NMPA/NHC Good Clinical Practice for Drugs (China GCP) — Audit-Readiness Rubric
What this is: the Chinese analogue to Regscribe's existing ICH E6 rubric (docs/ich-e6-rubric.md), for customers running or contracting drug/biologic clinical trials in mainland China, or submitting trial data generated in China to NMPA (China's National Medical Products Administration) in support of a domestic marketing application. China formally joined ICH in June 2017, and its 2020 GCP revision was explicitly built to align with ICH E6(R2) — but China layers several distinctive domestic requirements on top that have no ICH E6 analogue and carry real regulatory teeth: an Ethics Committee system with its own composition/independence rules, a clinical-trial-institution record-filing system (replacing a prior accreditation system), a mandatory public trial-registration/disclosure platform, and — the single most distinctive and highest-consequence China-specific control — Human Genetic Resources (HGR) administrative approval for any research activity touching genetic material collected in China. This document's job is to capture what China's own regulation requires and where it diverges from or adds to ICH E6 — not to restate ICH E6 with a translated cover page.
Standing project rule, restated because it governs every row below: never fabricate a citation — flag what could not be verified instead.
Sources used
Primary regulatory text — read directly this session:
- "Good Clinical Practice for Drugs" (药物临床试验质量管理规范), Announcement No. 57 of 2020, jointly issued by NMPA and the National Health Commission (NHC), released 26 April 2020, effective 1 July 2020. The authoritative text is Chinese; this rubric was built from a machine (Google Translate) English rendering, retrieved via the NIH/NIAID ClinRegs regulatory library at
clinregs.niaid.nih.gov/sites/default/files/documents/china/NMPA-GCP-No57-2020_Google-Translation.pdf(a 48-page, 83-article document; the PDF's own metadata shows a creation date of 2 November 2022, i.e. this is ClinRegs' archived translation of the 2020 text, not a fresh 2026 translation). This was read in full, page by page, this session — every chapter and article number cited below for the 2020 GCP was directly observed in that document, not inferred from secondary summaries. Caveat that must travel with every citation in this document: this is a machine translation of a legally-Chinese-language instrument, sourced via a third-party (US government/NIH) archive, not NMPA's own English text (NMPA does not appear to publish an official English translation of the full 83-article GCP itself — only English-language announcements about it, atenglish.nmpa.gov.cn). Article numbers and chapter boundaries are high-confidence (directly observed, internally consistent, article numbers run continuously 1→83 with no gaps found); exact clause wording/nuance should be treated as approximate, not quotable as official legal text. - The 2020 GCP's own confirmed structure (directly observed): 9 chapters, 83 articles — Chapter 1 General Provisions (Art. 1–10); Chapter 2 Terms and Definitions (Art. 11, containing 40 lettered sub-definitions); Chapter 3 Ethics Committee (Art. 12–15); Chapter 4 Researcher [Investigator] (Art. 16–28); Chapter 5 Sponsors (Art. 29–56); Chapter 6 Test Plan [Trial Protocol] (Art. 57–72); Chapter 7 Investigator's Handbook [Investigator's Brochure] (Art. 73–77); Chapter 8 Necessary Document Management [Essential Documents] (Art. 78–82); Chapter 9 Supplementary Provisions (Art. 83, stating the 1 July 2020 effective date). This contradicts some secondary sources found in search results (one described "8 chapters and 84 articles," a search-engine auto-summary described "six chapters and fifty-four articles" for what turned out to actually be describing the 2026 revision, not the 2020 text) — this rubric uses the structure directly observed in the primary-source document, not the secondary counts, precisely because of this kind of drift between summaries.
- For contrast/context, a Google-translated copy of the 2026 revision ("Good Clinical Practice for Drug Clinical Trials (Revised 2026)," Announcement No. 50 of 2026) was also fetched and partially read this session (first ~6 pages: Chapter 1 General Provisions Art. 1–13, Chapter 2 Ethics Review Committee Art. 14–17, start of Chapter 3 "Principal Investigators and Drug Clinical Trial Institutions" Art. 18–20) via the same ClinRegs archive (
NMPA-GCP-No50-2026_Google%20Translation.pdf) — not read in full, and not used as the basis for any row in Parts A–H below, since the 2026 text is not yet the operative standard as of the current date (see Part A). It is referenced only in the version-status discussion.
Version status / critical timing flag (do not miss this):
- Via direct WebFetch of
clinregs.niaid.nih.gov/country/china(a primary-source-aggregating NIH database, itself citing officialnmpa.gov.cnandgov.cnURLs): the 2020 GCP (Announcement No. 57 of 2020) is confirmed still in force through 31 August 2026, and is replaced in full by Announcement No. 50 of 2026 ("Good Clinical Practice for Drug Clinical Trials," issued 21 May 2026, published 25 June 2026), effective 1 September 2026 — i.e., roughly five weeks after the date this rubric was written. A separate, seemingly overlapping data point from the same ClinRegs page states that per a different notice (NMPA Announcement No. 125 of 2025), ICH E6(R3)-aligned technical requirements apply to "all Chinese drug clinical trials conducted after 31 March 2026" — this session could not reconcile whether that is the same transition as the September 2026 GCP replacement or a separate, earlier technical-guidance layer; flagged as unresolved rather than guessed at (see Gaps, item 1). - Practical consequence for Regscribe: this rubric is built on the 2020 GCP as the primary, currently-operative text (matching the task brief's framing), but given how close the 1 September 2026 replacement date is, this document has a materially shorter shelf life than Regscribe's other rubrics and should be flagged for near-term revision once the 2026 text can be read in full and reconciled against this one — see Gaps, item 1, and note the direct structural echo of how the ICH E6 rubric (
docs/ich-e6-rubric.md, Part A) handled the R2→R3 transition.
Secondary/tertiary sources — used only for content genuinely absent from the GCP text itself (Ethics Committee composition detail, institution filing, HGR, the trial registration platform, and enforcement history), each flagged individually:
- Ethics Committee composition (minimum 7 members, gender/outside-member/independence rules): these are not specified inside the GCP text itself (Art. 13(1) of the 2020 GCP defers composition/filing to "relevant regulations of the health authorities"). Sourced via WebFetch of
clinregs.niaid.nih.gov/country/china, which cites China's "Measures for Ethical Review of Life Science and Medical Research Involving Human Beings," issued jointly by NHC, the Ministry of Education, the Ministry of Science and Technology, and the National Administration of Traditional Chinese Medicine, published 18 February 2023 — confirmed independently via a second WebSearch pass citing the same instrument and the "at least 7 members" figure. This is a broader human-subjects-research ethics instrument, not GCP-specific — it governs ethics committees across all human life-science/medical research, of which drug trials are a subset. Treated as secondary (not independently fetched/read in full this session; content taken from search-engine synthesis of the primary text plus law-firm commentary, e.g. China Justice Observer, Biotechnology Law Report). - Clinical Trial Institution record-filing system: sourced via WebSearch confirming the "Measures for the Administration of Drug Clinical Trial Institutions," jointly issued by NMPA and NHC, released 29 November 2019, effective 1 December 2019 — confirmed by an
english.nmpa.gov.cnannouncement page (english.nmpa.gov.cn/2019-11/29/c_448732.htm) and cross-checked against a PMC (PubMed Central) article specifically on "China's drug clinical trial institution record-keeping system" (PMC10019591, referenced via search snippet, not independently fetched/read in full this session). The reform is confirmed as replacing a prior accreditation/qualification approval system with a record-filing (备案) system, per the Drug Administration Law's 2019 revision. Not independently verified this session: the exact filing platform name/URL, or the specific record-filing article/clause number within the Institution Measures themselves (the Measures document was not directly fetched). - Drug Clinical Trial Registration and Information Disclosure Platform: confirmed to exist and to require sponsor registration of trial initiation/progress information, via WebSearch citing
english.nmpa.gov.cnannouncement pages and Applied Clinical Trials Online commentary — but the specific NMPA notice number establishing the platform, and the precise legal trigger point (e.g., "before first subject enrolled" vs. "before drug regulatory review decision") were not confirmed against a primary-source document this session. Treated as confirmed-to-exist, unconfirmed-in-legal-detail — see Gaps, item 2. - Human Genetic Resources (HGR): the regulatory chain is: Biosecurity Law of the PRC (effective 15 April 2021, establishes HGR protection as a national security matter) → Regulation on the Administration of Human Genetic Resources (effective 1 July 2019) → Implementing Rules of the Regulation on the Administration of Human Genetic Resources (effective 1 July 2023). Confirmed via multiple independent sources (Freshfields, China Briefing, ChemLinked, Hogan Lovells, Covington) that substantively agree on structure and the July 2023 Implementing Rules' key liberalization: clinical trials conducted to obtain PRC marketing authorization for a drug/device are exempted from standard HGR collection-approval and the 3,000-subject population-study threshold, but HGR material/data leaving China, or foreign-party access to it, still requires either an International Scientific Research Cooperation approval or an International Cooperative Clinical Trial filing (via the
apply.hgrg.net"CHN-6" platform per one source, 20-working-day NHC review, no fee). Regulatory authority itself changed hands mid-cycle: HGR administration was transferred from the Ministry of Science and Technology (MOST) — which had run it via the China National Center for Biotechnology Development / Human Genetic Resources Administration of China office — to the National Health Commission (NHC), effective 1 May 2024, per the State Council's March 2023 Party-and-state-institution reform plan. One secondary source fetched this session (China Briefing) stated HGR filing "remains mandatory with MOST" without reflecting the May 2024 transfer — this is very likely stale/outdated content rather than a live contradiction, but it was not reconciled against a primary NHC/State Council document this session, so is flagged rather than silently corrected. - "722" clinical trial data-verification campaign (2015–2016): confirmed via WebSearch citing the PDA (Parenteral Drug Association) Letter, Lexology, and FiercePharma coverage (not independently fetched/read in full — content taken from search-result synthesis of these sources, which substantially agree on the figures). On 22 July 2015, China FDA (CFDA, NMPA's predecessor) ordered sponsors to self-inspect the data underlying 1,622 pending drug-registration applications; by the end of 2016, 1,165 were voluntarily withdrawn, 192 granted a "clinical trial waiver," and 30 refused registration approval — i.e., roughly 83% of applications under review were withdrawn or rejected. This predates the 2020 GCP (it is widely cited as a direct driver of China's 2017 ICH accession and the 2020 GCP rewrite) and is marked [HISTORICAL] below, used to illustrate why specific data-integrity clauses exist, not as a current enforcement trend.
- Current-era (2024–2026) named NMPA GCP inspection findings: none were found this session. Unlike EMA's GCP Inspectors' Working Group annual reports (used extensively in the ICH E6 rubric) or FDA's BIMO metrics, this session did not locate an NMPA-published aggregate GCP-inspection-finding dataset or any named recent enforcement action, despite several search attempts. This is the single biggest enforcement-evidence gap in this document — see Gaps, item 3.
Rubric legend
Each row: Requirement (plain language) → Citation → Evidence that demonstrates compliance → Common failure pattern
Failure-pattern cells are marked:
- [HISTORICAL] — a real, confirmed, publicly-documented regulatory episode, but predating or motivating the current rule, used to illustrate why a clause exists rather than as a current enforcement trend.
- [REPRESENTATIVE] — a well-established, textbook-level GCP failure mode consistent with the regulatory text and general auditor practice, not tied to a specific named 2020-GCP-era enforcement action verified this session.
- [ANALOGOUS] — a real, confirmed finding pattern, but drawn from a different jurisdiction's GCP inspection data (here: EMA's GCP IWG reports, already sourced in Regscribe's ICH E6 rubric) offered because no China-specific equivalent could be found — structurally on-point, not China-specific evidence.
- [UNVERIFIED — flagged, not used] — reserved marker; retained per the project's shared rubric convention. Not needed in this document's requirement rows since anything not independently confirmed was dropped or explicitly flagged in prose (see Sources/Gaps) rather than guessed at within a table cell.
Part A — Version status (read this before scoring anything)
| Milestone | Date | Source |
|---|---|---|
| 2003 GCP (predecessor) | superseded | Referenced in secondary sources as the prior version; not independently fetched this session |
| 2020 GCP (Announcement No. 57 of 2020) — issued | 26 April 2020 | english.nmpa.gov.cn/2020-04/26/c_500138.htm (direct fetch) |
| 2020 GCP — effective | 1 July 2020 | Primary source, Art. 83 (Chapter 9, Supplementary Provisions) — directly observed |
| China joins ICH | June 2017 | Widely corroborated across multiple secondary sources (Zhonglun law firm article, cisema.com); not independently verified against ICH's own member registry this session |
| Clinical Trial Institution record-filing system replaces accreditation | effective 1 December 2019 | english.nmpa.gov.cn/2019-11/29/c_448732.htm (direct fetch) |
| HGR Regulation (administrative) | effective 1 July 2019 | Multiple corroborating secondary sources |
| Biosecurity Law | effective 15 April 2021 | Multiple corroborating secondary sources |
| HGR Implementing Rules (liberalizes clinical-trial exemption) | effective 1 July 2023 | Multiple corroborating secondary sources |
| HGR administration transferred MOST → NHC | effective 1 May 2024 | Multiple corroborating secondary sources (China Briefing, Covington, Hogan Lovells) |
| 2026 GCP revision (Announcement No. 50 of 2026) — issued, replaces the 2020 GCP in full | issued 21 May 2026, published 25 June 2026 | clinregs.niaid.nih.gov/country/china (direct fetch, itself citing nmpa.gov.cn) |
| 2026 GCP — effective, 2020 GCP repealed | 1 September 2026 | Same source |
Bottom line for Regscribe's rubric scope: as of the date this rubric was written, the 2020 GCP (No. 57 of 2020) is the current, legally operative standard, and this document's Parts B–H below are built on it. It will be replaced in full in approximately five weeks (1 September 2026) by a text this session confirmed exists and partially read, but did not read in full or reconcile clause-by-clause against the 2020 text. Any customer document dated on or after 1 September 2026 should, strictly, be scored against the 2026 text once Regscribe builds that rubric — see Gaps, item 1.
Part B — Chapter 1: General Provisions (Art. 1–10)
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 1 | GCP governs the entire process of a drug clinical trial conducted for the purpose of applying for drug registration — design, organization, implementation, monitoring, auditing, recording, analysis, summarization, and reporting. All clinical-trial-related activities must comply. | 2020 GCP, Art. 1–2 | Trial-level SOP/quality-plan document explicitly invoking GCP compliance and NMPA drug-registration purpose | [REPRESENTATIVE] A document package for a China-site trial that never explicitly ties its conduct to a drug-registration purpose is a threshold scoping question — GCP applies to registration-directed trials; investigator-initiated/purely academic non-registration studies sit outside its direct scope (mirrors the ICH E6 rubric's own scoping caveat) |
| 2 | Trials must comply with the Declaration of Helsinki and related ethical requirements; subject rights and safety take precedence over scientific/social benefit; ethical review and informed consent are the named protective mechanisms. | 2020 GCP, Art. 3 | Ethics Committee approval record; signed informed consent forms | [REPRESENTATIVE] Same Helsinki-anchoring convention as ICH E6 — no CIOMS reference found in the 2020 GCP text itself, consistent with the ICH E6 rubric's own finding that CIOMS is absent from ICH E6 too |
| 3 | Trials must have sufficient scientific basis; may only proceed or continue when expected benefit reasonably outweighs risk to subjects and society. | 2020 GCP, Art. 4 | Protocol's risk/benefit justification section | [REPRESENTATIVE] |
| 4 | The trial protocol must be clear, detailed, and operable, and can only be executed after Ethics Committee consent — including for protocol amendments. | 2020 GCP, Art. 5 | Ethics-Committee-approved protocol version, with amendment approval history | [REPRESENTATIVE] A protocol amendment implemented at site before the corresponding EC approval letter is dated is a bright-line, easily-automatable gap check — structurally identical to the temporal-sequencing check the GLP rubric flags for Study-Director protocol sign-off (cfr58-glp-rubric.md, row 23) |
| 5 | Investigators must follow the trial protocol; medical judgments/decisions must be made by clinicians; personnel must have appropriate education/training/experience. | 2020 GCP, Art. 6 | Site delegation log with clinician credentials mapped to medical-decision responsibilities | [REPRESENTATIVE] |
| 6 | All paper and electronic trial data must be properly recorded, processed, and preserved so it can be accurately reported, interpreted, and confirmed; subject privacy/confidentiality must be protected. | 2020 GCP, Art. 7 | Data management plan; privacy/confidentiality SOP | [REPRESENTATIVE] |
| 7 | Investigational drug preparation must meet clinical-trial-drug production quality requirements; use must conform to the trial protocol. | 2020 GCP, Art. 8 | Certificate of analysis / GMP-for-clinical-trial-drugs documentation, cross-referenced to protocol dosing | [REPRESENTATIVE] Same substantive concern as ICH E6's Principle 11 (GMP-manufactured investigational product) |
| 8 | The clinical trial quality management system must cover the whole trial process, focused on subject protection, reliable results, and regulatory compliance. | 2020 GCP, Art. 9 | Sponsor-level QMS documentation | [REPRESENTATIVE] See Part D below (Sponsor QMS is elaborated in much greater detail in Chapter 5) |
| 9 | Trial conduct must comply with the principle of avoiding conflicts of interest. | 2020 GCP, Art. 10 | COI declaration/management records for investigators, EC members, and sponsor personnel | [REPRESENTATIVE] |
Part C — Chapter 3: Ethics Committee (Art. 12–15)
(Chapter 2, Art. 11, is a 40-item definitions section — used as reference throughout this document rather than as its own scoring row, mirroring how Regscribe's other rubrics treat definitions sections.)
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 10 | The Ethics Committee's duty is to protect subject rights/safety, with special attention to vulnerable subjects; it must review the trial protocol, ICF and updates, recruitment methods/materials, Investigator's Brochure, existing safety data, subject-compensation documentation, and investigator-qualification documentation. | 2020 GCP, Art. 12(1)–(4) | EC meeting minutes showing each required document category was actually reviewed, not just protocol/ICF | [REPRESENTATIVE] A "streamlined" EC review that only examines the protocol and ICF, skipping the investigator-qualification and existing-safety-data review items Art. 12 explicitly lists, is a documentation-completeness gap a document-analysis tool can catch directly |
| 11 | The EC must complete its review/filing within a reasonable timeframe and issue clear written opinions (agree; agree with modifications; disagree; terminate/suspend), stating required revisions or rejection reasons; it must require prompt investigator reporting of protocol deviations, all suspected/unexpected serious adverse reactions, and safety-relevant new information; and it has authority to suspend or terminate non-compliant trials or those causing unexpected serious subject harm. | 2020 GCP, Art. 12(9)–(12) | Written EC review-opinion letters bearing document name/version/date; EC correspondence log for safety-event notifications | [REPRESENTATIVE] |
| 12 | The EC must conduct periodic follow-up review of ongoing trials, at a frequency set by risk level but at least once a year. | 2020 GCP, Art. 12(13) | EC continuing-review schedule/log showing review dates no more than 12 months apart | [REPRESENTATIVE] A trial with a multi-year enrollment period but only a single initial EC approval on file, no annual continuing-review record, is a direct, checkable gap against this clause |
| 13 | EC composition and operation must meet health-authority requirements: all members trained in ethical review; work per the EC's own SOPs with written meeting records; voting members drawn from varied professional categories and different genders, meeting quorum; members independent of the clinical staff under review; the EC must maintain member qualification records. | 2020 GCP, Art. 13(1)–(6) | EC roster showing professional-category diversity, gender composition, and training-certificate records; independence-declaration records for voting members on each specific trial reviewed | [REPRESENTATIVE] — cross-reference: the 2020 GCP text itself defers the specific numeric/compositional threshold to external health-authority rules; the commonly-cited "at least 7 members" figure comes from a separate 2023 instrument (Measures for Ethical Review of Life Science and Medical Research Involving Human Beings), not from the GCP text itself — see Sources note above. A document-analysis tool should not hard-code "7" as a GCP citation; cite the 2023 Measures for that specific number. |
| 14 | The EC must establish and implement written SOPs covering: composition/organization/filing; meeting scheduling and notice; initial and follow-up review procedures; an expedited procedure for minor protocol amendments; timely notification of review opinions to investigators; and a procedure for handling disagreeing review opinions. | 2020 GCP, Art. 14(1)–(6) | EC's own SOP manual, checked section-by-section against these six required elements | [REPRESENTATIVE] An EC that has a general "SOP binder" but no distinct, documented expedited-review procedure for minor amendments is a common gap — the GCP text specifically calls this out as its own required SOP element, not something that can be folded into the general initial-review SOP |
| 15 | The EC must retain all ethics-review records (written review records, member information, submitted documents, meeting minutes, correspondence) for at least 5 years after trial completion; investigators/sponsors/drug regulatory authorities may request the EC's SOPs and reviewer list. | 2020 GCP, Art. 15 | EC archive with a demonstrable 5-year-minimum retention policy and a documented process for producing SOPs/reviewer lists on request | [REPRESENTATIVE] |
Part D — Chapter 4: Researcher [Investigator] (Art. 16–28)
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 16 | Investigators and the clinical trial institution must hold appropriate qualifications (practicing qualification at the institution, relevant professional knowledge/training/experience), stay current on the protocol/Investigator's Brochure/relevant drug information, and comply with GCP and applicable law; delegation of tasks to other individuals/entities requires the sponsor's advance consent, with complete oversight systems in place. | 2020 GCP, Art. 16–17 | Investigator CV/qualification file; delegation log with sponsor consent record for any non-institution delegate | [REPRESENTATIVE] |
| 17 | The investigator is responsible for subject medical care: must be (or be authorized as) a clinician bearing full responsibility for trial-related medical decisions; must ensure proper treatment and truthful disclosure for trial-related adverse events, including clinically significant lab abnormalities; must consider comorbidity treatments that could confound results or subject safety; subjects may withdraw without giving a reason. | 2020 GCP, Art. 18 | Site AE/SAE log with causality/co-morbidity annotation fields completed; withdrawal log with no coercive follow-up documentation required from the subject | [REPRESENTATIVE] |
| 18 | Investigators must not modify or deviate from the approved protocol without sponsor and EC consent, except to eliminate immediate hazard to subjects — in which case the deviation must be reported to the EC and sponsor (and, if necessary, the drug regulator) in a timely manner, with reasons explained. | 2020 GCP, Art. 20 | Protocol-deviation log distinguishing "immediate-hazard" deviations (reported after the fact, with justification) from ordinary amendments (pre-approved) | [REPRESENTATIVE] A deviation log with no distinction between "emergency, reported after" and "planned, approved before" deviation types — treating all deviations the same regardless of whether prior EC consent was obtained — obscures exactly the compliance question this clause is designed to test |
| 19 | Investigators/institutions are responsible for investigational drug management: qualified pharmacist/personnel assignment; receipt/storage/distribution/recovery/return/disposal records; storage-condition compliance; correct-use assurance and subject instruction; retained samples for bioequivalence trials kept at least 2 years after marketing. | 2020 GCP, Art. 21 | Drug-accountability log (dates, quantity, batch/lot, expiration, distribution code, signatures); retained-sample log with 2-year-post-marketing retention date calculated | [REPRESENTATIVE] |
| 20 | Investigators must follow randomization/blinding procedures; unblinding (accidental or emergency, tied to a serious adverse event) requires written explanation to the sponsor. | 2020 GCP, Art. 22 | Unblinding log with written sponsor notification for each unblinding event and its trigger reason | [REPRESENTATIVE] |
| 21 | Informed consent must follow Declaration of Helsinki ethical principles: latest EC-approved ICF version used; re-consent when new information could affect continued participation; no coercion/inducement; full disclosure including EC consent status; accessible language; signed/dated by subject and investigator, witnessed if the subject/guardian cannot read; specific provisions for incapacitated subjects, emergency consent, non-therapeutic trials, and pediatric assent. | 2020 GCP, Art. 23 (14 sub-items) | ICF version-control log tied to EC approval dates; re-consent records triggered by specific new-information events; documented pediatric-assent process where applicable | [REPRESENTATIVE] ICF not re-executed when materially new safety information emerges mid-study is the same cross-jurisdictional pattern flagged in the ICH E6 rubric (Part D, row 10) for drug trials generally — this is not a China-specific gap, but the GCP text's own Art. 23(2) makes it independently citable here |
| 22 | The ICF and other subject-facing materials must cover 20 specified content categories (trial overview, purpose, randomization possibility, procedures including traumatic ones, obligations, risks including reproductive/fetal risk, expected benefit including possibility of none, alternative treatments, damage compensation/treatment, compensation for participation, expected costs, voluntariness/withdrawal rights, confidentiality with named-inspector exceptions, new-information disclosure duty, contact information, termination circumstances, expected duration, estimated subject numbers). | 2020 GCP, Art. 24(1)–(20) | ICF template checked section-by-section against all 20 content categories | [REPRESENTATIVE] An ICF missing the reproductive/fetal-risk disclosure (item 8) or the "no benefit possible" disclosure (item 8, second clause) for a non-therapeutic or early-phase trial is a specific, checkable content gap — more granular than ICH E6's own consent-content list |
| 23 | Source data must meet ALCOA-type attributes: attributable, legible, contemporaneous, original, accurate, complete, consistent, durable; modifications must leave traceable marks without obscuring the original, with recorded reasons; electronic systems used for source data (e.g., electronic medical records for trials) require complete authority management and audit trails traceable to the record's creator/modifier. | 2020 GCP, Art. 25(2) | Source document/EMR audit-trail export showing no obscured corrections; system access-control/authority-management configuration for any electronic source-data system | [REPRESENTATIVE] Explicit ALCOA-type language appears directly in the GCP text itself (not just imported from ICH E6/FDA guidance by convention) — same substantive data-integrity check as the ICH E6 rubric's Part D row 11 and the GLP rubric's row 24, applied here to the China-specific electronic-medical-record integration context |
| 24 | Investigators must report safety events to the sponsor: serious adverse events reported immediately in writing with detailed follow-up; fatal events reported to both sponsor and EC with autopsy/final-medical-report information; SUSARs and safety updates from the sponsor must be reviewed/signed by the investigator and communicated to the EC. | 2020 GCP, Art. 26 | SAE log with immediate-notification timestamps and follow-up-report dates; investigator sign-off log for sponsor-provided SUSAR/safety-update packages | [REPRESENTATIVE] |
| 25 | Investigators must report trial progress: annual reports (or as EC requires) to the EC; immediate written reporting to sponsor/EC/institution for anything significantly affecting trial conduct or subject risk; a completion report to the institution, results summary to the EC, and required reports to the sponsor for drug-regulator submission. | 2020 GCP, Art. 28 | Annual-progress-report log; completion-report/results-summary distribution records | [REPRESENTATIVE] |
Part E — Chapter 5: Sponsors (Art. 29–56)
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 26 | Sponsors must establish a quality management system covering the whole trial process (design, implementation, recording, evaluation, result reporting, document filing); QA/QC methods must be proportionate to risk and trial-information importance; procedures/documents should be kept clear, concise, and operable — avoiding unnecessary complexity. | 2020 GCP, Art. 30 | Sponsor QMS documentation explicitly scoped to the whole trial lifecycle, not just data-collection | [REPRESENTATIVE] Directly parallels ICH E6(R3) Principle 6/7's "quality by design, proportionate to risk" language — the 2020 GCP's alignment with ICH E6(R2)-era risk-based QM is confirmed here in the primary text itself |
| 27 | Sponsors must conduct risk-based quality management: identify key trial links/data, assess risk at both system level (facilities, SOPs, computerized systems, personnel, suppliers) and trial level (investigational drug, design, data collection, informed consent), set risk tolerances, and periodically re-evaluate control measures based on new information; the trial report must describe the QM method used and any significant tolerance-limit breaches with remedial action. | 2020 GCP, Art. 31 | Written risk-management plan naming specific critical links/risk categories at both levels; final report's QM-methodology section cross-referenced against the plan | [REPRESENTATIVE] A risk plan produced once at trial start and never revisited — the same "frozen at initiation" pattern the ICH E6 rubric flags (Part C, row 1) for ICH E6(R3) Annex 1 §3.10.1 — applies identically here |
| 28 | Sponsor QA/QC: responsible for formulating/implementing/updating QA-QC SOPs; whole trial and lab-test process under QC to ensure reliable, correctly-processed data; contracts with all participating units (including investigators/institutions) must clarify responsibilities and confirm that sponsor and drug-regulator inspection can directly access source data/documents at the trial site. | 2020 GCP, Art. 32 | Sponsor QA-QC SOP set; site contracts with an explicit direct-access-to-source-data clause | [REPRESENTATIVE] |
| 29 | CRO delegation: sponsor may entrust part/all clinical trial work to a Contract Research Organization but retains ultimate responsibility for data quality/reliability; a designated responsible person must supervise CRO-performed work; CRO subcontracting requires sponsor's written approval; unclear entrustment of responsibilities defaults to the sponsor. | 2020 GCP, Art. 33 | CRO contract with explicit scope/SOP references; sponsor oversight/supervision records of CRO performance (not just the contract's existence) | [REPRESENTATIVE] A signed CRO contract treated as sufficient evidence of oversight, with no ongoing sponsor supervision trail, is the identical failure pattern the ICH E6 rubric documents for EMA's GCP IWG 2023 findings (Part C, row 3) — same underlying risk, independently confirmed here in China's own GCP text (Art. 33 explicitly retains sponsor responsibility) |
| 30 | Sponsors must maintain data/record governance: select qualified personnel for oversight/data processing/statistical analysis; may establish an independent data monitoring committee; the electronic data management system must undergo reliable system verification, maintain complete SOPs (setup, verification, functional testing, data collection/processing, system maintenance, change control, backup, recovery, contingency plans), keep audit trails traceable to record creator/modifier, and preserve blinding integrity throughout data entry/processing. | 2020 GCP, Art. 36 | Computerized-system validation records specific to the actual trial build; electronic audit-trail exports; independent data-monitoring-committee charter and meeting minutes (if used) | [REPRESENTATIVE] Validation records that reference only the EDC vendor's generic platform certification, not validation specific to this trial's configuration, is the same gap the ICH E6 rubric flags (Part C, row 6) for ICH E6(R3) Annex 1 §3.16/§4 — directly mirrored in the 2020 GCP's own Art. 36(3)–(6) language |
| 31 | Sponsors must select qualified investigators and institutions, provide the protocol/Investigator's Brochure with sufficient review time, and confirm biological-sample-testing labs meet qualification/quality requirements; genetic testing unrelated to the EC-agreed protocol is prohibited, and future/other use of remaining specimens requires separate subject informed consent. | 2020 GCP, Art. 37 | Investigator-selection rationale record; lab qualification documentation; specimen-consent records distinguishing protocol-scoped testing from any future/secondary use | [REPRESENTATIVE] Running unplanned genetic analyses on trial specimens without a separate consent/EC-scope check is a specific, named prohibition in the GCP text itself — worth a dedicated automated check, since it also intersects with the HGR requirements in Part G below |
| 32 | Before a clinical trial begins, the sponsor must submit trial materials to the drug regulatory department and obtain the trial license or complete the filing — submitted documents must show version number/date. | 2020 GCP, Art. 41 | NMPA clinical trial license/acceptance-of-filing document, version-and-date-matched to the protocol actually used | [REPRESENTATIVE] This is the GCP text's own reference point for the broader China drug-clinical-trial-application regulatory gate (governed in detail by NMPA's separate Provisions for Drug Registration, not fully explored in this session — see Gaps, item 2); a document package with no corresponding NMPA license/filing acknowledgment on file is a threshold, high-severity gap |
| 33 | Sponsors must clarify direct-access rights to trial records for inspectors/auditors/EC reviewers/drug-regulator inspectors, both to source data at trial sites and (with the subject's written agreement) to original medical records. | 2020 GCP, Art. 46 | Trial-site access agreements; subject written-agreement records for original-medical-record access by named categories of reviewer | [REPRESENTATIVE] Same substantive concern as ICH E6's monitor/source-record-access clauses (ICH E6 rubric, Part E, row 14) |
| 34 | Sponsor is responsible for safety assessment and adverse-drug-reaction reporting on required timelines; must analyze/evaluate safety information from any source, report SUSARs promptly to investigators/institutions/ECs and to the drug regulatory and health authorities; must provide periodic development safety update reports. | 2020 GCP, Art. 47–48 | SUSAR log with causality/expectedness assessment and multi-recipient notification timestamps; DSUR distribution records | [REPRESENTATIVE] |
| 35 | Sponsors must implement risk-based monitoring: document rationale for monitoring-strategy selection; combine on-site and centralized monitoring based on risk; produce a written monitoring plan emphasizing subject-rights protection, data authenticity, and key-risk handling; monitors must be trained, verify investigator qualifications/site readiness, confirm investigational-drug handling, confirm protocol/ICF compliance and case-report-form accuracy against source documents, and report findings in writing after each visit; sponsors must formulate inspection SOPs. | 2020 GCP, Art. 49–51 | Risk-based monitoring plan with documented strategy rationale; monitoring visit reports evidencing actual site-level findings review, not just a fixed visit-cadence checklist | [REPRESENTATIVE] A monitoring plan that states a risk-based/centralized approach on paper but produces visit reports with a fixed, unvarying cadence regardless of findings is the identical pattern the ICH E6 rubric documents (Part C, row 2) for ICH E6(R3) §3.11.4 — again independently confirmed in the 2020 GCP's own text |
| 36 | Sponsors may conduct audits: the auditor must be independent of the trial and not concurrently serve as a monitor; sponsor formulates audit procedures/plan based on submission content, subject numbers, trial complexity, and risk level; findings recorded in writing; audit certificate provided when needed. | 2020 GCP, Art. 52 | Auditor independence declaration (confirming no concurrent monitoring role for the same trial); audit plan and findings records | [REPRESENTATIVE] An "auditor" who also serves as the trial's monitor is a direct violation of Art. 52(2)'s independence requirement — a bright-line, checkable role-conflict gap |
| 37 | Sponsors must ensure trial compliance, conduct root-cause analysis and corrective/preventive action for significant compliance problems affecting subject safety/rights or data reliability, and may terminate an investigator/institution with serious, uncorrected non-compliance, reporting to the drug regulator. | 2020 GCP, Art. 53 | Non-compliance log with root-cause analysis and CAPA documentation; termination/reporting records for serious cases | [REPRESENTATIVE] Non-compliance logged without root-cause analysis — an audit trail of what happened without why — mirrors the identical gap flagged in the ICH E6 rubric (Part C, row 7) for ICH E6(R3) Annex 1 §3.12 |
| 38 | For multi-center trials, the sponsor must ensure all centers can comply with the trial program, provide the same protocol and case report form to every center, harmonize clinical/lab data evaluation, and ensure inter-investigator communication. | 2020 GCP, Art. 56 | Multi-center trial's centralized protocol/CRF distribution log; cross-site harmonization documentation | [REPRESENTATIVE] |
Part F — Chapters 6–8: Trial Protocol, Investigator's Brochure, Essential Documents (Art. 57–82)
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 39 | The trial protocol must specify (among many items across Art. 57–72): basic information (title, version, sponsor, signatories, medical expert, investigator, participating units); research background; trial purpose; design details including primary/secondary endpoints, control-group rationale, bias-control/randomization/blinding methods; treatment/dosage plans; suspension/termination criteria for subjects and the whole trial; investigational-drug management process; statistical plan (sample size rationale, significance level, handling of missing/illogical data, analysis population definitions); and quality control/assurance implementation. | 2020 GCP, Art. 57–72 | Protocol document checked section-by-section against the Art. 57–72 content list | [REPRESENTATIVE] — this is a rich, itemized China-specific protocol-content checklist, more granular in places than ICH E6(R3)'s own Appendix B; worth building as its own structured document-analysis checklist analogous to how the GLP rubric treats 21 CFR 58.120's protocol-content list (cfr58-glp-rubric.md, row 23) |
| 40 | The Investigator's Brochure must compile pharmaceutical, non-clinical, and clinical data on the investigational drug (chemistry, pharmacology, toxicology, prior human experience); may be simplified for a marketed drug the investigator already understands, substituting drug inserts/labeling; sponsor must formulate written revision procedures and review/update it at least annually, informing the investigator before updating and coordinating with the EC/drug regulator as needed. | 2020 GCP, Art. 73–77 | IB version-control log showing annual review cadence; investigator/EC notification records tied to each IB update | [REPRESENTATIVE] An IB not reviewed/updated within the required annual cycle, or updated without investigator/EC notification, is a direct, checkable gap |
| 41 | Essential/necessary documents ("documents for evaluating clinical trial implementation and data quality") must be kept in conditions preventing light exposure/damage/loss, with SOPs for identification/retrieval/relocation, and with electronic media integrity periodically tested; retention is at least 5 years after drug marketing approval (for registration-directed trials) or at least 5 years after trial termination (otherwise); sponsors must ensure investigators retain independent access to their own case-report data, not solely sponsor-controlled; all three parties (investigator, institution, sponsor) must establish document archives at trial start, reviewed/confirmed by the monitor at trial end. | 2020 GCP, Art. 78–82 | Document archive with demonstrable retention-period calculation tied to actual marketing-approval or termination date; investigator's own retained copy of submitted case-report data, independent of the sponsor's copy | [REPRESENTATIVE] Retention keyed to a future, trial-external event (marketing approval) rather than a fixed post-trial clock is structurally the same distinctive risk the GLP rubric flags for 21 CFR 58.195 (cfr58-glp-rubric.md, row 27) — a retention schedule calculated once at trial completion and never revisited when marketing approval is later granted (or denied, or delayed for years) risks premature destruction under this China-specific rule too |
Part G — China-distinctive requirements layered outside the GCP text itself
These are the requirements the task brief specifically flagged as most likely to matter, and — confirmed by the primary-source read above — none of them are actually inside the 83 articles of the 2020 GCP text. They are separate NMPA/NHC/State Council instruments that a China-site trial must also satisfy. Treating them as GCP clauses would be a fabrication; they are documented here as a distinct, cross-referenced layer.
| # | Requirement | Citation | Evidence | Common failure pattern |
|---|---|---|---|---|
| 42 | Clinical Trial Institution record-filing: since 1 December 2019, institutions conducting drug clinical trials are subject to a record-filing (备案) system, replacing the prior accreditation/qualification-approval system, per the Drug Administration Law's 2019 revision. | "Measures for the Administration of Drug Clinical Trial Institutions" (NMPA/NHC, effective 1 Dec 2019) — secondary-sourced, not independently fetched/read this session | Institution's record-filing confirmation/number, not an old-style accreditation certificate | [REPRESENTATIVE] A site document package that still cites a pre-2019-style institutional "accreditation" as its qualification basis, with no record-filing number, is now using outdated terminology for the applicable China regime — worth a dedicated automated flag distinct from generic institution-qualification checks used in Regscribe's other rubrics |
| 43 | Drug Clinical Trial Registration and Information Disclosure Platform: sponsors must register trial initiation/progress information and disclose investigator/site details publicly. | Referenced across multiple NMPA announcement pages and industry commentary — exact governing notice number and precise trigger-point (pre-enrollment vs. pre-regulatory-decision) not confirmed against a primary document this session | Platform registration record/screenshot cross-referenced against the actual trial's enrollment timeline | [UNVERIFIED — flagged, not used as a scored citation until the precise instrument/trigger-point is confirmed] — do not build an automated pass/fail gate on an unconfirmed trigger date; flag for follow-up (Gaps, item 2) |
| 44 | Human Genetic Resources (HGR) approval — the single highest-teeth China-specific control: any collection, preservation, or international-cooperation use of Chinese human genetic resources requires either exemption (clinical trials for PRC marketing authorization are exempted from standard collection-approval and the 3,000-subject threshold, per the July 2023 Implementing Rules) or a separate approval/filing route — an International Scientific Research Cooperation license (for foreign-party access/HGR export, NHC-administered since 1 May 2024, ~20-working-day review, no fee) or an International Cooperative Clinical Trial filing (for marketing-authorization-directed trials not exporting HGR materials). Foreign parties are barred from independently collecting/preserving Chinese HGR or selling/purchasing HGR materials; all HGR-derived records/data generated in China must be made fully available to the Chinese collaborating unit, with a backup retained by that unit. HGR information (genes/genome data/HGR-material-derived data) is defined narrowly and excludes clinical, imaging, protein, and metabolic data under the 2023 Implementing Rules. | Biosecurity Law (eff. 15 Apr 2021); Regulation on the Administration of Human Genetic Resources (eff. 1 Jul 2019); its Implementing Rules (eff. 1 Jul 2023) — all secondary-sourced this session (Freshfields, China Briefing, ChemLinked, Hogan Lovells, Covington), cross-corroborated across multiple independent law-firm summaries but not independently fetched/read from a primary NHC/State Council text | HGR exemption determination on file (citing the specific clinical-trial exemption) OR the relevant approval/filing certificate (International Cooperation license or International Cooperative Clinical Trial filing number); data-sharing/backup-to-Chinese-unit confirmation for any foreign-sponsored trial | [REPRESENTATIVE] A foreign-sponsored multi-region trial with a China arm that assumes ICH E6/FDA/EMA-style genetic-substudy consent and export practices are sufficient, without a distinct HGR determination on file, is the highest-consequence gap this rubric can flag — HGR non-compliance in China carries real regulatory/criminal exposure (per multiple sourced commentaries) that has no equivalent severity in the ICH E6 rubric's own scope. This is the row Regscribe should weight most heavily among the China-specific additions, pending better primary-source confirmation (see Gaps, item 4). |
| 45 | Ethics Committee composition minimums (at least 7 members, multi-disciplinary, gender-diverse, at least one non-institution member) — sits alongside, not inside, the 2020 GCP's own Art. 13. | "Measures for Ethical Review of Life Science and Medical Research Involving Human Beings" (NHC/MOE/MOST/NATCM, eff. 18 Feb 2023) — secondary-sourced | EC roster demonstrating the 7-member/diversity/independence composition, cited to the 2023 Measures rather than to the GCP text | [REPRESENTATIVE] Citing "GCP Art. 13" as the source for the "7 members" number (an error this session specifically avoided making) would be exactly the kind of plausible-but-wrong citation the project's no-fabrication rule exists to prevent — flag if found in any generated output |
Part H — Cross-reference: China GCP vs. ICH E6
| China GCP concept | ICH E6 analogue | Divergence/addition worth flagging |
|---|---|---|
| Chapter 3 Ethics Committee (Art. 12–15) | ICH E6(R3) Principle 3; Annex 1 references to IRB/IEC throughout | China's GCP text itself is comparatively light on EC composition specifics, deferring to external health-authority rules (2023 Measures) — ICH E6 similarly doesn't prescribe EC membership counts, so this is a shared gap, but China fills it via a distinct, separately-dated national instrument worth tracking independently |
| Chapter 4 Researcher / Chapter 5 Sponsors risk-based QM (Art. 31) | ICH E6(R3) Annex 1 §3.10.1 (Sponsor risk management) | Structurally near-identical — both require system-level and study/trial-level risk identification, tolerance-setting, and periodic re-evaluation; the 2020 GCP's Art. 31 predates ICH E6(R3)'s January 2025 finalization by nearly five years, consistent with China's stated 2020 alignment target being ICH E6(R2), not R3 |
| Clinical Trial Institution record-filing (Part G, row 42) | No ICH E6 analogue | Purely a China domestic administrative-reform artifact (2019 Drug Administration Law revision replacing institutional accreditation with filing) |
| Drug Clinical Trial Registration and Information Disclosure Platform (Part G, row 43) | ICH E6(R3) Principle 9 (transparency: registration on public databases, results posting) — same concept, different mechanism | ICH E6 speaks to transparency in principle; China operationalizes it via a specific named national platform with its own registration mechanics — worth noting as "same goal, different implementation," not simply "China copied ICH" |
| Human Genetic Resources approval (Part G, row 44) | No ICH E6 analogue whatsoever | The single cleanest example of a requirement genuinely outside ICH E6's scope — ICH E6 is silent on genetic-material sovereignty/export control because it is not a national-security-flavored instrument the way China's Biosecurity Law is |
| SUSAR/DSUR reporting (Art. 26, 47–48) | ICH E6(R3) Annex 1 §3.13; ICH E2A/E2F (cross-referenced by name, not detailed, in the ICH E6 rubric's own Gaps item 9) | Substantively aligned; the 2020 GCP explicitly names sponsor SUSAR-reporting duties to both drug regulatory and health authorities — a dual-agency reporting line (NMPA + NHC) not found in ICH E6 itself |
| ALCOA-type source-data integrity (Art. 25(2)) | ICH E6(R3) Annex 1 §2.12.2 (explicitly names ALCOA) | Near-verbatim conceptual match — both instruments name the same data-integrity attribute set, consistent with China's stated 2020 ICH E6(R2)-alignment goal |
Gaps / follow-ups before this goes into production scoring
- The 1 September 2026 GCP replacement (Announcement No. 50 of 2026) makes this rubric's shelf life unusually short. This session confirmed the replacement exists and read roughly the first six pages of a machine-translated copy (Chapter 1 General Provisions, start of Chapter 2 Ethics Review Committee, start of Chapter 3), but did not read it in full or reconcile it clause-by-clause against the 2020 text scored above. There is also an unresolved question about whether a separate NMPA Announcement No. 125 of 2025, said to apply ICH E6(R3)-aligned requirements to "trials conducted after 31 March 2026," is the same transition as the 1 September 2026 GCP replacement or an earlier, distinct technical-guidance layer — this was not resolved this session. Recommend treating this as the single highest-priority follow-up, on a similar timeline urgency to how the ICH E6 rubric flags its own E6(R3) Annex 2 as "not yet operative, needs a dedicated future pass" (
ich-e6-rubric.md, Gaps item 1) — except here the transition is imminent rather than distant. - The Drug Clinical Trial Registration and Information Disclosure Platform's exact governing instrument and legal trigger point were not confirmed against a primary source. Confirmed to exist and that sponsors must register trial initiation/progress information, but the precise notice number and whether registration must precede first-subject-enrollment or some other milestone was not verified. Flagged in Part G, row 43 as [UNVERIFIED — flagged, not used as a scored citation] rather than guessed at.
- No current-era (2024–2026) named NMPA GCP inspection findings or aggregate finding-frequency dataset were located this session, unlike the ICH E6 rubric's rich EMA GCP IWG sourcing. This is the single biggest evidentiary gap in this document relative to Regscribe's established rubric standard. Recommend a dedicated follow-up search pass targeting NMPA's own inspection-notice archive (if one is public) and Chinese-language pharma regulatory press, since English-language search coverage of NMPA enforcement specifics appears to be genuinely thin — a different sourcing problem than the GLP rubric's "search quota exhausted" gap, more like "the English-language secondary literature on this specific topic doesn't exist at volume."
- Every Human Genetic Resources (HGR) claim in Part G, row 44 is secondary-sourced (law-firm summaries, cross-corroborated across five independent firms but not independently read from a primary NHC/State Council/MOST text). Given that this session identified HGR as the highest-consequence, highest-regulatory-teeth China-specific requirement in this entire rubric, this is the single item most worth a dedicated primary-source follow-up — ideally a direct fetch/read of the Regulation on the Administration of Human Genetic Resources and its 2023 Implementing Rules themselves (Chinese-language, likely requiring translation), before Regscribe treats any HGR row as a hard pass/fail gate in production scoring.
- The specific article/clause number within the "Measures for the Administration of Drug Clinical Trial Institutions" that establishes the record-filing mechanism (Part G, row 42) was not confirmed — the existence and effective date of the record-filing reform is well-corroborated, but this rubric cites the Measures at the instrument level, not a specific article, unlike every 2020-GCP-sourced row above which cites a specific article number directly observed in the primary text.
- The "722" 2015–2016 data-verification campaign (Sources section, and referenced as [HISTORICAL] context) was sourced from search-result synthesis of the PDA Letter, Lexology, and FiercePharma coverage, not from a direct fetch/read of any one of those articles in full (the PDA Letter fetch attempt returned a 403). The specific figures (1,622 applications; 1,165 withdrawn; 192 waivers; 30 refused) are consistent across the search-engine's synthesis of multiple sources, which is somewhat reassuring, but no single source was read in full this session — flagged consistent with this project's standing sourcing-honesty rule.
- Chapter 2's 40-item definitions section (Art. 11) was read in full and used as background/reference throughout this document, but was not built out as its own scored table — consistent with how definitions sections are treated in Regscribe's other rubrics (e.g., the GLP rubric's Subpart A definitions row,
cfr58-glp-rubric.mdrow 2, is explicitly "N/A — reference section"). - Reconciliation with the ICH E6 rubric's own Gaps item 8 (SUSAR vs. Device Deficiency terminology reconciliation): this China GCP rubric adds a third vocabulary set to that reconciliation task (China's dual NMPA+NHC safety-reporting line, Part H above) — worth folding into that same future reconciliation pass rather than treating as a fourth standalone task.
- Traditional Chinese Medicine (TCM) and ethnic-medicine-specific Investigator's Brochure content (2020 GCP, Art. 77(17), referencing "theoretical basis, screening information, compatibility... ancient classic prescriptions") was observed directly in the primary text but not researched further — if Regscribe's customer base includes TCM-adjacent biotech/pharma companies, this is a distinctly China-specific IB-content requirement with no ICH E6 or FDA/EMA analogue at all, worth a dedicated follow-up if that customer segment materializes.