QPPV.COM

Article 31 Non-Pharmacovigilance Referral in EU Medicines Regulation: Quality, Efficacy and Union-Level Regulatory Assessment

A detailed guide to the Article 31 non-pharmacovigilance referral pathway, with emphasis on quality and efficacy concerns, CHMP assessment, authorisation scope and the transition from scientific opinion to regulatory implementation.

Audio Lesson 24 min
Knowledge Assessment Test your understanding of this article. Take the assessment →

Article 31 Non-Pharmacovigilance Referral in EU Medicines Regulation

Overall structure

This article is organised around the actual regulatory lifecycle rather than a collection of short procedural definitions.

  1. What an Article 31 non-pharmacovigilance referral is
  2. The legal basis and the Union-interest trigger
  3. How the procedure differs from an Article 31 pharmacovigilance referral
  4. What types of quality and efficacy concerns can lead to referral
  5. Who can initiate the procedure and how the notification works
  6. Determining the products and applications within scope
  7. Urgent national protective action during the procedure
  8. CHMP organisation, rapporteurs and assessment questions
  9. Scientific assessment of quality and efficacy evidence
  10. Responses, clock-stops and the assessment timetable
  11. CHMP opinion and the distinction from the final legal decision
  12. Re-examination and procedural rights
  13. European Commission decision-making
  14. National implementation and centrally authorised products
  15. Impact on product information, quality systems and regulatory obligations
  16. Practical interpretation of referral documents
  17. Common regulatory misreadings
  18. Inspection readiness and internal governance
  19. Practical checklist for regulatory and pharmacovigilance professionals
  20. References and regulatory note

1. What Is an Article 31 Non-Pharmacovigilance Referral?

Article 31 of Directive 2001/83/EC provides a Union-level referral mechanism for certain regulatory questions concerning medicinal products where the interests of the Union are involved.

The non-pharmacovigilance pathway applies when the procedure is initiated as a result of evaluating data other than pharmacovigilance data. EMA identifies quality and efficacy concerns as important examples.

The basic regulatory sequence is:

Regulatory concern
       ↓
Union interest
       ↓
Article 31 referral
       ↓
CHMP assessment
       ↓
CHMP opinion
       ↓
Applicable legal decision
       ↓
Implementation

The defining point is the source and regulatory character of the evidence. This is not the Article 31 pathway used for pharmacovigilance data.

2. Why the Term “Non-Pharmacovigilance” Matters

The term distinguishes this procedure from the pharmacovigilance Article 31 pathway.

An Article 31 pharmacovigilance referral is initiated because of pharmacovigilance data and follows the PRAC-led assessment pathway. A non-pharmacovigilance Article 31 referral is initiated because of other data, such as data concerning quality or efficacy, and the scientific assessment is performed by CHMP.

The distinction is therefore not merely semantic. It determines the principal scientific committee and the procedural architecture that follows.

A useful first question when reading any Article 31 referral is therefore:

What type of data gave rise to the referral?

Only after answering that question should the reader infer whether PRAC or CHMP is the lead scientific committee.

The principal legal basis is Article 31 of Directive 2001/83/EC. The subsequent Union referral procedure is governed principally by Articles 32, 33 and 34 of the Directive, subject to the precise circumstances of the individual procedure.

The legal framework should be read as a sequence rather than as isolated provisions:

Article 31
Referral mechanism
       ↓
Article 32
CHMP procedure
       ↓
Article 33
Opinion and related procedural consequences
       ↓
Article 34
Commission decision-making framework

The exact legal consequences depend on the products within scope and the route through which the referral proceeds.

4. The “Interests of the Union” Trigger

The Union-interest requirement is central to Article 31.

It prevents the referral mechanism from becoming a general route for transferring every national scientific or regulatory disagreement to the European level.

EMA describes Union interest in the context of medicinal products with particular reference to public-health interests, quality or efficacy concerns and the free movement of medicinal products within the Union.

The practical regulatory question is therefore not simply:

“Is there a scientific concern?”

It is closer to:

“Is there a sufficiently significant Union-level regulatory concern that the statutory referral mechanism should be used?”

The answer must be established from the legal and procedural circumstances of the individual referral.

5. Quality Concerns as a Basis for Referral

A quality concern concerns characteristics of the medicinal product, its manufacture, controls or the evidence supporting its pharmaceutical quality.

Depending on the case, relevant issues may include:

The presence of a quality concern does not automatically mean that Article 31 is appropriate. The concern must be considered within the applicable Union-interest and referral framework.

This distinction matters because many quality issues are managed through ordinary regulatory and quality procedures without becoming Union referrals.

6. Efficacy Concerns as a Basis for Referral

An efficacy concern concerns whether the available evidence continues to support the therapeutic claims or conditions of use associated with a medicinal product.

Depending on the referral, relevant questions can include:

As with quality, the existence of uncertainty about efficacy does not by itself establish that Article 31 must be used. The legal trigger and Union significance must still be established.

7. Article 31 Is Not a General Scientific Consultation

An Article 31 referral is a statutory regulatory procedure, not simply a mechanism for obtaining an expert scientific opinion.

The procedure has defined legal triggers, formal notification, identified products or applications, a committee assessment and a regulatory endpoint.

This distinction is important when interpreting the public record. A committee document should not be read as though it were an informal scientific discussion detached from the legal procedure.

The referral exists because a defined regulatory question has been placed within the Union framework.

8. Who Can Initiate a Non-Pharmacovigilance Article 31 Referral?

The current EMA explanation identifies competent authorities in Member States, the European Commission, marketing authorisation holders and applicants as potential initiators within the applicable framework.

The roles should nevertheless be distinguished.

An MAH or applicant does not simply create a Union referral unilaterally. Where the company seeks an Article 31 referral, the matter must enter the statutory process through the competent regulatory route and the referral question must be established within that framework.

For regulatory reconstruction, identify the actual initiating authority from the formal referral documentation rather than relying on a secondary description.

9. The Formal Referral Notification

The referral notification is the procedural anchor for the assessment.

It should be used to establish:

The notification should be read before the later scientific assessment reports because it defines the question that CHMP was asked to answer.

A safety, quality or efficacy discussion in a later document may be broader than the formal legal question. The referral notification provides the boundary against which the later reasoning should be interpreted.

10. What the Referral Does Not Automatically Mean

The existence of an Article 31 non-pharmacovigilance referral does not by itself establish:

Those conclusions require evidence and, where applicable, a specific regulatory finding.

The referral means that the defined issue has entered a formal Union-level assessment process.

That distinction is essential when communicating about an ongoing referral.

11. Who Can Initiate an Article 31 Non-Pharmacovigilance Referral?

The Article 31 mechanism can be initiated by a competent authority of a Member State, the European Commission, a marketing authorisation holder or an applicant, subject to the procedural framework applicable to the referral.

The practical distinction is between requesting Union-level consideration and formally defining the regulatory question. An MAH or applicant may bring a concern to the attention of a competent authority or the Commission, but the statutory referral process and its formal scope are established through the applicable legal and regulatory procedure.

The initiating document should therefore be treated as the procedural anchor.

12. The Referral Notification

The notification identifies the issue that CHMP is being asked to examine.

A robust regulatory record should identify:

The notification is particularly important because later scientific documents should be interpreted against the questions actually referred.

A broad background concern should not automatically be treated as though every aspect of that concern became a formal CHMP question.

13. The Start of the Procedure

Once the referral is formally triggered, EMA publishes information about the procedure through its established publication channels.

The public record can include the referral announcement, procedure page, list of products, active substances, marketing authorisation holders or applicants, CHMP questions and timetable.

For regulatory surveillance, the formal procedural date should be distinguished from the date on which the underlying concern was first detected.

This distinction is essential when reconstructing a historical timeline.

14. Determining the Scope of Products

The scope of an Article 31 non-pharmacovigilance referral is determined by the regulatory concern and the formal referral documentation.

The affected population can include nationally authorised products and, where applicable, centrally authorised products and applications.

The scope may therefore extend beyond the product in which the issue was first identified.

For an MAH, the correct question is not:

"Which of our products do we think are relevant?"

It is:

"Which products and applications are formally within the scope of the referral?"

15. Nationally Authorised Products

Nationally authorised products can fall within an Article 31 non-pharmacovigilance referral.

This includes products authorised through national procedures and products authorised through mutual-recognition or decentralised procedures where the legal conditions for the referral are met.

The significance is practical: a concern that initially appears national can become a Union-level assessment with consequences for multiple national marketing authorisations.

The subsequent implementation pathway must then be followed for each affected authorisation.

16. Centrally Authorised Products and Article 31

Centrally authorised products require particular attention because Regulation (EC) No 726/2004 contains a separate referral mechanism for certain concerns involving only centrally authorised products.

Where the relevant legal conditions for the central-only mechanism are satisfied, the Article 20 procedure rather than Article 31 is the applicable route.

This means that the first procedural question should always be:

What products are actually within scope?

Only after the authorisation categories are established can the correct referral mechanism be determined confidently.

17. Article 31 Versus Article 20

The distinction can be simplified as follows:

Situation Relevant mechanism
Concern involving nationally authorised products within the Directive framework Article 31 of Directive 2001/83/EC
Certain concern involving only centrally authorised products Article 20 of Regulation (EC) No 726/2004
Pharmacovigilance Article 31 Article 31 pathway with PRAC assessment
Non-pharmacovigilance Article 31 Article 31 pathway with CHMP assessment

This is a procedural classification, not merely a difference in terminology.

A referral should therefore never be classified from the headline alone. The underlying legal basis and product scope should be checked.

18. Can an MAH Exclude Its Product?

Inclusion in the referral is determined by the applicable regulatory scope rather than by an individual MAH's preference.

An MAH can provide scientific and regulatory arguments concerning its product, but it cannot unilaterally redefine the Union referral scope.

If the company believes that a product has been incorrectly included or that the scientific basis does not apply to its product, that position should be raised through the formal procedural mechanisms available during the assessment.

19. Urgent National Protective Measures

The Union referral procedure does not necessarily prevent a Member State from taking urgent protective action when the applicable legal conditions are satisfied.

Where immediate action is necessary to protect public health, the national authority may have powers to suspend an authorisation or prohibit use pending the Union-level outcome.

Such action should be distinguished from the final Union decision.

The existence of an urgent national measure therefore does not mean that CHMP has already reached its final scientific conclusion.

20. Notification of Urgent Action

Urgent national action carries its own notification requirements.

The Member State should communicate the action through the applicable EU regulatory channels and provide the reasons supporting it.

For an MAH, the operational consequence is that an urgent national measure can require immediate action even while the broader Article 31 assessment remains underway.

The company's regulatory and pharmacovigilance teams should therefore have a process for identifying urgent measures separately from the ordinary referral timetable.

21. CHMP Is the Scientific Committee for This Pathway

For a non-pharmacovigilance Article 31 referral, CHMP performs the scientific assessment.

CHMP is responsible for evaluating the referred evidence and adopting the opinion required by the applicable legal framework.

This is the principal institutional distinction from a pharmacovigilance Article 31 referral, where PRAC conducts the pharmacovigilance assessment.

The simplified model is:

Non-pharmacovigilance data
          ↓
       Article 31
          ↓
         CHMP
          ↓
   Scientific opinion

The subsequent legal effect depends on the regulatory route and final decision-making process.

22. Why PRAC Is Not the Lead Committee

The absence of PRAC as the lead committee does not mean that pharmacovigilance information can never be considered.

The distinction concerns the basis on which the referral was initiated and the committee assigned the principal scientific assessment.

A non-pharmacovigilance referral can exist in a broader regulatory context where safety information is relevant, but that does not automatically convert it into a pharmacovigilance Article 31 procedure.

The legal classification must remain anchored to the formal referral basis.

23. Appointment of Rapporteurs

CHMP appoints rapporteurs to lead the detailed scientific assessment.

The appointment should reflect the expertise required by the specific referral.

For a quality-related concern, relevant expertise may be different from that required for a complex efficacy or clinical-data question.

The rapporteur system allows the detailed assessment to be developed before the matter returns to the full committee for discussion and adoption of the opinion.

24. The Rapporteur Assessment

The rapporteur reviews the evidence and prepares an assessment for CHMP consideration.

Depending on the referral, this can involve:

The assessment should address the questions established by the referral and subsequent CHMP requests.

25. The List of Questions

CHMP may issue a List of Questions (LoQ) requesting clarification or additional evidence from the MAHs or applicants.

The LoQ is an important procedural document because it shows what the committee considers unresolved after its initial assessment.

For an MAH, each question should be mapped to:

  1. the requested evidence;
  2. the responsible internal function;
  3. the response position;
  4. supporting documentation;
  5. any remaining uncertainty.

This creates traceability between the regulator's question and the company's response.

26. Clock-Stop and Response Period

A referral assessment can contain a clock-stop period during which the MAHs or applicants prepare responses to CHMP questions.

The exact timetable should always be taken from the procedure-specific timetable rather than assumed from a generic article.

A clock-stop is not simply administrative delay. It is part of the procedural architecture that allows the applicant or MAH to provide information needed for the scientific assessment.

27. Additional Questions

CHMP can raise additional questions when the responses do not resolve an issue sufficiently.

The existence of additional questions does not by itself indicate that the final outcome will be negative.

It indicates that the committee considers further clarification or evidence necessary before reaching its opinion.

The MAH should therefore treat each additional question as a defined scientific and regulatory workstream rather than as a general request for another submission.

28. Oral Explanation

Where the applicable procedure provides for it, the MAH or applicant can be given an opportunity to provide an oral explanation to CHMP.

An oral explanation should not be regarded as a replacement for the written evidence.

Its value is to allow the committee to question the applicant's position directly and to clarify important scientific or regulatory points.

The precise procedural opportunity and timing should be verified in the referral-specific documentation.

29. Expert Consultation

A complex referral can require expertise beyond the routine committee membership.

CHMP can seek advice from scientific experts where appropriate.

This is particularly relevant when the question involves highly specialised areas such as:

External expertise supports the committee's assessment but does not replace CHMP's responsibility for adopting the opinion.

30. Quality Referrals: Scientific Assessment

A quality-related Article 31 referral can require assessment of whether the medicinal product continues to meet the applicable quality requirements.

Depending on the concern, the assessment can consider:

The precise evidence required depends on the nature of the quality concern.

31. Quality Defect Versus Regulatory Quality Concern

A quality defect and a broader regulatory quality concern should not automatically be treated as identical.

A specific batch defect may be managed through established quality-defect procedures and urgent corrective actions.

A systemic concern about whether an authorised product continues to meet its quality requirements can require a broader regulatory assessment.

The relevant question is therefore not simply whether a defect exists, but what regulatory question the evidence raises and whether that question has Union significance.

32. Manufacturing Evidence

Manufacturing evidence can be central to a quality-related referral.

The assessment may need to establish whether the observed issue is:

This distinction can materially affect the regulatory response.

For example, an isolated issue with an adequately controlled root cause may lead to a different outcome from evidence suggesting a systemic inability to assure product quality.

33. Stability and Shelf-Life Questions

Where the concern relates to stability, CHMP may need to evaluate whether the authorised shelf life and storage conditions remain supported by the available evidence.

Relevant information can include:

The regulatory consequence should be linked to the demonstrated quality risk rather than inferred from a single anomalous result.

34. Efficacy Referrals: Scientific Assessment

An efficacy-related referral focuses on whether the available evidence continues to support the relevant therapeutic claims.

The assessment can consider:

The committee's assessment should be read in relation to the specific indication and population under consideration.

35. Loss of Efficacy Versus Uncertainty About Efficacy

A distinction should be maintained between evidence that a treatment is ineffective and evidence that the existing efficacy conclusion is uncertain.

These situations can produce different regulatory consequences.

A new study may weaken confidence in a treatment effect without demonstrating that no treatment effect exists.

Conversely, sufficiently robust evidence can establish that the benefit previously relied upon is no longer supported.

The regulatory conclusion should therefore be derived from the evidence rather than from the existence of contradictory data alone.

36. Clinical Relevance of Efficacy Evidence

Statistical evidence should be interpreted alongside clinical relevance.

A statistically detectable effect may have little practical importance, while a clinically important effect may be difficult to estimate precisely in a small population.

The assessment should consider:

The final regulatory decision depends on the totality of evidence and the applicable benefit-risk framework.

37. Cross-Product Consistency

Where several products or applications are included in the same referral, CHMP may need to consider whether the scientific concern applies consistently across them.

The answer should not automatically be assumed to be identical.

Differences in:

can affect the relevance of the underlying concern.

This is why product-level analysis remains important even in a Union-wide referral.

38. The MAH's Scientific Response

The MAH should respond directly to the questions posed by CHMP.

A strong response should make clear:

The objective is not to maximise the volume of the submission. It is to make the scientific reasoning transparent and testable.

39. Handling Uncertainty in the Assessment

Scientific uncertainty is common in complex regulatory procedures.

CHMP may need to make a regulatory judgement even where the evidence is incomplete.

The relevant questions include:

The final regulatory measure should therefore be understood as a response to both evidence and uncertainty.

40. The Scientific Opinion Is Not the Commission Decision

CHMP's opinion is the scientific committee conclusion within the applicable regulatory procedure.

Where the procedure leads to a European Commission decision, that decision is the legally binding Union outcome.

The distinction is essential:

Evidence
   ↓
Rapporteur assessment
   ↓
CHMP scientific assessment
   ↓
CHMP opinion
   ↓
Commission decision, where applicable
   ↓
Implementation

A public summary of the CHMP opinion should therefore not automatically be described as the final legal decision.

41. Possible Regulatory Outcomes

A non-pharmacovigilance Article 31 referral can lead to different outcomes depending on the evidence and applicable legal framework.

Possible measures can include:

The referral itself does not predetermine which outcome will occur.

42. Product-Information Changes

Where the scientific assessment supports continued authorisation with modified conditions, the outcome can require changes to product information.

Depending on the issue, this can affect:

The exact wording should always be taken from the final regulatory documents rather than reconstructed from a summary.

43. Suspension and Revocation

Where the evidence demonstrates that the applicable legal requirements for continued authorisation are no longer satisfied, more restrictive action may be appropriate.

Suspension and revocation are materially different from a simple product-information variation.

The legal instrument should therefore be checked carefully to establish:

44. Re-Examination

Where the applicable legal procedure provides a right to request re-examination of the CHMP opinion, the MAH or applicant must comply with the applicable procedural deadlines.

The re-examination process should be treated as a defined scientific procedure rather than as an informal appeal.

The request should identify the grounds on which reconsideration is sought and should be supported by the relevant scientific or procedural arguments.

The current legislation and referral-specific EMA documentation should be used to determine the exact deadlines and requirements.

45. European Commission Decision

Where the referral proceeds to the Commission decision-making stage, the Commission adopts the legally binding Union decision through the applicable procedure.

The Commission decision should be read together with its annexes because these may contain the operative changes to the marketing authorisation or product information.

For regulatory implementation, the final decision—not an earlier draft or preliminary committee discussion—is the controlling document.

46. Implementation After the Decision

The regulatory process does not end when the scientific opinion is adopted.

The final outcome may require:

Each affected product should have a documented implementation plan.

47. Regulatory Implementation as a Controlled Process

Implementation should be tracked using the organisation's established quality and regulatory systems.

A simple control structure is:

Requirement Owner Deadline Evidence Status
Regulatory variation Regulatory Applicable deadline Submission Open/closed
Product-information update Regulatory/Medical Applicable deadline Approved text Open/closed
Quality action Quality/Manufacturing Defined action Completion record Open/closed
Communication Medical/Safety Defined action Distribution evidence Open/closed
Post-decision monitoring Relevant function Defined plan Monitoring record Open/closed

The actual fields should reflect the company's quality system.

48. Inspection and Audit Trail

An MAH should be able to reconstruct the complete regulatory history of the referral.

The record should connect:

  1. initiating concern;
  2. formal referral;
  3. CHMP questions;
  4. company responses;
  5. scientific assessment;
  6. CHMP opinion;
  7. re-examination, if applicable;
  8. Commission decision;
  9. implementation;
  10. subsequent lifecycle actions.

This is particularly important where a referral has consequences for several products or legal entities.

49. The Difference Between the Referral and Its Implementation

The referral is the regulatory assessment mechanism.

Implementation is what happens after the regulatory authority has determined the appropriate outcome.

The distinction matters because the same referral can produce different operational actions across products depending on their authorisation status and individual regulatory circumstances.

A referral-level conclusion should therefore not be copied directly into an operational action list without checking the final decision and product-specific consequences.

50. Closing the Regulatory Workstream

Once all regulatory actions are implemented, the exceptional referral workstream can be closed according to the company's quality system.

Closure should confirm that:

The referral should remain part of the product's permanent regulatory history.

51. Benefit-Risk Assessment in a Non-Pharmacovigilance Referral

Even when the initiating concern is not pharmacovigilance-related, the ultimate regulatory question can involve the overall balance between the therapeutic benefits and risks of the medicinal product.

The assessment should consider the specific concern that triggered the referral together with the product's established therapeutic benefits and the available alternatives.

Relevant considerations can include:

A quality or efficacy concern therefore should not automatically be interpreted as a negative benefit-risk conclusion. The regulatory response depends on the evidence and the legal requirements for continued authorisation.

52. Regulatory Proportionality

The regulatory response should correspond to the nature and significance of the demonstrated concern.

Depending on the evidence, proportionate action may range from continued authorisation with no substantive change to restrictions, suspension or revocation.

Proportionality should not be interpreted as a presumption that the least restrictive measure will always be selected. Where the evidence demonstrates that the legal requirements for authorisation are no longer satisfied, a more restrictive measure may be required.

The final decision should therefore be understood as the consequence of the evidence and applicable legal standard, not as a predetermined hierarchy of interventions.

53. Uncertainty Does Not Automatically End the Procedure

A referral can reach a regulatory conclusion even when some scientific uncertainty remains.

The committee may need to determine whether the remaining uncertainty is compatible with continued authorisation, whether it can be reduced through additional evidence or whether it requires additional regulatory controls.

The assessment should therefore distinguish:

This distinction is especially important for complex efficacy assessments and quality concerns where evidence may be incomplete but still sufficient for a regulatory decision.

54. The Importance of the Exact Referral Questions

The referral questions define the boundary of the formal scientific assessment.

A company should therefore avoid answering a broader question than the committee has actually posed unless the broader issue is directly relevant to the requested assessment.

A useful internal response matrix is:

CHMP question Evidence Company conclusion Residual uncertainty
Question 1 Defined evidence set Conclusion Remaining issue
Question 2 Defined evidence set Conclusion Remaining issue
Question 3 Defined evidence set Conclusion Remaining issue

This makes the submission easier to review and makes later regulatory reconstruction more reliable.

55. Evidence Should Be Traceable

A referral response should allow material conclusions to be traced to the underlying evidence.

For quality questions, this may mean linking conclusions to batch data, analytical reports, validation records or stability studies.

For efficacy questions, it may mean linking conclusions to clinical-study reports, statistical analyses, datasets or published evidence.

The traceability principle is the same:

Regulatory conclusion
        ↓
Scientific conclusion
        ↓
Underlying evidence
        ↓
Source record

This is both a scientific-quality principle and an inspection-readiness principle.

56. Conflicting Evidence

Conflicting evidence should be addressed explicitly rather than selectively omitted.

The assessment may need to explain differences in:

A scientifically defensible response should explain why the conflicting evidence does or does not alter the overall conclusion.

57. Quality Evidence: Batch-Level Findings

A batch-level finding can be important without necessarily demonstrating a systemic quality failure.

The assessment should consider whether the finding is:

The regulatory significance depends on the evidence establishing the scope of the problem.

58. Quality Evidence: Analytical Methods

Analytical-method performance can be central to a quality referral.

The assessment may need to examine:

An apparent quality signal can sometimes arise from analytical methodology rather than from an actual change in the product.

The committee therefore needs sufficient evidence to distinguish analytical artefact from genuine product-quality concern.

59. Quality Evidence: Impurities and Degradation Products

Where the concern involves impurities or degradation products, the assessment can require characterisation of the impurity, its formation pathway and its toxicological or clinical significance.

Relevant questions can include:

The appropriate regulatory response depends on the demonstrated risk and the ability to control it.

60. Quality Evidence: Manufacturing-Site Differences

Where several manufacturing sites are involved, evidence should be evaluated at site level as well as product level.

Differences in:

can influence whether a concern applies uniformly across products.

A Union-level referral does not eliminate the need for this granular assessment.

61. Efficacy Evidence: Endpoint Interpretation

Efficacy conclusions depend heavily on the endpoint being assessed.

A statistically significant effect on a surrogate endpoint does not necessarily establish a clinically meaningful benefit on a patient-important outcome.

The assessment should therefore consider:

This is particularly important where a referral questions whether an existing indication remains adequately supported.

62. Efficacy Evidence: Study Design

Study design can materially influence the reliability of efficacy evidence.

Relevant factors include:

A referral assessment should not treat the numerical result independently of the design that generated it.

63. Efficacy Evidence: External Validity

Even a well-controlled study may not answer whether the observed benefit remains applicable to the current authorised population.

External validity considerations can include:

Where the referral concerns continued validity of an indication, these factors may be important to the regulatory conclusion.

64. Efficacy Evidence: Loss of Effect Over Time

A referral can arise when later evidence suggests that a previously demonstrated treatment effect is smaller, less durable or less consistent than originally believed.

The assessment should determine whether the apparent change reflects:

The final regulatory conclusion should be based on the integrated evidence rather than on a single later study.

65. Cross-Functional Governance

Non-pharmacovigilance referrals frequently require coordination between regulatory affairs, quality, manufacturing, clinical, statistics, medical and safety functions.

A defined governance structure should establish:

The governance model should reflect the actual referral rather than automatically using a pharmacovigilance referral template.

66. QPPV Relevance

A non-pharmacovigilance referral does not automatically make the QPPV the procedural owner.

However, the QPPV can still have an important oversight role where the outcome affects the pharmacovigilance system, product safety information, risk management or ongoing benefit-risk monitoring.

The QPPV should therefore be informed when the referral has material implications for pharmacovigilance obligations, even if the principal scientific issue is quality or efficacy.

This distinction prevents the QPPV role from being either overstated or unnecessarily excluded.

67. Pharmacovigilance Consequences of a Quality Referral

A quality-related referral can produce pharmacovigilance consequences.

For example, a manufacturing or contamination issue may alter the clinical risk profile of the product and require enhanced monitoring, case identification or communication.

The MAH should therefore assess the final regulatory outcome for implications across the entire product-safety system.

The initiating legal category does not determine every downstream pharmacovigilance consequence.

68. Pharmacovigilance Consequences of an Efficacy Referral

An efficacy-related regulatory decision can also affect the benefit-risk assessment used in pharmacovigilance activities.

Changes to an indication, patient population or conditions of use may alter:

The pharmacovigilance system should therefore incorporate the final regulatory outcome where it changes the product's benefit-risk framework.

69. Risk Management Plan Implications

Where the final outcome changes the safety profile, conditions of use or risk-management requirements, the RMP may need to be updated.

The update should be based on the final regulatory conclusion rather than preliminary hypotheses discussed during the referral.

Possible consequences can include changes to:

The applicable current RMP guidance should be used to determine the exact requirements.

70. Implementation Across Multiple Marketing Authorisations

Where several marketing authorisations are affected, implementation should be tracked separately for each authorisation.

Relevant fields can include:

A Union-level decision can therefore generate multiple operational implementation records.

71. Product Information as an Implementation Instrument

Where the outcome requires changes to product information, the final authorised wording should be treated as the controlling text.

Internal summaries should not replace the approved wording.

The implementation process should reconcile:

Consistency is important because conflicting internal and approved information can create additional compliance risk.

72. Communication Following the Decision

A regulatory decision can require or justify communication to healthcare professionals, patients or other stakeholders.

The communication should remain within the scope of the final regulatory outcome.

It should not transform an uncertain scientific finding into an unsupported statement of fact or imply that a regulatory measure has a broader scope than the decision actually establishes.

The final approved regulatory language should therefore be the starting point for communication development.

73. Monitoring Regulatory Commitments

The implementation team should maintain a controlled list of commitments resulting from the referral.

Each commitment should have:

This is particularly important when implementation crosses legal entities, manufacturing sites or Member States.

74. Deviations and Delayed Implementation

If an implementation action is delayed, the issue should be managed under the applicable quality and regulatory processes.

The assessment should consider:

A referral does not remove ordinary quality-system responsibilities.

75. Inspection Readiness

An MAH should be able to demonstrate that it understood and implemented the final outcome.

A useful inspection package should connect:

Referral
   ↓
CHMP assessment
   ↓
Company response
   ↓
Final opinion / decision
   ↓
Implementation
   ↓
Ongoing lifecycle control

The record should be sufficiently organised that an inspector can reconstruct the sequence without relying on individual staff memory.

76. Document Control

Referral documentation should be maintained under appropriate document-control procedures.

This includes distinguishing:

The final regulatory conclusion should always be traceable to its authoritative source.

77. Public Documents Versus Primary Regulatory Documents

Public summaries are useful for understanding a referral but should not automatically be treated as the authoritative legal source.

A practical hierarchy is:

  1. applicable legislation;
  2. formal referral documents;
  3. final Commission decision or applicable national/Union regulatory instrument;
  4. final CHMP opinion;
  5. assessment reports and supporting committee documents;
  6. EMA procedural guidance;
  7. public summaries and secondary explanations.

The appropriate document depends on the question being asked.

78. How to Cite the Referral

Regulatory writing should identify the referral precisely enough for another professional to reproduce the source.

Where relevant, cite:

A generic reference to "an EMA referral" is insufficient where the conclusion depends on a specific procedure.

79. Regulatory Fact Versus Interpretation

A defensible article or regulatory assessment should distinguish three types of statement:

Regulatory fact: what the legal or procedural document says.

Scientific finding: what the evidence demonstrates or supports.

Interpretation: the reviewer's conclusion based on those facts and findings.

This distinction prevents secondary interpretation from being presented as though it were a legal requirement.

80. Article 31 Is a Process, Not a Single Document

A referral should be reconstructed as a sequence of documents and decisions.

The central sequence is:

Initiating concern
      ↓
Formal referral
      ↓
CHMP assessment
      ↓
Questions / responses
      ↓
CHMP opinion
      ↓
Re-examination, if applicable
      ↓
Commission decision, where applicable
      ↓
Implementation

Each document answers a different question and has a different procedural status.

81. Common Misreading: Referral Equals Failure

The existence of a referral does not establish that the product has failed to meet the requirements for authorisation.

The referral is the mechanism through which the question is assessed.

The final conclusion depends on the evidence and applicable legal standard.

This is particularly important for public-facing descriptions, where the word "referral" can otherwise be interpreted as a finding of non-compliance.

82. Common Misreading: Quality Concern Equals Quality Defect

A quality concern can be broader than a confirmed quality defect.

The referral may ask whether the available evidence establishes a systemic problem, whether the authorised specifications remain appropriate or whether additional controls are necessary.

The final regulatory conclusion should therefore be used to determine what has actually been established.

83. Common Misreading: Efficacy Concern Equals No Efficacy

Evidence that challenges an efficacy claim does not automatically demonstrate absence of efficacy.

The committee may conclude that evidence remains sufficient, that the indication should be narrowed, that additional conditions are required or that the benefit is no longer adequately demonstrated.

The specific conclusion matters.

CHMP's opinion and the final legal decision should be distinguished whenever the procedure proceeds to the Commission decision-making stage.

The opinion explains the committee's scientific conclusion. The final decision establishes the legally effective Union measure.

For implementation, the final applicable regulatory instrument should be used.

85. Common Misreading: The Public Summary Is the Whole Procedure

A public summary can omit procedural detail that is important for regulatory interpretation.

The full record may contain:

For a live regulatory assessment, the primary documents should therefore be reviewed.

86. Practical Checklist for an Article 31 Non-Pharmacovigilance Referral

When reviewing a referral, ask:

  1. What is the exact legal basis?
  2. Who initiated the procedure?
  3. Why are the interests of the Union involved?
  4. Which products are in scope?
  5. Are the products nationally or centrally authorised?
  6. Is Article 31 or another referral mechanism applicable?
  7. What are the formal questions?
  8. What evidence has CHMP assessed?
  9. What questions were sent to the MAHs or applicants?
  10. What did the company submit?
  11. What did CHMP conclude?
  12. Was re-examination requested?
  13. What is the final legal outcome?
  14. What implementation actions are required?
  15. What quality, efficacy, safety or RMP consequences follow?
  16. How has completion been documented?

87. Practical Takeaways for Regulatory and Quality Professionals

The most important principles are:

88. Final Perspective

An Article 31 non-pharmacovigilance referral is best understood as a Union mechanism for resolving a defined regulatory concern where the interests of the Union are involved and the applicable legal conditions for the referral are satisfied.

Its value lies in converting a potentially fragmented national or product-specific concern into a structured Union scientific and regulatory assessment.

The essential chain is:

Regulatory concern
      → Article 31 referral
      → CHMP assessment
      → Scientific opinion
      → Commission decision where applicable
      → Product-level implementation
      → Ongoing lifecycle control

The procedure should not be reduced to the existence of a referral, the identity of the committee or the publication of a public summary. The legally relevant question is what the competent authorities ultimately concluded, through which legal instrument, and what that conclusion requires the affected marketing authorisation holders and competent authorities to do.

For regulatory, quality and pharmacovigilance professionals, that distinction is the foundation for interpreting Article 31 correctly.

References

  1. European Parliament and Council. Directive 2001/83/EC on the Community code relating to medicinal products for human use, as amended. In particular Article 31 and the provisions governing Union referral procedures, marketing authorisations and implementation of referral outcomes.
  2. European Parliament and Council. Regulation (EC) No 726/2004, as amended, laying down Union procedures for the authorisation, supervision and pharmacovigilance of medicinal products for human and veterinary use and establishing the European Medicines Agency. In particular the provisions relevant to centrally authorised products and Article 20 referrals.
  3. European Medicines Agency. Questions and answers: Article 31 referrals. Current EMA procedural information concerning referrals under Article 31 of Directive 2001/83/EC, including the distinction between pharmacovigilance and non-pharmacovigilance referrals.
  4. European Medicines Agency. Questions and answers: Article 31 non-pharmacovigilance referrals. Current procedural information concerning the legal basis, scope and handling of non-pharmacovigilance Article 31 referrals.
  5. European Medicines Agency. Committee for Medicinal Products for Human Use (CHMP). Current information concerning CHMP responsibilities, scientific assessment and referral procedures.
  6. European Medicines Agency. Guide to information on human medicines evaluated by EMA — What EMA publishes and when. Current information concerning publication of referral announcements, opinions, assessment reports, annexes and related regulatory documents.
  7. European Commission. Notice to Applicants, Volume 2A — Procedures for marketing authorisation, Chapter 3: Union Referral Procedures. The current version should be consulted for the applicable procedural framework and implementation requirements.
  8. European Commission. Notice to Applicants, Volume 2A — Decision-making procedure for the adoption of Commission Decisions. The current version should be consulted when a referral proceeds to a Commission decision.
  9. European Medicines Agency and CMDh. Procedural guidance relevant to nationally authorised medicinal products, mutual-recognition and decentralised procedures. Current guidance should be consulted where the referral affects products authorised through these routes.
  10. European Medicines Agency. Quality, efficacy and benefit-risk guidance applicable to human medicinal products. The current scientific and regulatory guidance relevant to the particular issue should be consulted for live procedures.

Primary-document hierarchy

When a general explanation conflicts with the documents governing an individual referral, the procedure-specific primary sources should prevail. A practical hierarchy is:

  1. applicable EU legislation;
  2. formal referral notification and procedure-specific documents;
  3. final European Commission decision or other applicable binding regulatory instrument;
  4. final CHMP opinion;
  5. procedure-specific assessment reports and annexes;
  6. current EMA and European Commission procedural guidance;
  7. secondary summaries and explanatory material.

The hierarchy is practical rather than absolute: the appropriate primary document depends on the regulatory question being asked. A scientific question may require the assessment report, while the operative legal requirement may be found in the final decision and its annexes.

Regulatory Note

This article is an educational and regulatory-reference document. It is intended to explain the structure and practical interpretation of Article 31 non-pharmacovigilance referrals for medicinal products for human use in the European Union. It is not legal advice and should not be used as a substitute for applicable legislation, official regulatory guidance or procedure-specific documents.

The article has been prepared using the EU regulatory framework and official EMA/European Commission materials relevant to the subject. EU medicines legislation, procedural guidance and publication practices can change. A live regulatory assessment must therefore use the current consolidated legislation and current procedure-specific documentation.

The article deliberately distinguishes general procedural principles from case-specific facts. The existence, scope, timetable, questions, committee documents, re-examination rights, implementation requirements and final outcome of an individual Article 31 referral must be established from the documents for that specific procedure.

In particular, the existence of an Article 31 referral should not be interpreted as proof that a medicinal product has failed to meet the requirements for authorisation. The referral is the mechanism through which the defined regulatory concern is assessed. The final regulatory conclusion depends on the evidence and the applicable legal standard.

Do not infer the final legal outcome solely from a CHMP discussion, preliminary assessment, public summary or secondary source. Where applicable, the final European Commission decision and its annexes should be checked for the operative Union requirements.

Where an individual regulatory procedure is underway, the procedure-specific EMA timetable, communications, formal regulatory documents, applicable national measures and applicable EU legal provisions take precedence over generic descriptions in this article.

Nothing in this article should be interpreted as creating a regulatory obligation that is not contained in applicable EU legislation, a binding regulatory decision or another applicable regulatory requirement.

Revision History

Related Articles

Article 31 Pharmacovigilance Referral: Legal Basis, Procedure and Regulatory Consequences in the EU

A textbook-quality guide to Article 31 pharmacovigilance referrals, explaining why the procedure is initiated, how the Union-level assessment operates, how PRAC, CMDh and CHMP interact, and how the final regulatory outcome is implemented and incorporated into pharmacovigilance systems.

Article 20 Pharmacovigilance Procedures for Centrally Authorised Medicines

A textbook explanation of the Article 20 pharmacovigilance procedure for centrally authorised medicines, including its legal trigger, initiation by the European Commission, PRAC and CHMP roles, MAH participation and implementation of the regulatory outcome.

Article 30 Referral in EU Medicines Regulation: Divergent National Decisions and Harmonisation

A structured explanation of Article 30 referrals, the Union mechanism used to resolve qualifying divergent national decisions concerning the same medicinal product and establish a harmonised regulatory position.

Last reviewed: 2026-08-23