chapter 01 · the regulatory spine
The eight GCP pillars · seven regulators.
Quick-reference grid of the eight Good-Clinical-Practice pillars · then the cross-regulator comparison drilldown for each. ICH E6(R3) Step 4 (6 January 2025) · FDA · EMA · PMDA · CDSCO · MHRA · Health Canada. The flagship chapter for the clinical-trial spine.
/ 00
The deep reference: ICH E6(R3) Step 4.
Adopted 6 January 2025 · liveThe clinical-trial spine pivoted on 6 January 2025, when the ICH Assembly adopted E6(R3) at Step 4. R3 is a structural rewrite of the global Good Clinical Practice standard, not an addendum: twelve overarching principles, Annex 1 (Interventional) at Step 4 alongside the principles, and Annex 2 (Non-traditional designs) still in development. The 2016 R2 addendum stays in force until R3 is regionally implemented — FDA via guidance through 2025–2026, EMA scientific guideline EMA/CHMP/ICH/135/1995 superseding the R2-era version. Every other major regulator now sits in a defined posture relative to R3.
This chapter sits each of the eight operational GCP pillars against seven regulators side by side. Pick a pillar. Read the convergence/divergence summary. Compare ICH E6(R3) · FDA 21 CFR 312/50/56 · EMA CTR 536/2014 + CTIS · PMDA J-GCP · CDSCO NDCT 2019 · MHRA UK Clinical Trials Regulations · Health Canada Division 5. The pillars were chosen for what auditors actually find findings against in 2026 inspections.
/ Foundation document
7 regulators · 8 GCP pillars · one reference.
Designed for the sponsor regulatory lead, the QA auditor, the CRO project manager, the cross-region investigator-site. Where regulators converge, treat the pillar as harmonised. Where they diverge, the divergence is the story — and the audit risk.
ICH E6(R3)FDAEMA CTRPMDACDSCOMHRAHealth Canada
/ 00a
The twelve overarching principles of E6(R3).
Adopted 6 January 2025R3 starts with twelve principles that govern every operational pillar below. They do not replace the pillars — they sit above them. R3 is the first GCP text to make quality-by-design, risk proportionality, and participant-centric language doctrinal rather than aspirational. Where Annex 1 is silent, the principles bind.
i
Conduct in accordance with ethical principles.
Declaration of Helsinki at the apex. Belmont and CIOMS recognised. Local ethics review independent of sponsor.
ii
Informed consent before any trial-related activity.
Voluntary, current, documented. Re-consent on material amendment. Capacity assessment for vulnerable participants.
iii
Independent ethical review (IRB/IEC).
Composition, scope, working procedures, decision documentation. Approval before any participant enrols.
iv
Scientific soundness; clear, detailed protocol.
Protocol drives every other artefact. Pre-specified endpoints; statistical analysis plan locked before unblinding.
v
Qualified investigators; medical decisions by qualified physicians.
Investigator CV + delegation log. 21 CFR 312.53 in the US; J-GCP investigator qualification dossier in Japan.
vi
Quality by design; risk proportionate to participant.
R3's signature shift. Critical-to-quality factors identified at protocol design, not retroactively at audit.
vii
Operations proportionate to risk.
Risk-based monitoring (RBM) replaces 100% SDV as default. ICH E6(R2) §5.0 lineage retained.
viii
Reliable trial results.
ALCOA+ data integrity (Attributable, Legible, Contemporaneous, Original, Accurate, plus Complete, Consistent, Enduring, Available).
ix
Roles and responsibilities clear.
Sponsor, investigator, vendor, IRB. Delegation must be documented at the level of the activity.
x
Investigational products manufactured under GMP.
Storage, handling, accountability traceable. ICH Q7/Q9/Q10 lineage on the product side.
xi
Systems and processes assure quality.
SOPs, training records, vendor oversight, computerised system validation (21 CFR Part 11 / EU Annex 11).
xii
Trial registration and results disclosure.
ClinicalTrials.gov / EU CTIS / Japan jRCT / India CTRI. Lay summary required under EU CTR within 12 months of trial end.
/ 01
The eight GCP pillars.
The operational spineThese eight pillars are what every clinical trial must satisfy to be GCP-complete. Below: a quick-reference grid of all eight · then the cross-regulator drilldown for each. Informed consent and safety reporting are where most trials draw the most-frequent inspection findings; data integrity (★) is where the cross-regulator divergence runs deepest in 2026.
Quick reference · the eight pillars.
/ 1.1
Informed consent.
Voluntary, current, documented. ICH E6(R3) Annex 1 §3.3; 21 CFR 50 Subpart B; EU CTR Article 28-31. Re-consent on material amendment. eConsent recognised by FDA and EMA.
/ 1.2
IRB / IEC oversight.
Independent ethics review before enrolment. 21 CFR 56; EU CTR Part II Member State; J-GCP IRB; CDSCO Form CT-04 ethics approval. Composition, quorum, decision documentation.
/ 1.3
Investigator qualification.
CV, training, GCP, delegation log, 1572 (US). 21 CFR 312.53; ICH E6(R3) Annex 1 §2; PMDA investigator dossier. Qualifications must match the protocol's medical demands.
/ 1.4
Monitoring · RBM.
Risk-based monitoring · centralised + on-site. ICH E6(R3) Annex 1 §3.10; FDA RBM Guidance Aug 2013/2023. Critical-to-quality data drive monitoring intensity, not blanket 100% SDV.
/ 1.5
Data integrity. ★
ALCOA+ across paper and electronic. 21 CFR Part 11 + EU Annex 11; MHRA Data Integrity Guidance Mar 2018; PIC/S PI 041-1. Where the regulator divergence runs deepest.
/ 1.6
Safety reporting.
SAE/SUSAR within 7/15 days. ICH E2A definitions, E2B(R3) electronic submission, E2D post-marketing. EudraVigilance, FAERS, JADER, VigiBase.
/ 1.7
TMF · Reference Model 3.3.
Trial Master File, paper or electronic. ICH E6(R3) Annex 1 §3.16; DIA TMF Reference Model 3.3 (2022); Inspection-ready throughout, not at lock.
/ 1.8
Source documentation.
Source data → CRF traceability. Certified copies, eSource, EHR-to-EDC. ICH E6(R3) Annex 1 §3.15; FDA eSource Guidance Sep 2013; EMA Reflection Paper EMA/INS/GCP/454280/2010.
Cross-regulator comparison · ICH · FDA · EMA · PMDA · CDSCO · MHRA · Health Canada.
ICH E6(R3)
FDA 21 CFR
EMA CTR 536/2014
PMDA J-GCP
CDSCO NDCT 2019
MHRA UK CTR
Health Canada Div 5
/ 1.1 Informed consent.Voluntary, current, documented. Re-consent on amendment. eConsent. +
At a glance · convergence vs divergence
CONVERGE
All seven regulators require written, dated, signed informed consent before any trial-related procedure. Declaration of Helsinki at apex. Re-consent on material amendment.
DIVERGE
EU CTR Article 28-31 mandates a participant-friendly summary plus the full ICF; CDSCO NDCT Schedule III requires audio-video recording for vulnerable populations and trials in illiterate participants; FDA 21 CFR 50.27 permits short-form consent + witness signature; EMA + FDA 2024 Q&A formally recognise eConsent; PMDA J-GCP §50 requires a separate "explanation document" distinct from the consent form.
ICH E6(R3)2025-01-06
Annex 1 §3.3 · principle ii
Voluntary informed consent, current at time of activity. Documented, dated, signed. Capacity-appropriate language. Re-consent on amendment material to participant interest.
Acceptance criteriaPre-trial signature · re-consent on amendment · capacity assessment
FDA21 CFR 50
Subpart B · short-form OK
21 CFR 50.20 · 50.25 elements (8 basic + 6 additional). 50.27 short-form + witness allowed. eConsent per Dec 2016 Q&A; updated 2024.
Acceptance criteria14 elements · witness for short-form · eConsent
EMA CTR536/2014
Article 28-31 · lay summary
CTR Article 28 (general), 29 (informed consent in clinical trials in emergency situations), 30 (incapacitated subjects), 31 (minors). Lay summary mandatory.
Acceptance criteriaFull ICF + lay summary · CTIS upload
PMDA J-GCPMHLW 28/1997
§50, 51 · explanation doc
Explanation document distinct from consent form. Reviewed by IRB. Reading-level appropriate. Re-consent triggers explicitly enumerated.
Acceptance criteriaTwo-document model: explanation + consent
CDSCO NDCT2019
Schedule III · audio-video
Rules 16-19 + Schedule III. Audio-video recording mandatory for vulnerable populations, illiterate participants. Form CT-01 acknowledgement.
Acceptance criteriaAV recording for vulnerable groups
MHRA UK CTRSI 2004/1031
Schedule 1 Part 2 · ICH-aligned
UK Clinical Trials Regulations Schedule 1 Part 2. ICH-aligned post-Brexit; HRA/NHS site-level approvals separate from MHRA CTA.
Acceptance criteriaICH-aligned · HRA approval distinct
Health CanadaDivision 5
FDR C.05 · ICH-aligned
Food and Drug Regulations Division 5 (Drugs for Clinical Trials). ICH-aligned consent. REB approval mandatory before participant enrolment.
Acceptance criteriaICH-aligned · REB approval
Inspector's eye
For India-bound trials, the AV recording rule under Schedule III is the most-flagged finding for non-Indian sponsors. For EU CTIS submissions, the lay-summary upload is the most-frequent reason a Part I assessment is paused. Both pre-2019 ICFs and pre-CTIS ICFs typically need a refresh.
/ 1.2 IRB / IEC oversight.Independent ethics review · composition · scope · quorum. +
At a glance · convergence vs divergence
CONVERGE
All require independent ethics review before participant enrolment. Composition: scientific + non-scientific + lay member at minimum. Decisions documented.
DIVERGE
EU CTR moves national ethics into Part II of the CTIS submission — one harmonised dossier, but each Member State runs national ethics in parallel. FDA 21 CFR 56 recognises central IRBs (WCG IRB, Advarra, Salus) for multi-site studies; this is uncommon in EU. CDSCO Form CT-04 is a separate Indian Ethics Committee registration step distinct from study-specific approval.
ICH E6(R3)2025-01-06
Annex 1 §1 · principle iii
Independent ethics review before any trial-related activity. Composition criteria, working procedures, decision documentation, ongoing review at intervals appropriate to risk.
Acceptance criteriaIndependent · documented · ongoing review
FDA21 CFR 56
Central IRB recognised
21 CFR 56 Subparts A-D. Central IRB (WCG, Advarra, Salus) standard for multi-site. Single IRB rule for NIH cooperative research.
Acceptance criteriaCentral IRB OK · FDA 1572 lists IRB
EMA CTR536/2014
CTIS Part II per MS
CTR Article 4-5. Single-portal CTIS submission; Part I scientific assessment (RMS-led); Part II national ethics in each MS. Ireland's NREC, France's CPP, etc.
Acceptance criteriaCTIS Part II · per Member State
PMDA J-GCPMHLW 28/1997
§27-34 · institution IRB
Institution-based IRB the historical norm; central IRB (Joint IRB / shared IRB) recognised since 2017 J-GCP revision. Annual continuation review required.
Acceptance criteriaInstitution IRB · annual continuation
CDSCO NDCT2019
Form CT-04 EC registration
Rules 7-15. Ethics Committee registration on Form CT-04 separate from study approval. Composition: 7-15 members; minimum 1 lay, 1 woman, 1 legal expert.
Acceptance criteriaEC registration distinct from CT approval
MHRA UK CTRSI 2004/1031
REC + HRA
Research Ethics Committee (REC) review via Health Research Authority (HRA). HRA Approval combines REC favourable opinion + NHS R&D capacity confirmation.
Acceptance criteriaREC favourable · HRA Approval
Health CanadaDivision 5
REB approval · TCPS2
FDR C.05.006(1)(c) requires REB approval. TCPS2 (Tri-Council Policy Statement) governs ethics-committee composition and procedures.
Acceptance criteriaREB · TCPS2 compliance
Inspector's eye
For multi-EU-MS trials, the most-flagged Part II finding is misalignment between Member-State ethics opinions on the same protocol; harmonised RMS Part I is no defence. For US-cooperative-group trials, single-IRB rule from NIH cooperative-research policy still surprises non-cooperative-group sponsors.
/ 1.3 Investigator qualification.CV, training, GCP, delegation, 1572 (US) / J-GCP dossier (Japan). +
At a glance · convergence vs divergence
CONVERGE
All require qualified PI by training, education, experience matched to the protocol's medical scope. CV, GCP training, delegation log; medical decisions by a qualified physician.
DIVERGE
FDA Form 1572 Statement of Investigator is the US-specific instrument; PMDA requires institution-level investigator dossier with annual update; CDSCO Schedule Y / NDCT 2019 Rule 17 mandates investigator registration with CDSCO before nomination on any CT application; EMA CTR embeds investigator CVs in CTIS Part I but has no equivalent of the 1572.
ICH E6(R3)2025-01-06
Annex 1 §2 · principle v
Investigator qualification: education, training, experience commensurate with protocol's clinical and procedural demands. CV, GCP training, delegation log per activity.
Acceptance criteriaCV · GCP · delegation log per activity
FDA21 CFR 312.53
Form 1572 mandatory
312.53(c) Form FDA 1572 Statement of Investigator. Lists IRB, sub-investigators, facilities. False statement = federal felony. Form updated when sub-investigators change.
Acceptance criteria1572 signed · updated on change
EMA CTR536/2014
CTIS Part I CV upload
CTR Article 49(2)(d). Investigator CV in CTIS Part I; no separate signed statement equivalent to 1572. Site-level investigator change reportable as substantial modification.
Acceptance criteriaCV in CTIS · change = substantial modif.
PMDA J-GCPMHLW 28/1997
Institution dossier annual
§42-43. Institution-level investigator qualification dossier; annual update. Sub-investigator delegation documented at institution level.
Acceptance criteriaInstitution dossier · annual update
CDSCO NDCT2019
Rule 17 · pre-registration
Investigator registration with CDSCO mandatory before nomination on any CT application. Bilingual CV; Indian Council of Medical Research GCP training.
Acceptance criteriaPre-registration with CDSCO
MHRA UK CTRSI 2004/1031
Schedule 1 Part 4
Investigator brochure + CV in CTA. ICH-aligned. NHS R&D capacity assessment for site-level qualification.
Acceptance criteriaCV + IB · NHS R&D capacity
Health CanadaDivision 5
FDR C.05.012 records
FDR C.05.012 records of qualifications. ICH-aligned. Qualified Investigator (QI) attestation maintained at site.
Acceptance criteriaQI attestation · ICH-aligned
/ 1.4 Monitoring · risk-based.RBM · centralised + on-site · critical-to-quality data drive intensity. +
At a glance · convergence vs divergence
CONVERGE
All major regulators have moved from 100% Source Data Verification (SDV) as the default to Risk-Based Monitoring (RBM) as the default. Critical-to-quality factors define monitoring intensity.
DIVERGE
FDA RBM Guidance (final Aug 2013) + FDA RBM Q&A guidance (final April 2023) formalised RBM ahead of ICH; EU CTR Article 48 (Monitoring; Article 48(2) is the legal basis for risk-based approach to monitoring; RMS assessment governed by Articles 5–8); PMDA historically slower to accept centralised-only monitoring — 2024 J-GCP guidance now aligned; CDSCO still expects on-site monitoring frequency proportionate to enrolment, not just risk.
ICH E6(R3)2025-01-06
Annex 1 §3.10 · principle vii
Sponsor monitoring plan based on critical-to-quality factors. Centralised + on-site + remote in any combination justified by risk. Trigger-based escalation.
Acceptance criteriaRisk-based monitoring plan · CtQ factors
FDA2013/2023
RBM Guidance · oversight
A Risk-Based Approach to Monitoring of Clinical Investigations Aug 2013, revised Mar 2023. Monitoring plan part of QMS. 21 CFR 312.50 sponsor general responsibilities.
Acceptance criteriaRBM plan · centralised data review
EMA CTR536/2014
Article 48 · Monitoring
Article 48(2) is the legal basis for risk-based monitoring. Sponsor monitoring proportionate to risk. Reflection paper on RBM 2013 still operational. RMS assessment governed by Articles 5–8.
Acceptance criteriaRisk-proportionate · documented plan
PMDA J-GCPMHLW 28/1997
2024 RBM alignment
§21-22 monitoring. 2024 PMDA guidance accepts centralised-only monitoring for low-risk endpoints, aligning with E6(R3).
Acceptance criteriaRBM accepted · 2024 alignment
CDSCO NDCT2019
On-site frequency expected
Rules 22-23. On-site monitoring expected at site initiation, interim, and close-out as a minimum — in addition to RBM. Triggered escalation.
Acceptance criteriaOn-site at SIV/IM/COV minimum
MHRA UK CTRSI 2004/1031
RBM-aligned · GCP IT
UK MHRA GCP Inspectorate published RBM expectations 2014; updated 2023. Aligned with E6(R3).
Acceptance criteriaRBM · centralised data review
Health CanadaDivision 5
FDR C.05.010(c)
Sponsor responsibility for trial conduct. ICH-aligned. RBM accepted; HPFB Inspectorate aligned to E6(R2)/(R3).
Acceptance criteriaRBM · ICH-aligned
Inspector's eye
For India-bound trials, dropping all on-site monitoring in favour of centralised + remote is the most-frequent CDSCO finding in 2026. SIV, interim, and COV physical visits are still expected. For EU CTR trials, the absence of a documented critical-to-quality factor list is the most-frequent Article 48 finding.
/ 1.5 Data integrity. ★ALCOA+ · 21 CFR Part 11 · EU Annex 11 · PIC/S PI 041-1 · deepest divergence. +
At a glance · convergence vs divergence · the deepest pillar
CONVERGE
All recognise ALCOA+ (Attributable, Legible, Contemporaneous, Original, Accurate · plus Complete, Consistent, Enduring, Available). Audit trails required for any computerised system holding GCP records.
DIVERGE
21 CFR Part 11 (US, since 1997) and EU Annex 11 (Volume 4 of EudraLex) take different approaches to electronic-signature validation; MHRA Data Integrity Guidance Mar 2018 is the most-cited inspectorate document worldwide; PIC/S PI 041-1 consolidates the inspectorate position. PMDA historically required printed-and-signed copies of electronic CRFs; this is being relaxed in 2024-2026 J-GCP revisions. CDSCO data-integrity expectations follow PIC/S; eCRF-only trials are still uncommon for Indian-only studies.
ICH E6(R3)2025-01-06
Principle viii · reliable results
Reliable trial results principle. ALCOA+ as data quality reference; cross-references E2A/E2B(R3), Q9(R1) quality risk management, M11 protocol structure.
Acceptance criteriaALCOA+ · audit trail integrity
FDA21 CFR Part 11
eRecords + eSignatures
21 CFR Part 11 (1997). eSignature = handwritten signature equivalent. Data Integrity Guidance Dec 2018. eSource Sep 2013.
Acceptance criteriaPart 11 · eSig validation
EMA CTREU Annex 11
EudraLex Vol 4 Annex 11
EU Annex 11 Computerised Systems. EMA Notice to Sponsors 2010 + Reflection Paper EMA/INS/GCP/454280/2010 on eSource and EHRs.
Acceptance criteriaAnnex 11 · CSV proportionate
PMDA J-GCPMHLW 28/1997
Print + sign legacy
Historically required printed eCRFs signed by investigator. 2024-2026 J-GCP revisions relax in line with E6(R3); paper still common at older institutions.
Acceptance criteriaPrint + sign legacy → relaxing
CDSCO NDCT2019
PIC/S-aligned
Schedule III data-integrity expectations follow PIC/S PI 041-1 in practice. eCRF acceptable; institutional ethics committees often still require paper printout.
Acceptance criteriaPIC/S aligned · paper still common
MHRA UK CTR2018 / 2023
★ The most-cited DI doc
MHRA GxP Data Integrity Guidance Mar 2018. The most-cited data-integrity guidance worldwide. Dynamic data, true copies, audit trail review at intervals.
Acceptance criteriaAudit-trail review at defined intervals
Health CanadaDivision 5
FDR C.05.012 records
FDR C.05.012 records retention 25 years. Aligned to ICH E6 + PIC/S PI 041-1 in HPFB inspections.
Acceptance criteria25-yr retention · PIC/S alignment
Inspector's eye
Audit-trail review at defined, documented intervals (rather than ad hoc) is the single most-flagged data-integrity finding in 2025-2026 EU and UK inspections. For Japan-bound trials, plan for paper-printout-and-signature workflows at older institutions even where the eCRF is fully validated. For US Part 11 systems, the eSignature meaning component is the most-frequent validation gap.
/ 1.6 Safety reporting.SAE / SUSAR · 7-day fatal/life-threatening · 15-day other · E2A/E2B(R3). +
At a glance · convergence vs divergence
CONVERGE
All adopt ICH E2A definitions (SAE, SAR, SUSAR, ADR). All use E2B(R3) electronic submission format. Sponsor expedited reporting timelines: 7 calendar days for fatal or life-threatening SUSARs, 15 calendar days for other SUSARs.
DIVERGE
EU CTR Article 41-43 centralises SUSAR submission via EudraVigilance Clinical Trial Module; FDA uses FAERS + IND Safety Reports per 21 CFR 312.32; PMDA uses JADER + PMDA-direct expedited reports; CDSCO Form CT-08 is India-specific; the Annual Safety Report (DSUR per E2F) is universally required but the date of submission varies.
ICH E6(R3)2025-01-06
Annex 1 §3.13 · cross-ref E2A/E2B/E2F
Cross-reference to ICH E2A definitions, E2B(R3) ICSR format, E2F DSUR. SUSARs to regulator and ethics; sponsor maintains overall safety profile.
Acceptance criteria7/15 day SUSAR · annual DSUR
FDA21 CFR 312.32
IND Safety Report
312.32 IND Safety Reports. Final Rule 2010 narrowed to suspected ADR by sponsor assessment. 7/15 day. FAERS post-marketing.
Acceptance criteriaIND Safety Report 7/15 day
EMA CTR536/2014
Article 41-43 · EudraVigilance
CTR Article 41-43. EudraVigilance Clinical Trial Module (EVCTM) the single submission point for SUSARs in EU.
Acceptance criteriaEudraVigilance CTM · 7/15
PMDA J-GCPPAL Article 80-2
JADER + direct
Pharmaceuticals and Medical Devices Act §80-2. JADER post-marketing; PMDA direct expedited reports for clinical trials. 7/15 day.
Acceptance criteriaPMDA direct · JADER
CDSCO NDCT2019
Form CT-08 SUSAR
Rules 27-30. Form CT-08 expedited safety report. SUSARs to CDSCO and Ethics Committee. Compulsory compensation for CT-related injury (Rule 39).
Acceptance criteriaForm CT-08 · 7/15 day
MHRA UK CTRSI 2004/1031
SUSAR via MHRA portal
Schedule 1 Part 5. Post-Brexit MHRA portal for SUSARs (no longer EudraVigilance). 7/15 day. Annual Safety Report (DSUR) due 60 days after development international birth date.
Acceptance criteriaMHRA portal · DSUR DIBD+60
Health CanadaDivision 5
FDR C.05.014
FDR C.05.014 reporting of serious unexpected ADRs. 7/15 day. Canada Vigilance database; ICH E2A/E2B aligned.
Acceptance criteria7/15 day · Canada Vigilance
/ 1.7 TMF · Reference Model 3.3.Trial Master File · paper or electronic · inspection-ready throughout. +
At a glance · convergence vs divergence
CONVERGE
All require TMF · sponsor and investigator portions distinct · inspection-ready throughout study, not at study end. DIA TMF Reference Model 3.3 (2022) the de-facto industry standard structure.
DIVERGE
EMA Reflection Paper on TMF (2018) sets the most explicit timing expectations: documents in TMF within agreed timeframes proportionate to risk. FDA guidance silent on TMF as a structure — 21 CFR 312.62 / 812.140 records-and-reports requirement is what the TMF satisfies. PMDA has institution-level Investigator Site File (ISF) expectations that pre-date the DIA Reference Model. Retention varies: EU 25 years, US 2 years post-marketing, Japan 3-5 years post-end, India 5 years.
ICH E6(R3)2025-01-06
Annex 1 §3.16 · principle xi
TMF as quality system component. Sponsor TMF + Investigator Site File. Inspection-ready throughout. ICH E6(R3) deletes the prescriptive document list of R2 · principle-based.
Acceptance criteriaInspection-ready · principle-based
FDA21 CFR 312.62
Records-and-reports
312.62 sponsor + investigator records. No TMF as structure · the records-and-reports requirement is what TMF satisfies. 2-year retention post-marketing or post-IND withdrawal.
Acceptance criteria2-yr retention post-marketing
EMA CTR536/2014
Article 57-58 · 25-yr
CTR Article 57 sponsor TMF + Article 58 investigator TMF. 25 years archive. EMA Reflection Paper on TMF 2018.
Acceptance criteria25-yr archive · sponsor + invest TMF
PMDA J-GCPMHLW 28/1997
ISF institution-level
§36 essential records. Investigator Site File at institution level pre-dates DIA Reference Model. Retention 3 years post-end of study or 5 years post-approval, whichever later.
Acceptance criteria3-5 yr · ISF institution-level
CDSCO NDCT2019
Rule 24 · 5-yr
Rule 24 records retention 5 years post-completion or post-discontinuation. Investigator File + Sponsor File. Form CT-09 progress reports retained.
Acceptance criteria5-yr retention
MHRA UK CTRSI 2004/1031
25-yr post-Brexit
Schedule 1 Part 7. Aligned with EU CTR 25-year retention. MHRA GCP Inspectorate uses TMF Reference Model 3.3 in practice.
Acceptance criteria25-yr retention · TMF RM 3.3
Health CanadaDivision 5
FDR C.05.012 · 25-yr
FDR C.05.012 records retention 25 years. ICH E6 + DIA TMF RM 3.3 in practice.
Acceptance criteria25-yr retention · ICH-aligned
Inspector's eye
eTMF completeness at inspection — not retrospectively rebuilt on the eve of audit — is the single most-flagged finding across MHRA, FDA BIMO, and EU CTR inspections. The DIA TMF Reference Model 3.3 zone structure (Trial Management, Central Trial Documents, Regulatory, IRB/IEC, Site Management, Investigational Product, Safety Reporting, Statistics, Data Management, Third Parties, IP & CRO oversight) is what every modern eTMF tool ships with.
/ 1.8 Source documentation.Source data → CRF traceability · certified copies · eSource · EHR-to-EDC. +
At a glance · convergence vs divergence
CONVERGE
All require source documentation that is contemporaneous, attributable, traceable to the CRF entry. Certified-copy concept recognised. eSource (direct entry into EDC at point of care) accepted in principle.
DIVERGE
FDA eSource Sep 2013 is the foundational guidance — eCRF as source data permitted with adequate protocol justification; EMA Reflection Paper EMA/INS/GCP/454280/2010 + 2024 update covers EHR-to-EDC. PMDA requires explicit identification of source vs CRF in protocol; mismatched expectations remain frequent in non-Japanese sponsors. CDSCO ethics committees frequently require paper source even where eSource is protocol-defined.
ICH E6(R3)2025-01-06
Annex 1 §3.15 · ALCOA+
Source data identified in protocol. ALCOA+. Direct access to source for monitor, auditor, regulator. Certified copies recognised.
Acceptance criteriaSource identified in protocol · ALCOA+
FDA2013 / 2024
eSource Guidance
Electronic Source Data in Clinical Investigations Sep 2013. eCRF as source allowed with protocol justification. EHR-to-EDC consistent with this.
Acceptance criteriaeCRF-as-source · protocol justification
EMA CTR2010 / 2024
EMA/INS/GCP/454280/2010
Reflection paper on expectations for electronic source data and data transcribed to electronic data collection tools in clinical trials. Updated 2024 for EHR.
Acceptance criteriaEHR-to-EDC · EMA reflection paper
PMDA J-GCPMHLW 28/1997
Source list in protocol
§47 case report form. Explicit identification of source vs CRF in protocol. Mismatched expectations remain frequent for first-time-in-Japan sponsors.
Acceptance criteriaSource-vs-CRF list in protocol
CDSCO NDCT2019
EC paper-source ask
Rules 22-23 monitoring + Schedule III. Ethics Committees frequently require paper source even where eSource is protocol-defined.
Acceptance criteriaPaper source still common at EC level
MHRA UK CTR2018 / 2023
DI Guidance + GCP
MHRA Data Integrity Guidance Mar 2018 covers source data. eSource accepted; EHR-to-EDC accepted with risk assessment.
Acceptance criteriaeSource · risk assessment for EHR
Health CanadaDivision 5
ICH-aligned
FDR C.05.012 records. ICH-aligned. eSource accepted; HPFB inspections expect protocol-level identification.
Acceptance criteriaICH-aligned · protocol identification