3, or FDA Standardized Cardiovascular/Stroke Definitions, enforce full independence from sponsor and CRO, and define a 2–3 reviewer blinded workflow with chair-resolved disagreements. If your endpoints are objective, binary, and instrument-measured (sensor threshold, lab value, device-reported parameter), skip the CEC — it adds cost without value.

What Is a Clinical Events Committee (CEC/EAC), and What Does It Actually Do?

A CEC is a standing committee of clinical domain experts — cardiologists, neurologists, surgeons, pathologists, or other specialists matched to the device's therapeutic area — who are independent of the sponsor, CRO, and investigational sites. The committee operates under a written charter and reviews source documents for each suspected endpoint event, applying standardized, pre-specified clinical definitions to classify the event.

The CEC's core functions are:

Event classification. Reviewing investigator-reported suspected events (and events identified through other trigger sources, such as adverse event reports, core lab findings, or electronic data capture flags) and determining whether they meet the protocol's endpoint definitions.
Causal attribution. Determining whether an event is device-related, procedure-related, or unrelated — a judgment particularly complex in device trials where the boundary between device failure and procedural technique is often unclear.
Consistency enforcement. Ensuring that the same clinical criteria are applied uniformly across all sites, investigators, and regions, reducing the variability that would otherwise accumulate in a multi-center, multi-country trial.

Unlike a Data Monitoring Committee (DMC/DSMB), which reviews unblinded aggregate data and can recommend stopping or modifying a trial, the CEC works at the individual-event level and remains blinded to treatment assignment. Unlike an imaging core laboratory, which provides standardized quantitative measurements (echo parameters, angiographic diameter stenosis, CT volumetrics), the CEC makes clinical judgments that integrate imaging, laboratory, clinical, and pathological data.

The term "Endpoint Adjudication Committee" (EAC) is used interchangeably with CEC in regulatory and academic literature. FDA's 2006 guidance on Data Monitoring Committees introduced the EAC terminology alongside DMCs, reinforcing independent-committee oversight as a regulatory expectation.

Origin and Context

CECs emerged in the 1990s from large cardiovascular trials, where academic cardiologists including Robert Califf and Kenneth Mahaffey at Duke recognized that investigator-reported event classifications varied too much across sites and countries to support reliable efficacy conclusions. The PURSUIT trial (eptifibatide in acute coronary syndromes) became a landmark illustration: when a central CEC independently adjudicated suspected MI events, the classifications disagreed materially with site-investigator assessments, demonstrating that without independent adjudication, trial results could be biased by inconsistent endpoint classification.

While CECs became routine in cardiovascular drug trials by the 2000s, their adoption in medical device trials has lagged. As one industry assessment noted, independent endpoint adjudication committees remain "increasingly commonplace in pharma trials but remain a bit of a novelty in device trials" — a gap that FDA, the Cardiac Safety Research Consortium (CSRC), and MDEpiNet have worked to close.

When Does a Medical Device Trial Need a CEC vs When Is It Unnecessary?

There is no blanket FDA regulation mandating a CEC for all device trials. The necessity depends on endpoint subjectivity, device masking feasibility, operator dependence, and risk class. The following decision framework consolidates guidance from FDA's perspective paper (Farb & Zuckerman, American Heart Journal 2017), the 2016 FDA/CSRC/MDEpiNet Think Tank, the Held et al. systematic review (2018/2019), and ISO 14155:2026 §6.12, which now sets a formal standard trigger for when a CEC is required.

CEC Necessity Decision Framework

Factor	CEC Expected / Strongly Recommended	CEC Optional / Usually Unnecessary
Endpoint subjectivity	Endpoints require medical judgment: composite MACE, MI classification, stroke severity/type, device-relatedness determination, "device success" composites	Endpoints are objective, binary, instrument-measured: sensor threshold, lab value, device parameter, all-cause mortality with death certificate
Masking feasibility	Device trial cannot be masked (surgical implant, visible device, sham vs active procedure) — higher risk of investigator bias in event classification	Trial is effectively masked (indistinguishable sham device, software-only intervention with identical interface)
Operator dependence	Device performance depends on operator skill (surgical technique, catheter placement, implant positioning) — events may reflect technique vs device failure, requiring expert disambiguation	Device operates autonomously after deployment; no operator-dependent variables
Risk class	Class III / PMA / IDE significant-risk pivotal trial; high-risk EU MDR Class III; Breakthrough Device with clinical endpoints	Class I or II; early feasibility study (EFS) with safety-only endpoints; non-significant-risk investigation
Regulatory pathway	PMA, De Novo with clinical data, HDE with clinical endpoints, EU MDR Class III clinical investigation	510(k) with predicate-comparison bench/animal data only; exempt device

The key decision rule: If your trial's primary endpoint requires a physician to interpret clinical data and make a classification judgment (rather than reading a number off a calibrated instrument), you need a CEC. If the endpoint is a number from a machine, you usually do not.

FDA's Position

FDA's published perspective (Farb & Zuckerman, 2017, PMID 28888271) states that "for almost all high-risk devices, regulatory decision-making focuses on clinical study results" and that "raw clinical data must be translated into clinical events (or nonevents) through an adjudication process most often performed by a clinical events committee." This is not a binding regulation but an explicit statement of reviewer expectations.

In March 2016, the FDA Center for Devices and Radiological Health (CDRH), the Cardiac Safety Research Consortium (CSRC), and MDEpiNet co-sponsored a Think Tank titled "The Role of Endpoint Adjudication in Medical Device Clinical Trials." The resulting consensus whitepaper (published in American Heart Journal, 2017, DOI 10.1016/j.ahj.2017.05.009) recommended that "the majority of medical device studies be adjudicated" and detailed best practices for charter design, event identification, independence, and device-specific adjudication challenges.

When Adjudication Is Unnecessary

For early feasibility studies (EFS) with small sample sizes (typically fewer than 15 patients) and safety-only surveillance endpoints, a formal CEC is usually not required. Similarly, trials whose endpoints are purely instrument-measured and binary (e.g., a blood glucose monitor's analytical accuracy against a reference standard, a sensor's signal-to-noise ratio exceeding a threshold) do not benefit from adjudication because there is no clinical judgment to standardize.

Held et al.'s systematic review (PMC6454409) also documented that in trials with hard, objective endpoints, central adjudication frequently does not change results materially compared to investigator classification — adding cost without altering conclusions.

CEC vs DMC vs Core Lab: What Is the Difference, and How Do They Interact?

Device trials commonly involve three independent oversight or assessment bodies. Sponsors and clinical teams often conflate them, leading to governance gaps. The following matrix clarifies each body's role and the handoff relationships.

Dimension	CEC / EAC	DMC / DSMB	Imaging / Measurement Core Lab
Primary function	Blinded adjudication of individual suspected endpoint events	Unblinded review of aggregate interim safety and efficacy data	Standardized quantitative measurement (imaging, ECG, biomarker)
Blinding	Blinded to treatment assignment	Unblinded (sees treatment arms)	Usually blinded to treatment assignment
Scope	Individual events: "Was this an MI? Was it device-related?"	Aggregate trends: "Is the device arm showing excess mortality?"	Individual measurements: "What is the paravalvular leak grade?"
Authority	Classifies events; does not stop or modify the trial	Can recommend stopping, modifying, or continuing the trial	Provides measurements; does not classify clinical events or stop the trial
Independence	Independent of sponsor, CRO, and sites	Independent of sponsor and sites	Independent of sponsor and sites
Charter	Pre-specified endpoint definitions, workflow, quorum, disagreement resolution	Stopping rules, interim analysis schedule, communication plan	Measurement protocol, quality-control procedures, reader certification
Output	Adjudicated event dataset feeding statistical analysis	Recommendations to the sponsor (continue / modify / stop)	Quantitative measurement dataset feeding the CEC and/or SAP

The Handoff Chain

The critical interaction is: Core Lab → CEC → DMC → Statistical Analysis.

The core lab produces standardized measurements (e.g., echocardiographic paravalvular leak grade, angiographic late lumen loss, CT fracture assessment).
The CEC integrates core lab measurements with clinical data (symptoms, biomarkers, clinical presentation) to adjudicate whether an endpoint event occurred and whether it was device-related.
CEC-adjudicated events are aggregated and presented to the DMC as part of interim data reviews — the DMC does not re-adjudicate individual events.
CEC-adjudicated events also feed the statistical analysis plan (SAP) for primary and secondary endpoint analysis.

The firewall rules are equally important: the CEC should not receive unblinded treatment-arm information from the DMC, and the DMC should not attempt to override individual CEC adjudications. Both should be independent of the sponsor. For detailed guidance on DMC structure, chartering, and FDA's 2006/2024 expectations, see our companion DMC/DSMB guide.

What Must a CEC Charter Include for FDA Acceptance?

FDA reviewers do not just check whether a CEC existed — they scrutinize the charter's specificity, the committee's independence, and whether the adjudication process was pre-specified before first-patient enrollment. A poorly designed charter is a common source of FDA deficiency letters and review delays.

The following charter anatomy maps each element to the FDA scrutiny points identified in the Farb & Zuckerman perspective and the CSRC/MDEpiNet whitepaper.

CEC Charter Anatomy

1. Membership and qualifications

Specify the number of adjudicators (typically 3–7 physicians with domain expertise matching the device's therapeutic area).
Include at least one member with expertise in the device's mechanism of action and at least one with endpoint-definition expertise (e.g., a cardiologist familiar with ARC-2 MI definitions for a coronary-device trial, a neurologist familiar with modified Rankin Scale for a neurovascular device).
Document each member's qualifications and confirm no disqualifying conflicts.

2. Independence and conflict of interest

Adjudicators must have no direct financial relationship with the sponsor (consulting, equity, speaker fees).
No adjudicator may serve as a study-site investigator or sub-investigator in the trial.
If a CRO provides the CEC infrastructure, document a strict organizational firewall between the CRO's operational team and the CEC members.
Disclose and manage all other potential conflicts (advisory board membership, competitive-device consulting, institutional relationships).

3. Pre-specified endpoint definitions

Reference the standardized endpoint definitions the CEC will apply (ARC-2, VARC-3, FDA Standardized Cardiovascular and Stroke Endpoint Definitions, BARC bleeding criteria, or device-specific consensus definitions).
Include the complete decision algorithms, not just endpoint names. A charter that says "MI will be adjudicated" without specifying which MI definition (periprocedural vs spontaneous, biomarker thresholds, ECG criteria) is insufficient.
For novel devices without established consensus definitions, the charter must develop and document device-specific endpoint criteria before the trial starts.

4. Event identification and trigger sources

Do not rely solely on investigator-reported events. Specify all trigger sources: adverse event reports, core lab findings, electronic data capture flags, lab value alerts, and protocol-specified surveillance assessments.
The CSRC whitepaper emphasizes that restricting event identification to investigator reports under-ascertains events, because investigators may not recognize or report borderline cases.

5. Adjudication workflow

Define the review process: typically 2–3 independent, blinded reviewers per event.
Specify the disagreement-resolution mechanism: third-reviewer tiebreak, committee chair override, or consensus meeting.
Define turnaround times and escalation procedures for urgent events (deaths, SAEs requiring DMC notification).

6. Quorum and decision rules

Specify the minimum number of reviewers required to adjudicate an event.
Define whether decisions are by majority vote, unanimity, or chair decision after discussion.

7. Quality control and validation

A common validation practice is to externally review approximately 10% of adjudicated events (up to approximately 100 events) to measure inter-reviewer concordance and identify systematic classification drift.
Document re-adjudication triggers (new clinical information, protocol amendment changing endpoint definitions, core lab measurement revision).

8. Sponsor and DMC firewall

The CEC should not receive unblinded interim analysis results from the DMC.
The sponsor should not have access to individual CEC adjudication decisions until database lock.
Define the one-way data flow: CEC adjudicated events → aggregate → DMC review.

9. Documentation and regulatory deliverables

The charter should specify what documents the CEC will produce for the regulatory submission (adjudication methodology report, concordance/discordance analysis, final adjudicated-event dataset with audit trail).

How Does Adjudication Change Trial Results (and Does It Matter for Devices)?

The strongest argument for independent adjudication is the evidence that it materially changes results — and that sponsor decision-making can depend on which classification (investigator vs CEC) is used.

Discordance Evidence

PURSUIT trial (Mahaffey et al., 2001). In this large acute coronary syndrome trial of eptifibatide, the central CEC and site investigators disagreed on MI endpoint assessments. The CEC identified MIs that investigators missed and reclassified events that investigators had called MIs but that did not meet protocol criteria. This landmark study established that investigator-reported endpoints are systematically noisy and that adjudication is not merely a bureaucratic exercise.

CHARM-Preserved trial. In this candesartan heart failure trial, a treatment benefit was observed using investigator-adjudicated hospitalization-for-heart-failure events — but the benefit was not confirmed when a CEC independently adjudicated the same events using standardized definitions. This case demonstrated that adjudication can change the direction of a trial's conclusions, not just the magnitude.

The Counter-Argument: When Adjudication Doesn't Change Results

Held et al.'s systematic review (PMC6454409, published 2018/2019 in the Journal of Clinical Epidemiology) found that in many trials — particularly those with hard, objective endpoints and well-trained investigator networks — central adjudication did not materially change the treatment effect or the study conclusions. The review also noted that adjudication is costly and can introduce its own variability when definitions are ambiguous.

The practical implication for device sponsors: Adjudication adds the most value when (1) the endpoint requires medical judgment, (2) the trial spans many sites and countries with variable investigator experience, (3) the device cannot be masked, and (4) the regulatory submission will be scrutinized (PMA, De Novo with clinical data, EU MDR Class III). When these conditions are absent, the cost may not be justified.

How Should Sponsors Handle CEC Independence, the CRO Firewall, and FDA Scrutiny?

Independence is the single most scrutinized aspect of CEC design in FDA review. A CEC that is technically independent but operationally embedded within the sponsor's CRO, or staffed by investigators who also enroll patients in the trial, will draw deficiency questions.

Independence Architecture

The CSRC/MDEpiNet whitepaper and FDA perspective paper converge on three non-negotiable requirements:

No direct financial relationship between adjudicators and the sponsor. Consulting fees, equity, speaker honoraria, and advisory-board compensation all create conflicts. Adjudicators should be compensated only for their CEC work, at fair-market rates, through the CEC's managing organization (not the sponsor directly).
No study-site investigators as adjudicators. An investigator who enrolls patients and reports events should not also adjudicate those events — or events from other sites, because the same biases and knowledge contamination apply.
CRO firewall. If a full-service CRO both conducts the trial and provides the CEC administrative infrastructure, the charter must document a strict organizational separation: separate teams, separate data access controls, separate reporting lines. Best practice favors a CEC managed by an organization independent of the operational CRO entirely.

Practical Risk: The CRO-Integrated CEC

Many device sponsors, especially smaller companies running their first IDE pivotal trial, default to their CRO's CEC offering because it is bundled and convenient. This is not automatically disqualifying, but it raises FDA scrutiny risk. The charter must demonstrate that the CRO's CEC team cannot access unblinded data held by the CRO's data management team, that the CRO's project management cannot influence adjudication decisions, and that the adjudicators themselves have no financial relationship with the sponsor beyond CEC compensation.

How Do Endpoint Definition Standards (ARC-2, VARC-3, FDA Standardized Definitions) Fit In?

A CEC is only as good as its endpoint definitions. Vague or inconsistent definitions are the primary source of adjudication variability and regulatory challenge. The medical device field benefits from several consensus standardized-definition frameworks:

Key Standardized Endpoint Definition Frameworks

Academic Research Consortium-2 (ARC-2) Garcia-Garcia et al., Circulation 2018;137(24):2635–2650 (PMID 29891620). Provides standardized endpoint definitions for coronary intervention trials: periprocedural MI (Type 4a), spontaneous MI (Type 1), stent thrombosis (definite, probable, possible), target-lesion revascularization (TLR), and target-vessel revascularization (TVR). ARC-2 is the default framework for coronary device CECs (drug-eluting stents, bioresorbable scaffolds, coronary atherectomy devices).

Valve Academic Research Consortium-3 (VARC-3) Genereux et al., JACC 2021;77(21):2717–2746 (PMID 33888385). Standardized definitions for transcatheter and surgical aortic valve trials: device success, early safety (30-day), clinical efficacy (1-year), time-related valve safety, paravalvular leak grading, and structural valve deterioration. VARC-3 is the reference framework for TAVR and SAVR device CECs.

FDA Standardized Cardiovascular and Stroke Endpoint Definitions Hicks et al., Circulation 2018;137(2):e61–e72 (DOI 10.1161/CIRCULATIONAHA.117.033502). Originally developed through a CDRH/ACC/AHA collaboration (public drafts circulating since 2009), these definitions standardize CV death classification, MI, stroke (ischemic, hemorrhagic, undetermined), stent thrombosis, and bleeding for FDA submissions. They represent FDA-endorsed (though not FDA-mandated) consensus and are used for CEC adjudicator training.

BARC Bleeding Definitions The Bleeding Academic Research Consortium (BARC) definitions classify periprocedural and spontaneous bleeding events by severity (Types 0–5). Used by CECs in trials of antithrombotic devices, vascular access/closure devices, and any device trial where bleeding is a safety endpoint.

Beyond Cardiology: Device-Specific Endpoint Challenges

For devices outside cardiovascular medicine, standardized consensus definitions may not exist. CECs for orthopedic implant trials may adjudicate revision surgery, implant loosening, or periprosthetic fracture using device-specific criteria developed in the charter. Neuromodulation device CECs may classify responder status using patient-reported outcome thresholds. Aesthetic device CECs may adjudicate treatment success using blinded expert panels applying photographic grading scales.

When no published consensus definitions exist, the charter must develop and document device-specific endpoint criteria — and these criteria become subject to the same FDA scrutiny as standardized definitions. Sponsors should engage clinical experts and biostatisticians during protocol development to ensure definitions are pre-specified, clinically meaningful, and operationally reproducible.

How Do EU MDR and ISO 14155 Treat Endpoint Adjudication?

EU MDR (Regulation 2017/745) requires clinical investigations for high-risk devices to follow Good Clinical Practice, and ISO 14155 is the device-specific GCP standard the regulation points to. As of March 23, 2026, ISO 14155:2026 (the fourth edition) is the current published edition, replacing ISO 14155:2020 with no transition period — so any new clinical investigation must apply it from the outset, and ongoing investigations should be gap-assessed against it.

The 2026 edition is significant for this topic because, for the first time, it places Clinical Events Committees directly in the standard. Section 6.12 introduces formal CEC requirements: a sponsor must establish a CEC where clinical event classification could be subject to interpretation bias, where events are complex and require specialist clinical judgment, or where event adjudication directly affects primary endpoint assessment. ISO 14155:2026 also specifies that CEC members must be independent of the sponsor and the investigation sites, operate under a documented charter (event definitions, adjudication process, blinding, disagreement resolution, and reporting obligations), and make decisions without knowledge of treatment assignment where the design permits.

This is a meaningful shift for device sponsors. Before 2026, independent adjudication was an FDA reviewer expectation and a best-practice recommendation from the CSRC/MDEpiNet Think Tank — but it was not a black-letter requirement in the international device GCP document. Under ISO 14155:2026, a CEC is now a normative expectation for appropriate EU MDR clinical investigations, giving EU notified bodies and national competent authorities an explicit standard basis to expect adjudication rather than relying only on the parallel FDA expectation.

ISO 14155:2026 also reinforces the surrounding requirements any CEC must satisfy:

Independent safety oversight for clinical investigations (provided by a DMC, a CEC, or both depending on design — and the DMC must now define conditions for stopping the study under §6.11).
Pre-specified endpoint definitions in the clinical investigation plan (CIP).
Documented procedures for adverse event assessment and device-relatedness determination.
Source data verification and audit-trail requirements that apply equally to CEC-adjudicated datasets.

The regulatory timing is worth noting for planning: on January 28, 2026, the European Commission published Implementing Decision (EU) 2026/193, formally harmonizing the prior EN ISO 14155:2020/A11:2024 with EU MDR — the first harmonization since the MDR entered force — only for ISO 14155:2026 to supersede it weeks later. For a full walkthrough of the standard's GCP requirements, see our dedicated ISO 14155:2026 guide.

For guidance on how adjudicated trial results feed the EU MDR Clinical Evaluation Report (CER), see our CER guide.

What Does Endpoint Adjudication Cost, and Can Registry Data Reduce the Burden?

Cost is the most common sponsor objection to independent adjudication, and it is a legitimate concern. The CSRC/MDEpiNet Think Tank acknowledged the cost barrier; Held et al.'s systematic review argued that when adjudication does not change results, it represents an avoidable expense.

Cost Drivers

CEC costs are driven by:

Adjudicator compensation. Fair-market-value payments to 3–7 senior clinical experts for per-event review, training, charter meetings, and consensus meetings. Device trials with high event rates (cardiovascular MACE composites in large trials) can generate hundreds or thousands of events requiring adjudication.
Infrastructure. Secure, validated adjudication platforms (21 CFR Part 11 and GDPR-compliant electronic systems for document exchange, reviewer assignment, blinding maintenance, and audit trail). Many sponsors use vendor platforms from CROs or specialized adjudication service providers.
Operational complexity. Source document collection, redaction (to maintain blinding), distribution, tracking, query resolution, and re-adjudication management.
Scale. Vendor self-reported figures (illustrative, not authoritative, as of mid-2026) suggest the scale of established CEC operations: Cardialysis reports managing CECs for over 100 cardiology trials with over 90,000 potential events adjudicated; Stanford's Department of Medicine Clinical Research Hub reports involvement in over 210 trials with over 120,000 endpoints adjudicated; the TIMI Study Group has handled trials ranging from fewer than 100 to more than 20,000 events adjudicated.

Can Registry Data or Alternative Designs Reduce CEC Costs?

Held et al.'s systematic review raised the possibility that national patient registries (such as Sweden's SWEDEHEART, Denmark's Western Denmark Heart Registry, or the U.S. STS/ACC TVT Registry) might reduce the need for prospective adjudication by providing independently validated outcome data through routine healthcare records.

For medical device trials, this approach has limitations:

Registry endpoint definitions may not align with the trial's protocol-specified definitions.
Registry data capture may not include the granularity (imaging, biomarkers, clinical narrative) that a CEC needs for classification.
Regulatory acceptance of registry-adjudicated endpoints varies by jurisdiction and device type.

However, for post-marketing studies, PMCF investigations, and real-world evidence (RWE) submissions, registry-based outcome validation is increasingly accepted — particularly when the registry itself employs standardized endpoint definitions and its own adjudication process.

Cost-Benefit Decision Rule

Invest in a CEC when:

The trial is a pivotal IDE or EU MDR Class III clinical investigation.
Primary endpoints require medical judgment.
The device cannot be masked.
The regulatory submission will be scrutinized (PMA, De Novo, MDR Class III).
The cost of a failed or delayed submission (from endpoint credibility challenges) exceeds the cost of adjudication.

Consider alternatives when:

Endpoints are objective and instrument-measured.
The trial is early feasibility, low-risk, or not intended for a marketing application.
A validated registry provides independently adjudicated outcome data.

Practical CEC Implementation Checklist for Device Sponsors

For sponsors designing a pivotal device trial and establishing a CEC for the first time, this checklist consolidates the charter, independence, workflow, and regulatory requirements discussed above.

Pre-Protocol Phase

Determine whether your endpoints require adjudication using the necessity decision framework above.
Identify which standardized endpoint definitions apply (ARC-2, VARC-3, FDA Standardized Definitions, BARC, or device-specific definitions to be developed).
If no consensus definitions exist, engage domain experts and biostatisticians to develop device-specific endpoint criteria before finalizing the protocol.

Charter Development (Before First-Patient Enrollment)

Draft the CEC charter with all elements from the charter anatomy above.
Specify membership qualifications and confirm no disqualifying conflicts.
Document the complete adjudication workflow: reviewer assignment, blinding maintenance, disagreement resolution, turnaround times.
Define event-identification trigger sources (not just investigator reports).
Define the quality-control/validation process (e.g., 10% external review).
Document the sponsor and DMC firewall.
Have the charter reviewed by regulatory counsel and the biostatistician responsible for the SAP.

Operational Phase

Train all adjudicators on the endpoint definitions and the adjudication platform before the first event.
Implement blinding procedures: redact treatment-identifying information from source documents before CEC review.
Establish the one-way data flow: CEC adjudicates → aggregate adjudicated data → DMC → SAP.
Track concordance between adjudicators; investigate systematic drift if inter-reviewer agreement falls below protocol-specified thresholds.
Document all re-adjudications and the reasons for each.

Regulatory Submission Phase

Prepare the CEC methodology report for inclusion in the PMA/De Novo/MDR technical dossier.
Include concordance/discordance analysis (investigator vs CEC classifications).
Provide the complete audit trail of adjudicated events.
Be prepared for FDA review questions about CEC independence, event-identification completeness, and definition consistency.

Frequently Asked Questions

What is the difference between a CEC and a DMC (DSMB)?

A CEC/EAC performs blinded adjudication of individual suspected endpoint events — determining whether each event meets the protocol's clinical definitions (e.g., was this an MI? was it device-related?). The CEC does not see unblinded treatment assignments and does not stop or modify the trial. A DMC/DSMB reviews unblinded aggregate interim safety and efficacy data across treatment arms and can recommend stopping, modifying, or continuing the trial. CEC-adjudicated events typically feed the DMC's interim analysis reports. Both committees should be independent of the sponsor.

Does FDA require a CEC for medical device trials?

There is no blanket FDA regulation mandating a CEC for every trial. However, FDA's published perspective (Farb & Zuckerman, American Heart Journal 2017, PMID 28888271) states that for almost all high-risk devices, raw clinical data must be translated into events through an adjudication process most often performed by a CEC, and the 2016 FDA/CSRC/MDEpiNet Think Tank recommended that the majority of medical device studies be adjudicated. Since March 2026 this has moved from expectation toward requirement on the GCP-standards side: ISO 14155:2026 §6.12 now requires a sponsor to establish a CEC where event classification could be subject to interpretation bias, requires specialist clinical judgment, or directly affects primary endpoint assessment — so for EU MDR clinical investigations meeting those triggers, adjudication is now a standard requirement, not merely a recommendation.

Can the CRO running our trial also run the CEC?

It is feasible but raises FDA scrutiny risk. A CEC nested inside the same CRO that conducts the trial can create an independence problem. Best practice favors adjudicators who have no financial relationship with the sponsor or the operational CRO and who do not serve as study-site investigators. If a full-service CRO provides the CEC, the charter must document a strict organizational firewall between the CRO's data management and the CEC, and the adjudicators themselves must be free of sponsor financial ties beyond CEC compensation.

When is a CEC NOT needed?

When endpoints are objective, binary, and measured by calibrated instrumentation without medical judgment — such as a device-reported parameter exceeding a defined threshold, a blood test result, or a yes/no sensor reading — and the trial is low-risk or early feasibility. Adjudication adds the most value when endpoints require clinical interpretation (stroke classification, MI adjudication, device-relatedness determination, composite MACE) or when the device trial cannot be masked.

How many reviewers should adjudicate each event?

Typically 2–3 independent, blinded reviewers per event. Disagreements are resolved by a third reviewer, committee chair decision, or consensus meeting. A common quality-control practice is to externally review approximately 10% of adjudicated events (up to approximately 100 events) to measure concordance and identify systematic drift.

Which endpoint definitions should our device CEC use?

Use the standardized consensus framework that matches your device's therapeutic area: ARC-2 for coronary interventions, VARC-3 for aortic valve devices, FDA Standardized Cardiovascular and Stroke Endpoint Definitions for broader cardiovascular/neurovascular trials, BARC for bleeding endpoints. For devices outside these areas (orthopedic, neuromodulation, aesthetic, IVD), develop device-specific criteria in the charter with domain-expert input and document them before first-patient enrollment.

Key Sources

Farb A, Zuckerman BD. Clinical event adjudication in cardiovascular device trials: An FDA perspective. American Heart Journal. 2017;191:62–64. PMID 28888271.
FDA/CSRC/MDEpiNet Think Tank. Use of endpoint adjudication to improve the quality and validity of endpoint assessment for medical device development and post marketing evaluation. American Heart Journal. 2017. DOI 10.1016/j.ahj.2017.05.009.
FDA. Guidance for Clinical Trial Sponsors: Establishment and Operation of Clinical Trial Data Monitoring Committees. March 2006.
Garcia-Garcia HM et al. Standardized End Point Definitions for Coronary Intervention Trials: The Academic Research Consortium-2 (ARC-2). Circulation. 2018;137(24):2635–2650. PMID 29891620.
Genereux P et al. Valve Academic Research Consortium 3 (VARC-3): Updated Endpoint Definitions. JACC. 2021;77(21):2717–2746. PMID 33888385.
Hicks KA et al. 2017 Cardiovascular and Stroke Endpoint Definitions for Clinical Trials. Circulation. 2018;137(2):e61–e72. DOI 10.1161/CIRCULATIONAHA.117.033502.
ISO 14155:2026. Clinical investigation of medical devices for human subjects — Good clinical practice (4th edition). Published 2026-03-23, replacing ISO 14155:2020 with no transition period; §6.12 introduces formal Clinical Events Committee requirements.
Held C et al. When do we need clinical endpoint adjudication in clinical trials? Journal of Clinical Epidemiology. 2018/2019. PMC6454409.
Mahaffey KW et al. Disagreements between central CEC and site investigator assessments of MI endpoints: PURSUIT study. Curr Control Trials Cardiovasc Med. 2001;2(4):187–194.
Lopes RD et al. Clinical event committees in coronary stent trials: insights and recommendations (RESOLUTE All-Comer). EuroIntervention.