Important Protocol Deviations in 2026: One Departure, Four Outputs Under ICH E6(R3), China GCP, EU/UK Serious-Breach Rules and FDA's Draft

TL;DR. One departure in a multiregional trial now produces four outputs: the ICH E6(R3) section 3.9.3 importance grade; the China article 23 report to the ethics committee and the sponsor from 1 September 2026, with no day count; the EU or UK seven-day serious-breach decision under the EMA guideline and UK regulation 29A; and the FDA draft paths, still labeled Draft and "Not for implementation" on 28 September 2026. [1] [2] [3] [4] [5] [6] FDA's September 2025 final E6(R3) guidance repeats section 3.9.3's "may," and there "should" means suggested or recommended, but not required. [7] ICH E3 Q&As (R1) question 7, dated 6 July 2012, uses those "may" verbs; the December 2024 draft's quotation uses "might significantly affect" for both clauses. [8] [9]

Four outputs for one departure

Those four outputs are the sentences below. 21 CFR 312.66 sits beside the draft column as the in-force US sentence for IRB reporting of changes and unanticipated problems. 21 CFR 312.60 is the in-force US sentence for conducting the investigation according to the investigational plan. [10] [11]

The row later in this article is where those four outputs, and the 312.60 and 312.66 questions, are written for one departure.

The sentence to copy, and the quotation to leave in the draft

Copy section 3.9.3 into the protocol criterion, and leave the draft's "might significantly affect" quotation in the draft column. The two sentences are different documents. Sponsor classification practice is a separate question from that wording comparison. [1] [8] [9]

ICH E6(R3) section 3.9.3 reads:

The sponsor should determine necessary trial-specific criteria for classifying protocol deviations as important. Important protocol deviations are a subset of protocol deviations that may significantly impact the completeness, accuracy and/or reliability of the trial data or that may significantly affect a participant's rights, safety or well-being. [1]

FDA's E6(R3) guidance page displays September 2025 and the label Final. [12] The PDF repeats that operative sentence, including "may significantly impact," "may significantly affect," "trial data," and "participant." [7] The same PDF's preface says:

The use of the word should in Agency guidances means that something is suggested or recommended, but not required. [7]

The English Step 4 glossary has no protocol-deviation entry, so the definition used here sits in section 3.9.3. [1] [12]

ICH E3 Q&As (R1), question 7, current version dated 6 July 2012, answers what the clinical study report should treat as important:

Important protocol deviations are a subset of protocol deviations that may significantly impact the completeness, accuracy, and/or reliability of the study data or that may significantly affect a subject's rights, safety, or well-being. [8]

Question 7 calls a protocol deviation any change, divergence, or departure from the study design or procedures defined in the protocol. For a particular trial, that definition depends in part on study design, the critical procedures, the data, the subject protections, and the planned analyses. Sponsors may amend E3's examples, which include enrollment in violation of key eligibility criteria and failure to collect data needed for primary endpoints. Both clauses use "may." [8]

The December 2024 draft says it is adopting the ICH E3(R1) definitions and places "might significantly affect" in quotation marks for both the data clause and the rights clause. [9] The footnote points to FDA's January 2013 E3 questions-and-answers guidance. The comparison here is with the ICH file dated 6 July 2012. Later in the draft the same wording returns without quotation marks. [8] [9]

Section 3.9.3 supplies no universal example list. E3 names key eligibility and primary-endpoint data as illustrations the sponsor may amend. [8] Setting the classification for the trial, confirming it by someone other than the identifier, and doing that work before protocol finalisation are already in E3 question 7, section 3.9.3, and TransCelerate's March 2021 guide. [13] The 2026 increment is the status map: the verb in the final texts, the verb in the December 2024 quotation, the China sentence that changed on 1 September 2026, the US IRB "shall" in force today, and the December 2024 paths that remain in the draft. [6] [9]

What the investigator documents, and what section 3.12.2 notifies

Section 3.12.2 has two thresholds: root-cause analysis with corrective and preventive action, and a notification for what the text calls serious noncompliance. 21 CFR 312.66 is the in-force US sentence for IRB reporting of changes and unanticipated problems. The draft's wider IRB recommendation stays in the draft column. [1] [9] [10]

Section 2.5.3 says the investigator should document all protocol deviations, including those the sponsor communicates when they bear on that investigator's participants and conduct. For a deviation deemed important, the investigator should explain it and take measures to prevent a recurrence, where applicable. The cross-reference is section 3.9.3. [1] Sections 2.5.4 and 2.5.5 allow a departure only to eliminate an immediate hazard: inform the sponsor promptly, and report the hazard, the change, and any proposed amendment to the IRB or independent ethics committee and, where applicable, regulatory authorities. [1] Section 3.11.4.5.1(b) says monitoring should inform the investigator of relevant deviations, highlight the important ones, and focus remediation on them as appropriate. [1]

Section 3.12.2 then splits the sponsor's response:

If noncompliance that significantly affects or has the potential to significantly affect the rights, safety or well-being of trial participant(s) or the reliability of trial results is discovered, the sponsor should perform a root cause analysis, implement appropriate corrective and preventive actions and confirm their adequacy unless otherwise justified. Where the sponsor identifies issues that are likely to significantly impact the rights, safety or well-being of the trial participant(s) or the reliability of trial results (i.e., serious noncompliance), the sponsor should notify the regulatory authority and/or IRB/IEC, in accordance with applicable regulatory requirements, and/or investigator, as appropriate. [1]

FDA's final E6(R3) guidance repeats that two-tier structure, under the same preface that "should" means suggested or recommended, but not required. [7] The notification sentence names applicable regulatory requirements and states no seven-day clock of its own. [1]

If significant noncompliance by an investigator, institution, or service provider persists despite remediation, section 3.12.3 says the sponsor should consider terminating that participation and, in those circumstances, promptly notify the regulatory authority and IRB or independent ethics committee of the serious noncompliance, as appropriate, and take actions to minimise the impact on participants and on the reliability of the results. [1]

Appendix C says that if the listed trial records are produced, they are considered essential and should be retained. The row for records and reports of noncompliance, including protocol deviations and corrective and preventative actions, names no system. [1]

21 CFR 312.66, as displayed by the Legal Information Institute's e-CFR page on 28 September 2026, reads:

The investigator shall also assure that he or she will promptly report to the IRB all changes in the research activity and all unanticipated problems involving risk to human subjects or others, and that he or she will not make any changes in the research without IRB approval, except where necessary to eliminate apparent immediate hazards to human subjects. [10]

21 CFR 312.60, on the same host and date, says an investigator is responsible for ensuring that an investigation is conducted according to the signed investigator statement, the investigational plan, and applicable regulations, and for protecting the rights, safety, and welfare of subjects under the investigator's care. [11] Neither displayed section contains the term important protocol deviation. [10] [11]

The December 2024 draft cites 312.66, then adds a wider recommendation: "Where possible, FDA recommends that protocol deviations that are classified as important be submitted by the investigator to the IRB when they are identified, and in accordance with the IRB's written procedures." [9] One sponsor time frame in that draft is left as "within x days." Footnote 2 widens "protocol," inside this draft, to the monitoring plan and the statistical analysis plan. Footnote 6 cites ICH E6(R2), while the September 2025 final E6(R3) guidance remains the later final text. Lines 35 to 39 say "should" is suggested or recommended, but not required. [7] [9] [12]

21 CFR 812.150(a)(4) is a device rule: notice to the sponsor and the reviewing IRB of an emergency deviation to protect life or physical well-being, no later than 5 working days after the emergency, and the sponsor's prior approval for other changes or deviations. The page's next clause reads "FDA and IRB in accordance with § 812.35(a) also is required" and omits "approval." [14] The draft's further claim of prior FDA and IRB approval, where the change may affect scientific soundness or participants' rights, safety, or welfare, stays unconfirmed from that page. [9]

China, from 1 September 2026

The 2026 standard, republished by the Shanghai Municipal Medical Products Administration from NMPA announcement 2026 No. 50, dated 21 May 2026, takes effect on 1 September 2026 and abolishes announcement 2020 No. 57 on that date. [2] [3]

Article 23 opens with the principal investigator and authorised researchers complying with the protocol. Clause (1) says:

应当按照伦理审查委员会批准的试验方案实施临床试验。如偏离试验方案，应当记录，对重要方案偏离进行解释，采取适当的纠正和预防措施，并向伦理审查委员会和申办者报告。 [2]

A close rendering: the trial shall be conducted according to the protocol approved by the ethics committee. If the protocol is departed from, it shall be recorded; for an important protocol deviation, an explanation shall be given, appropriate corrective and preventive measures shall be taken, and a report shall be made to the ethics committee and the sponsor.

应当 appears once, before 记录. The rendering that attaches the explanation, the measures, and the report to an important protocol deviation follows the topic phrase 对重要方案偏离. The sentence states no day count. Article 54 defines five terms, and 重要方案偏离 lies outside that list. [2]

Clause (2) is a different departure, the one made without prior ethics-committee agreement to remove an urgent hazard:

为消除对试验参与者的紧急危害，在未获得伦理审查委员会同意的情况下，偏离试验方案，主要研究者应当及时向伦理审查委员会、申办者报告，并说明理由。 [2]

The principal investigator shall promptly report to the ethics committee and the sponsor and give the reasons. The 2020 phrase 必要时报告药品监督管理部门 is absent from this clause. [2]

Article 14(1) lists 重要方案偏离报告 among the ethics committee's review files. Article 14(5) calls for focused, timely attention to an urgent-hazard departure and to serious and continuing noncompliance. [2]

The 2020 standard, in the State Council gazette of announcement 2020 No. 57, already required a departure to be written down. Article 20(3) says:

研究者或者其指定的研究人员应当对偏离试验方案予以记录和解释。 [15]

A search of that gazette text found no occurrence of 重要方案偏离. The result describes the retrieved 2020 instrument. Article 20(4) adds the urgent-hazard report to the ethics committee and the sponsor and, when necessary, to the drug regulatory department. [15]

Article 43(2), in full for the verb:

发现对试验参与者的权益和安全，或者对试验结果的可靠性产生或者可能产生显著影响的不依从问题时，申办者应当及时分析根本原因，采取适当且充分的纠正和预防措施，及时书面报告伦理审查委员会。 [2]

Article 43(3):

发现严重、持续的不依从问题时，申办者应考虑终止主要研究者、临床试验机构或者服务供应商继续参加临床试验，及时书面报告伦理审查委员会，并采取措施将对试验参与者和试验结果的可靠性的影响降至最小。如违反试验方案或者本规范的问题严重时，申办者可追究相关人员的责任，并报告药品监督管理部门。 [2]

应当, 应考虑, and 可 stay on the clauses just quoted. Article 43(2)'s 应当 report names the ethics committee. Article 43(3)'s report to the drug regulatory department is 可. The 2020 article 53(3) shall-report to the drug regulatory department sits on a different sentence, tied to termination for serious or uncorrected noncompliance. Section 3.12.2's notification tier says "regulatory authority and/or IRB/IEC," "in accordance with applicable regulatory requirements." [1] [2] [15]

NMPA announcement 2025 No. 125, dated 22 December 2025 and read from the Shanghai republication, says drug clinical trials implemented after 31 March 2026 apply the ICH E6(R3) technical guideline. The Center for Drug Evaluation is named for technical guidance. [16] The 2026 questions and answers call the GCP a normative document and a basis for enforcement. Operational detail, recommended practice, and conceptual description in E6(R3) stay out of the GCP and are handled by full-text implementation. [17] That remains 适用 and full-text implementation, short of any claim that every ICH "should" is a free-standing Chinese penalty provision.

Article 54 says other terms may be taken from the Chinese E6(R3) glossary. That glossary file remains unretrieved, and the English Step 4 glossary has no protocol-deviation entry, so the grade used here remains section 3.9.3. [1] [2] The sentence that changed in 2026 is the important-deviation report. The 2020 standard already required a departure to be recorded and explained. [15]

The seven-day test is a different decision

The seven-day serious-breach decision uses its own sentences. UK regulation 29A, as read on 28 September 2026, uses "shall" and "likely to effect." [5]

The EMA guideline quotes Article 52 of Regulation (EU) No 536/2014, and the quotation matches the EUR-Lex consolidated text of 5 December 2022. In section 1 the guideline sets out:

"1. The sponsor shall notify the Member States concerned about a serious breach of this Regulation or of the version of the protocol applicable at the time of the breach through the EU portal without undue delay but not later than seven days of becoming aware of that breach.

2. For the purposes of this Article, a 'serious breach' means a breach likely to affect to a significant degree the safety and rights of a subject or the reliability and robustness of the data generated in the clinical trial." [4]

On protocol deviations, the guideline says an important deviation as defined in the ICH E3 questions and answers (R1) "is not equivalent to the definition of a serious breach and therefore an important deviation is not necessarily also a serious breach and vice versa," and that not every important deviation needs to be reported through the Clinical Trials Information System as a serious breach. [4] In the responsibilities list, reporting through that system should be done within 7 calendar days of the sponsor becoming aware. Where the sponsor has reasonable grounds based on evidence, the guideline expects the report first, within 7 days, with the investigation in parallel. [4]

UK regulation 29A, shown up to date on 28 September 2026, says:

The sponsor of a clinical trial shall notify the licensing authority in writing of any serious breach of (a) the conditions and principles of good clinical practice in connection with that trial; or (b) the protocol relating to that trial, as modified from time to time in accordance with regulations 20 to 22C, within 7 days of becoming aware of that breach. [5]

The statutory definition uses "effect": likely to effect to a significant degree the safety or physical or mental integrity of the participants, or the scientific value of the trial. [5] The EU quotation uses "affect," and its interests are the safety and rights of a subject, or the reliability and robustness of the data. [4] Both are seven-day sponsor notifications. Section 3.9.3 is a different sentence. [1]

Legislation notes attribute "modified," the regulations 20 to 22C cross-reference, and "participants" to SI 2025/538 on 28 April 2026. Regulation 29A was inserted in 2006. [5] MHRA's page, published that day, says the duty "is unchanged in the amended Clinical Trials Regulations and reflects the same requirements as the previous unamended Clinical Trials Regulations (the version prior to 28 April 2026)." [18] The seven-day duty dates from 2006.

The same page says protocol and GCP deviations occur commonly, that the majority are technical, and that not every deviation is reported to the MHRA as a serious breach. "Majority" is MHRA's word. Sponsors are told to keep these notifications clear of punitive use against sites or service providers. HRA procedures require notice to the research ethics committee. Not every serious adverse event or suspected unexpected serious adverse reaction is a serious breach. The clock starts when the party who holds the notification becomes aware. [18]

EMA's ICH E6 page puts the principles and Annex 1 into effect on 23 July 2025 and Annex 2 into effect on 15 January 2027. [19] A search of the Annex 2 Step 4 PDF found no occurrence of "protocol deviation," "serious breach," or "3.9.3." Section 3.9 of that annex is safety assessment. What Annex 2 does for a decentralised element sits outside that word search. [20]

What the docket asked, inside a document still marked draft

The reporting sentences ACRO asked FDA to replace on 28 February 2025 sit in the draft the FDA page, re-fetched on 28 September 2026, still marks not for implementation. [6] [21]

ACRO's comment on docket FDA-2023-D-5016 quotes draft lines 194 to 198, including the sentence that investigators should report to the sponsor all protocol deviations of which they are aware, using procedures that highlight the important ones. ACRO asks FDA to replace that passage with this one:

When the investigator identifies that an important deviation, as defined by the sponsor in the protocol or other study-specific instructions, has occurred, they should immediately report the deviation to the sponsor via a reporting process defined in the protocol or other study-specific instructions. The sponsor should then provide instruction on how to proceed, including decisions around a patient's ability to continue in the study or other steps to minimize risk to the patient or trial integrity. [21]

ACRO's requested sentence puts the sponsor's definition in control of the grade and routes the investigator's report through instructions the protocol already gives. The same letter says a small number of trials use EDC or another site-facing system, and that most sponsors and CROs track and manage protocol deviations in a CTMS that does not have a direct reporting interface for sites. "Most" is ACRO's word, the association's description of its members' practice. FDA's response remains unretrieved. [21]

In the same letter, ACRO's parenthetical of section 3.9.3 reads "i.e., those that impact the rights, safety and well-being of trial participants and the reliability of results." [21] The Step 4 sentence uses "may significantly" and includes the completeness, accuracy and/or reliability of the trial data. [1] The parenthetical records the comment. The sentence to copy remains section 3.9.3.

A later ACRO comment, dated 17 September 2025 and posted on 29 October 2025, asks in its subject line for targeted modifications to strengthen the draft's risk-based framework, to mitigate over-reporting, and to align the draft with EU and UK guidance. The attachment remains unread, so no line edit from that letter is used. [22] Other docket files, and FDA's response, remain unretrieved. That gap is separate from a verified absence of comments. [6]

A row for the departure on the desk

Write one row: the event, the section 3.9.3 criterion, the sponsor-confirmed grade kept apart from the investigator's account, the China report when article 23 applies, the serious-breach yes or no on the EU or UK wording, the 312.60 and 312.66 questions, and a blank draft-only cell unless the team has chosen to follow the December 2024 text. The row applies those sentences. Storage practice among sponsors is outside what the sentences show. [1] [2] [4] [5] [9] [10]

TransCelerate's March 2021 guide already says to store both classes in a validated repository and names CTMS, EDC, a trial master file, or a custom system as examples. Section 10.1 leaves an expedited authority report to the local requirement. The guide predates E6(R3) Step 4, the December 2024 draft, and the 2026 China standard. Appendix C names no system. [1] [13] ACRO's description of where its members' logs sit stays ACRO's description. [21]

EClinCloud's electronic data capture (EDC) product and its clinical trial management (CTMS) product are separate products. A team can keep the captured departure in one and the sponsor-confirmed grade, the China report status, and the serious-breach decision in the other, as distinct fields.

Two public write-ups, read against the row

Two public write-ups leave the important grade unassigned. The EMA GCP Inspectors Working Group 2024 report, adopted 15 December 2025, includes site examples of undetected deviations, late detection of investigational-product timing, and an eligibility assessment left incomplete before randomisation. [23] FDA's warning letter to Sourav K. Mishra, M.D., 29 April 2026, after an inspection on 17 to 20 March 2025, cites 21 CFR 312.60: Protocol 0270-22 prohibited CYP3A4 and CYP2D6 inhibitors, and all three enrolled subjects received aprepitant. The letter supplies no "important protocol deviation" sentence and concerns a site in India. [11] [24] On the row, the letter fills the event and the 312.60 cell. The section 3.9.3 grade, China article 23, and the seven-day tests stay on their own wording. An incomplete assessment before randomisation remains a different fact from E3's enrollment example. [8]

Frequently asked questions

Is the December 2024 protocol-deviations guidance in force on 28 September 2026?

The FDA guidance page re-fetched on that date displays December 2024, the label Draft, and "Not for implementation. Contains non-binding recommendations." [6] The E6(R3) guidance page on the same site displays September 2025 and the label Final, and the final PDF repeats section 3.9.3's "may significantly impact" and "may significantly affect." [7] [12]

Does section 3.12.2 require notice to a regulator for every important protocol deviation?

Section 3.12.2 uses two thresholds. Root-cause analysis and corrective and preventive action attach to noncompliance that significantly affects, or has the potential to significantly affect, participant rights, safety or well-being, or the reliability of trial results. Notification attaches to issues likely to significantly impact those interests. The text calls that tier serious noncompliance and says the notice follows applicable regulatory requirements. [1] In FDA's final E6 guidance, "should" means suggested or recommended, but not required. [7]

Does China article 23 set a seven-day report to the drug regulatory authority?

Article 23 names the ethics committee and the sponsor and states no day count. [2] The seven-day clocks retrieved here are the EMA guideline's quotation of Article 52 and UK regulation 29A. [4] [5] Article 43(3) uses 可 for a report to the drug regulatory department when a violation of the protocol or the standard is serious, again with no day count. [2]

Sources

1. International Council for Harmonisation. ICH E6(R3) Guideline for Good Clinical Practice, Step 4, 6 January 2025. Sections 2.5.3 to 2.5.5, 3.9.3, 3.11.4.5.1(b), 3.12.2, 3.12.3, and Appendix C. Accessed 28 September 2026. https://database.ich.org/sites/default/files/ICH_E6%28R3%29_Step4_FinalGuideline_2025_0106.pdf

2. National Medical Products Administration. 药物临床试验质量管理规范（2026年修订）, attachment 1 to announcement 2026 No. 50. Articles 14, 23, 43, and 54. Read from the Shanghai Municipal Medical Products Administration republication. Accessed 28 September 2026. https://yjj.sh.gov.cn/cmsres/01/019e5e2d868c470f9927c3a9672fc59c/203167cf5e796696423edf05cb8c3174.doc

3. Shanghai Municipal Medical Products Administration, republication of NMPA announcement 2026 No. 50, dated 21 May 2026. Effective 1 September 2026; announcement 2020 No. 57 abolished on that date. The page prints the NMPA address https://www.nmpa.gov.cn/xxgk/ggtg/ypggtg/ypqtggtg/20260608103856150.html. Accessed 28 September 2026. https://yjj.sh.gov.cn/qtgzwj/20260608/89cbf084b5cb4eab8fd0fb56d8538a1d.html

4. European Medicines Agency. Guideline for the notification of serious breaches of Regulation (EU) No 536/2014 or the clinical trial protocol, EMA/698382/2021. Section 1 quotation of Article 52, and the protocol-deviation discussion. Article 52 checked against the EUR-Lex consolidated text of 5 December 2022 on 28 September 2026. Accessed 28 September 2026. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-notification-serious-breaches-regulation-eu-no-5362014-or-clinical-trial-protocol_en.pdf

5. The Medicines for Human Use (Clinical Trials) Regulations 2004, regulation 29A, as shown up to date on 28 September 2026. https://www.legislation.gov.uk/uksi/2004/1031/regulation/29A

6. U.S. Food and Drug Administration. Protocol Deviations for Clinical Investigations of Drugs, Biological Products, and Devices. Guidance page, December 2024, label Draft, "Not for implementation. Contains non-binding recommendations." Docket FDA-2023-D-5016. Accessed 28 September 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/protocol-deviations-clinical-investigations-drugs-biological-products-and-devices

7. U.S. Food and Drug Administration. E6(R3) Good Clinical Practice: Guidance for Industry, September 2025. Preface and sections 3.9.3 and 3.12.2. Accessed 28 September 2026. https://www.fda.gov/media/169090/download

8. International Council for Harmonisation. E3 Q&As (R1), current version dated 6 July 2012. Question 7. Accessed 28 September 2026. https://database.ich.org/sites/default/files/E3_Q%26As_R1_Q%26As.pdf

9. U.S. Food and Drug Administration. Protocol Deviations for Clinical Investigations of Drugs, Biological Products, and Devices: Draft Guidance for Industry, December 2024. Header "Draft — Not for Implementation"; definitions; lines 194 to 216; lines 233 to 250; lines 351 to 356; footnotes 2 and 6. Accessed 28 September 2026. https://www.fda.gov/media/184745/download

10. 21 CFR 312.66, Assurance of IRB review, as displayed by the Legal Information Institute's e-CFR page on 28 September 2026. https://www.law.cornell.edu/cfr/text/21/312.66

11. 21 CFR 312.60, General responsibilities of investigators, as displayed by the Legal Information Institute's e-CFR page on 28 September 2026. https://www.law.cornell.edu/cfr/text/21/312.60

12. U.S. Food and Drug Administration. E6(R3) Good Clinical Practice (GCP), guidance page, September 2025, label Final. Accessed 28 September 2026. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/e6r3-good-clinical-practice-gcp

13. TransCelerate BioPharma. Protocol Deviation Process Guide, March 2021. Sections 3.5, 4.1, 7, 8, and 10.1. Accessed 28 September 2026. https://www.transceleratebiopharmainc.com/wp-content/uploads/2021/03/Protocol-Deviations_Process-Guide_March-2021.pdf

14. 21 CFR 812.150(a)(4), as displayed by the Legal Information Institute's e-CFR page on 28 September 2026. The closing clause on that page is grammatically incomplete. https://www.law.cornell.edu/cfr/text/21/812.150

15. State Council gazette. 药物临床试验质量管理规范, NMPA announcement 2020 No. 57. Articles 20 and 53. Accessed 28 September 2026. https://www.gov.cn/gongbao/content/2020/content_5525106.htm

16. Shanghai Municipal Medical Products Administration, republication of NMPA announcement 2025 No. 125, dated 22 December 2025. Drug clinical trials implemented after 31 March 2026 apply the E6(R3) technical guideline. The page prints the NMPA address https://www.nmpa.gov.cn/xxgk/ggtg/ypggtg/ypqtggtg/20251224111804135.html. Accessed 28 September 2026. https://yjj.sh.gov.cn/qtgzwj/20251225/0ad0d30639e24be2b53065903468e293.html

17. National Medical Products Administration. Questions and answers on the 2026 Drug Clinical Trial Quality Management Standard, attachment 2 to announcement 2026 No. 50. Answers 1 to 3. Read from the Shanghai Municipal Medical Products Administration republication. Accessed 28 September 2026. https://yjj.sh.gov.cn/cmsres/6c/6c38f69ae6a942e684221c90927db3c1/b6bd1d09af642f8f1107a80d3e0a3f59.doc

18. Medicines and Healthcare products Regulatory Agency. Notification of Serious Breaches of GCP or the trial protocol. Guidance published 28 April 2026. Accessed 28 September 2026. https://www.gov.uk/government/publications/clinical-trials-for-medicines-notification-of-serious-breaches-of-gcp-or-the-trial-protocol/notification-of-serious-breaches-of-gcp-or-the-trial-protocol

19. European Medicines Agency. ICH E6 Good Clinical Practice scientific guideline. Principles and Annex 1 effective 23 July 2025; Annex 2 effective 15 January 2027. Accessed 28 September 2026. https://www.ema.europa.eu/en/ich-e6-good-clinical-practice-scientific-guideline

20. International Council for Harmonisation. ICH E6(R3) Annex 2, Step 4, adopted 3 June 2026. Accessed 28 September 2026. https://database.ich.org/sites/default/files/ICH_E6%28R3%29_Annex%202_Guideline_Step%204_2026_0603_0.pdf

21. Association of Clinical Research Organizations. Comment on FDA-2023-D-5016, 28 February 2025. Accessed 28 September 2026. https://www.acrohealth.org/wp-content/uploads/2025/03/ACRO-FINAL-Protocol-Deviations-02-28-25.pdf

22. Regulations.gov. Comment FDA-2023-D-5016-0056, Association of Clinical Research Organizations, posted 29 October 2025. Subject line of the 17 September 2025 letter. The attachment remains unread. Accessed 28 September 2026. https://www.regulations.gov/comment/FDA-2023-D-5016-0056

23. European Medicines Agency. Annual Report of the Good Clinical Practice Inspectors Working Group 2024, EMA/INS/GCP/37608/2025, adopted 15 December 2025. Section 3 and investigator-site examples under protocol compliance. Accessed 28 September 2026. https://www.ema.europa.eu/en/documents/report/annual-report-good-clinical-practice-inspectors-working-group-2024_en.pdf

24. U.S. Food and Drug Administration. Warning letter to Sourav K. Mishra, M.D., All India Institute of Medical Sciences, 29 April 2026. MARCS-CMS 724881. Citation under 21 CFR 312.60, Protocol 0270-22. Accessed 28 September 2026. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/sourav-k-mishra-md-all-india-institute-medical-sciences-724881-04292026