GVP Module VIII: When Is a PASS Needed?

Introduction

A post-authorisation safety study (PASS) is not simply a study performed after marketing authorisation. Under EU pharmacovigilance legislation, a PASS has a specific safety purpose: identifying, characterising or quantifying a safety hazard, confirming the safety profile of a medicinal product, or measuring the effectiveness of risk-management measures. A PASS can be interventional or non-interventional, although GVP Module VIII places particular emphasis on non-interventional PASS. citeturn0search18

The practical question is not simply whether a study can be performed after authorisation. It is whether a PASS is the appropriate and proportionate pharmacovigilance activity for the uncertainty or safety question that needs to be addressed.

Not every unanswered safety question requires a PASS. Some questions can be addressed through routine pharmacovigilance, signal management, targeted follow-up, literature monitoring or existing data. Other questions require structured study-based evidence.

1. Start With the Safety Question

The decision should begin with a clearly defined safety question. Examples include whether a suspected association is real, how frequently an event occurs, which factors increase risk, which populations are vulnerable, whether exposure changes risk, whether a risk-minimisation measure works in practice, or whether previously missing information can be characterised.

The question determines what evidence is needed. A study should not be selected first and justified afterwards.

2. Four Broad PASS Situations

GVP Module VIII identifies situations in which non-interventional PASS may be conducted: studies imposed by an EU competent authority, specific obligations associated with certain authorisation circumstances, studies included in an RMP to investigate a safety concern or evaluate risk-minimisation effectiveness, and studies conducted voluntarily by an MAH. citeturn0search18

The legal basis and procedural consequences differ between these categories.

3. Imposed PASS

An EU competent authority can impose a non-interventional PASS in specified circumstances. EMA describes imposed non-interventional PASS as potentially being required as a condition of a marketing authorisation or as a specific obligation associated with a conditional marketing authorisation or a marketing authorisation under exceptional circumstances. citeturn0search0turn0search5

An imposed study should therefore be treated as a regulatory obligation, not merely as an internal pharmacovigilance project. The organisation needs controlled processes for identifying the obligation, understanding its legal basis, establishing milestones, preparing the protocol, obtaining required regulatory review, conducting the study, reporting results and demonstrating fulfilment.

4. RMP-Driven PASS

A PASS can form part of the pharmacovigilance plan in the RMP. The pharmacovigilance plan is used to describe how safety concerns will be further characterised, including investigation of potential risks, characterisation of risks, investigation of missing information and measurement of risk-minimisation effectiveness. citeturn0search20

An RMP study should therefore be connected to a defined safety-management need. The organisation should be able to explain what the study is intended to answer and how the result may affect the safety profile, risk-management system or regulatory understanding.

5. Voluntary PASS

An MAH may conduct a PASS voluntarily when additional evidence is scientifically useful even though the study has not been imposed by an authority.

Voluntary does not mean scientifically uncontrolled. The MAH remains responsible for appropriate rationale, methodology, governance and quality. EMA recommends that companies submit protocols and study reports in the same manner for voluntary PASS as for imposed PASS, although this is not mandatory. citeturn0search0

6. When Routine Pharmacovigilance May Be Enough

A PASS should not automatically be selected whenever there is uncertainty. Routine pharmacovigilance is the minimum set of activities required for medicinal products and includes activities such as adverse-reaction reporting and signal detection. citeturn0search20

Routine methods may be sufficient where the question can reasonably be addressed through spontaneous reporting, literature monitoring, signal detection, case follow-up, medical review or aggregate analysis.

The important question is whether the available method can answer the safety question with sufficient reliability.

7. When a PASS May Add Value

A PASS becomes more compelling when the required evidence cannot be obtained reliably from routine sources.

An epidemiological study may, for example, provide a denominator, incidence estimate, comparative risk estimate or analysis of risk factors that spontaneous reports cannot provide reliably. A structured study may also be appropriate when determining whether a risk-minimisation intervention changed prescribing, exposure, behaviour or clinical outcomes.

8. Fit-for-Purpose Evidence Generation

A sophisticated study is not automatically a better pharmacovigilance study. The design should reflect the question, available data, feasibility, expected bias and required precision.

An unnecessarily complex study can introduce delay and operational risk without improving the answer. The objective should be fit-for-purpose evidence generation.

9. Proportionality

The intensity of additional pharmacovigilance should be proportionate to the risks and uncertainties of the product.

Safety concern / uncertainty
          ↓
What exactly must be known?
          ↓
What evidence already exists?
          ↓
Can routine PV answer the question?
          ↓
If not, what additional activity is appropriate?
          ↓
Would a PASS provide meaningful additional evidence?
          ↓
Select the least burdensome scientifically adequate design

This prevents the study itself from becoming the starting point of the reasoning.

10. PASS Versus Other Additional Pharmacovigilance Activities

Not every additional pharmacovigilance activity is a PASS. Depending on the question, alternatives may include enhanced spontaneous-report monitoring, targeted follow-up, specific questionnaires, registries, literature analyses, database studies or signal analyses.

The organisation should document why the selected activity is appropriate for the question.

11. PASS and Risk-Minimisation Effectiveness

A study may be used to measure risk-minimisation effectiveness. Implementation alone does not demonstrate effectiveness.

The study question should distinguish whether a measure was implemented, reached the intended population, changed behaviour and ultimately affected the safety outcome. These are different questions and may require different evidence.

12. PASS and Missing Information

Missing information can sometimes justify additional data generation, but does not automatically mean a PASS is required.

The MAH should determine what information is missing, why it matters, whether routine data may resolve the uncertainty, whether another activity could address it and whether the remaining uncertainty materially affects benefit-risk.

The study should address a defined information gap rather than merely produce more data.

13. PASS and Important Risks

An important risk may require further characterisation. A PASS may be appropriate when evidence is needed about incidence, relative risk, risk factors, susceptible populations, severity, duration, clinical consequences or effectiveness of an intervention.

The presence of an important risk is therefore a reason to assess whether additional evidence is required, not an automatic instruction to conduct a PASS.

14. Documenting the Decision

Whether a PASS is selected or rejected, material reasoning should be documented. A useful decision record can capture the safety question, current evidence, residual uncertainty, alternatives considered, limitations of those alternatives, proposed objectives and the rationale for the final decision.

This creates a defensible bridge between scientific assessment and the selected pharmacovigilance activity.

Key Takeaways

A PASS is justified by a defined safety or risk-management question, not simply by the fact that a medicinal product is authorised.

The main situations include imposed PASS, specific obligations, RMP-driven studies and voluntary studies. The legal consequences differ between them.

Routine pharmacovigilance may be sufficient for some questions. A PASS becomes useful when structured study-based evidence is needed to resolve uncertainty that routine methods cannot adequately answer.

The governing principle is fit-for-purpose, proportionate evidence generation.

References

1. European Medicines Agency. GVP Module VIII — Post-authorisation safety studies (Rev. 3). citeturn0search18
2. European Medicines Agency. Post-authorisation safety studies (PASS). citeturn0search0
3. European Medicines Agency. GVP Module V — Risk management systems (Rev. 2). citeturn0search20
4. Directive 2001/83/EC, as amended.
5. Regulation (EC) No 726/2004, as amended.
6. Commission Implementing Regulation (EU) No 520/2012, as amended.

Regulatory Note

This article is an educational explanation of when a PASS may be required or scientifically appropriate. It does not replace current EU legislation, GVP Module VIII, Module V, EMA procedural guidance or approved procedures.

Regulatory requirements and procedures may change. The current legal basis and procedural guidance should be verified before making a regulatory commitment or initiating a PASS.

15. Regulatory Obligation Versus Scientific Need

The first distinction in a PASS decision is whether the study is required or merely useful.

An imposed PASS is driven by a regulatory obligation. An RMP PASS is connected to a defined pharmacovigilance need. A voluntary PASS is initiated by the MAH because additional evidence is scientifically useful. These categories should remain distinguishable in the quality system. citeturn0search18

16. PASS as a Condition of Authorisation

A PASS can be imposed as a condition of a marketing authorisation. The study then forms part of the regulatory conditions associated with the product. The organisation should maintain a clear link between the authorisation condition, study objective, protocol, regulatory milestone, final result and evidence demonstrating fulfilment. citeturn0search0

17. Specific Obligations

A PASS may also be imposed as a specific obligation associated with a conditional marketing authorisation or a marketing authorisation under exceptional circumstances. This is a regulatory commitment and should not be managed as an ordinary voluntary research project. citeturn0search0

18. PASS in the RMP

A PASS can be included in the RMP where additional evidence is needed to investigate a safety concern or evaluate risk-minimisation activities. GVP Module V describes the pharmacovigilance plan as a structured means of further characterising safety concerns and measuring risk-minimisation effectiveness. citeturn0search20

A useful RMP description should identify the safety concern, knowledge gap, study objective, information required and milestone. Simply writing "conduct PASS" does not explain the scientific purpose.

19. Voluntary Evidence Generation

A voluntary PASS may be appropriate where an MAH identifies important residual uncertainty that cannot be adequately addressed with existing evidence. Examples include estimating incidence, investigating risk factors, characterising a population or evaluating a safety intervention.

Voluntary does not mean scientifically uncontrolled. EMA recommends that companies submit protocols and study reports in the same manner for voluntary PASS as for imposed PASS, although this is not mandatory. citeturn0search0

20. When a PASS Is Probably Not the Right Tool

A PASS may be inappropriate when another method can answer the question more reliably or efficiently.

Examples include an individual case requiring follow-up, an initial signal that can be characterised using existing cases, a literature question addressed through systematic monitoring, a data-quality problem requiring database remediation, or a regulatory question requiring clarification rather than new evidence.

A study should not be used to compensate for a failure in an existing PV process.

21. The Denominator Problem

Spontaneous reporting is valuable for signal detection but generally does not provide a reliable denominator for estimating incidence. Where the question specifically requires an estimate of occurrence in an exposed population, a suitable epidemiological study may provide evidence that routine reporting cannot.

The study must still address its own sources of bias and uncertainty.

22. The Comparative-Risk Problem

Some questions require comparison with an appropriate reference population or comparator exposure. Spontaneous reports alone may be insufficient for such questions. An observational study may provide comparative evidence, but comparator selection, confounding control and outcome definition become important.

The existence of a comparative question does not dictate one particular design.

23. Risk-Factor Characterisation

A PASS may be useful when the organisation needs to identify factors associated with an adverse outcome, such as age, organ impairment, concomitant medication, dose, duration or underlying disease.

The study should specify which factors are clinically meaningful and how they will be evaluated.

24. Risk-Minimisation Effectiveness

A risk-minimisation measure may be implemented correctly but fail to produce its intended effect. A structured study may therefore be needed to determine whether the measure reached the intended population, was understood, changed behaviour or reduced the targeted risk.

Implementation, reach, behavioural change and clinical effectiveness are different endpoints.

25. Study Feasibility

Before committing to a PASS, the organisation should consider population availability, data-source coverage, linkage, outcome ascertainment, follow-up, sample size, geographic coverage and other feasibility constraints.

A scientifically important question does not justify a study that cannot generate interpretable evidence.

26. Evidence Is Question-Specific

There is no universal hierarchy in which one evidence source is always superior. Spontaneous reports may be highly sensitive for rare-signal detection; a database study may be needed for incidence; a targeted questionnaire may be sufficient for another question.

The appropriate method depends on the causal question, outcome, exposure, population and uncertainty.

27. Existing Data

A PASS does not necessarily require prospective data collection. Existing healthcare or other data sources may provide suitable information for a non-interventional study when the source and design are appropriate for the objective.

The organisation should avoid creating new data collection merely because it is administratively easier than evaluating existing data.

28. PASS and Registries

A registry may support a PASS but is not synonymous with a PASS. A registry is a data-collection structure; a PASS is a study conducted for the safety purposes established in EU legislation.

A registry can therefore be a data source for a PASS, part of a PASS, an additional pharmacovigilance activity or an activity serving another purpose.

29. PASS and Signal Management

Signal management can identify the need for further investigation, but a signal should not automatically become a PASS. Existing cases, literature, clinical review or other evidence may be sufficient. A PASS becomes relevant when a structured study is needed to resolve an important residual question.

30. PASS and PSUR

A PSUR provides periodic evaluation of the safety profile and benefit-risk balance. A PASS generates additional evidence that can feed into that evaluation. The processes should therefore be connected but not confused.

PASS findings should be incorporated into subsequent aggregate safety assessment when relevant.

31. PASS and RMP Lifecycle

When a PASS is included in the RMP, its progress and results can affect subsequent RMP versions.

Safety concern
      ↓
RMP pharmacovigilance activity
      ↓
PASS objective
      ↓
Study result
      ↓
Updated safety understanding
      ↓
RMP / PV / risk-minimisation consequences

This prevents the study from becoming an isolated research activity.

32. Practical Decision Matrix

This is a reasoning aid, not a regulatory decision tree.

33. Practical Scenario: No PASS Needed

Suppose a potential signal concerns a rare event and the MAH has sufficient cases for clinical assessment supported by relevant literature and medical review. The organisation may determine that a new study is not yet necessary.

That decision should be based on evidence and uncertainty, not a desire to avoid research.

34. Practical Scenario: Incidence Cannot Be Estimated

Suppose a safety concern requires estimating incidence among exposed patients, but spontaneous reporting cannot provide a reliable denominator. A suitable database or other epidemiological study may therefore be justified.

The protocol should define the population, exposure, outcome, comparator where applicable and analytical approach sufficiently to answer the question.

35. Practical Scenario: RMM Effectiveness

Suppose an additional risk-minimisation measure requires healthcare professionals to follow specific prescribing or monitoring behaviour. The organisation should distinguish distribution of materials from actual behavioural change.

A study designed around the appropriate effectiveness endpoint may be needed to determine whether the measure is achieving its intended purpose.

Key Takeaways

The decision to conduct a PASS should be anchored in the source of the obligation, the safety question, the evidence gap and the ability of alternative methods to answer that question.

A PASS is most useful when it generates evidence that existing pharmacovigilance methods cannot provide adequately.

References

1. European Medicines Agency. GVP Module VIII — Post-authorisation safety studies (Rev. 3). citeturn0search18
2. European Medicines Agency. Post-authorisation safety studies (PASS). citeturn0search0
3. European Medicines Agency. GVP Module V — Risk management systems (Rev. 2). citeturn0search20
4. Directive 2001/83/EC, as amended.
5. Regulation (EC) No 726/2004, as amended.
6. Commission Implementing Regulation (EU) No 520/2012, as amended.

Regulatory Note

Examples and decision matrices are educational and illustrative and should not be interpreted as automatic regulatory criteria.

36. A PASS Decision Should Survive Challenge

A mature PASS decision can be explained by someone who was not involved in making it. The organisation should be able to show the question, evidence, uncertainty, alternatives considered and rationale for selecting or rejecting a PASS.

The decision should also identify whether the activity is legally imposed, an RMP commitment or voluntary.

37. Regulatory Obligation Mapping

For an imposed study, the organisation should map the obligation to a controlled set of milestones:

Regulatory obligation
        ↓
Study objective
        ↓
Protocol
        ↓
Regulatory review / endorsement
        ↓
Study conduct
        ↓
Final report
        ↓
Regulatory outcome
        ↓
Commitment closure

This prevents the organisation from treating study completion as sufficient evidence of regulatory compliance when the actual obligation includes additional milestones.

38. Protocol Quality Is Part of the Decision

A PASS should not be considered justified merely because its topic is important. The proposed protocol must be capable of answering the question.

The organisation should consider whether the proposed design adequately addresses:

• study population;
• exposure definition;
• outcome definition;
• comparator where relevant;
• confounding;
• bias;
• data quality;
• follow-up;
• sample size or precision;
• and analysis.

A weak protocol can convert a valid regulatory or scientific question into an uninformative study.

39. When the Question Changes

The need for a PASS should be reassessed if the underlying safety question changes.

For example, new evidence may:

• confirm the suspected risk;
• refute it;
• change its importance;
• identify a different population at risk;
• or make another evidence source more appropriate.

The organisation should therefore avoid continuing a study simply because it was once considered appropriate.

For imposed PASS, however, any change must also be managed within the applicable regulatory procedure and commitment framework.

40. Substantial Protocol Changes

Where an imposed non-interventional PASS is underway, substantial amendments to the agreed protocol require the applicable regulatory process before implementation. EMA identifies examples of substantial changes as those likely to affect participant safety or the study results or their interpretation, including changes to objectives, population, sample size, design, data sources, data collection, exposure/outcome definitions or statistical analysis. citeturn0search0

This illustrates why the original scientific rationale must remain traceable throughout the study lifecycle.

41. PASS Milestones and Governance

A PASS should have visible milestones appropriate to its regulatory context.

Governance should address:

• protocol development;
• regulatory submission or notification where applicable;
• study initiation;
• data collection;
• analysis;
• final report;
• regulatory review;
• commitments;
• and follow-up.

For imposed non-interventional PASS, EMA maintains a specific timetable for protocol and final-results procedures; the current timetable was updated in July 2026. citeturn0search0

42. The HMA-EMA Catalogue

For imposed non-interventional PASS, the applicable legal framework includes registration and publication requirements in the HMA-EMA catalogue of real-world data studies.

EMA states that MAHs should register relevant PASS information and that the protocol is to be entered before data collection, with the abstract of the final study report submitted within the specified period after finalisation. citeturn0search0

This creates an additional transparency and records-management interface that should be incorporated into the study workflow.

43. Study Completion Is Not Automatically Commitment Closure

A final study report does not necessarily mean that all regulatory consequences have been completed.

The organisation should determine whether the result requires:

• regulatory action;
• product-information changes;
• RMP updates;
• further pharmacovigilance;
• risk-minimisation changes;
• additional studies;
• or other follow-up.

Commitment closure should therefore be based on the applicable regulatory requirement, not merely the existence of a final report.

44. Practical Scenario: New Evidence Makes the Study Unnecessary

Suppose a voluntary PASS was initiated to characterise a potential risk. Subsequent evidence from a robust external study conclusively answers the original question.

The MAH should reassess whether continuing its own study remains scientifically justified.

For a voluntary study, discontinuation may be appropriate if the rationale has disappeared, subject to appropriate governance and commitments already made.

For an imposed study, the MAH cannot simply stop because the scientific rationale has changed; the applicable authority and regulatory procedure must be considered.

45. Practical Scenario: Study Is Too Small

Suppose an important safety question remains, but interim feasibility analysis shows that the available population is substantially smaller than anticipated.

The organisation should determine whether the study can still answer the question with meaningful precision.

Options may include methodological revision, additional data sources or another evidence-generation approach. Any change to an imposed protocol must follow the applicable regulatory process.

46. Practical Scenario: PASS Result Changes the Safety Profile

A PASS may generate evidence that materially changes understanding of a safety concern.

The result should then feed into the broader pharmacovigilance system, including as appropriate:

• signal management;
• PSUR assessment;
• RMP revision;
• risk-minimisation evaluation;
• and regulatory assessment.

The study should therefore have a defined pathway for translating results into pharmacovigilance action.

47. Inspection Perspective

An inspector may ask:

• Why was this PASS initiated?
• What exact safety question was it intended to answer?
• Was the study imposed or voluntary?
• Where is the legal or scientific rationale?
• Why was this design selected?
• What alternatives were considered?
• How are milestones controlled?
• How are protocol amendments governed?
• How are results incorporated into the RMP and PSUR?
• What happened after the final result?

The organisation should be able to answer using contemporaneous records rather than reconstructing the rationale retrospectively.

48. Common PASS Decision Failures

Potential weaknesses include:

• treating every safety uncertainty as a reason for a PASS;
• failing to distinguish imposed and voluntary studies;
• starting a study without a sufficiently precise question;
• selecting an inappropriate design;
• failing to reassess a changing safety question;
• weak protocol governance;
• treating completion as equivalent to regulatory closure;
• and failing to feed study results back into the wider PV system.

These are illustrative deficiency patterns, not a catalogue of confirmed inspection findings.

49. QPPV Oversight

For significant PASS activity, the QPPV should have sufficient visibility to understand:

• why the study exists;
• whether it is imposed or voluntary;
• whether milestones are controlled;
• whether important deviations or delays are escalated;
• whether the study remains scientifically appropriate;
• and whether results are incorporated into the pharmacovigilance system.

The QPPV need not personally manage every study activity, but should have appropriate assurance over significant pharmacovigilance obligations.

50. PASS Decision Checklist

Before deciding that a PASS is needed, ask:

• What is the exact safety or risk-management question?
• What is already known?
• What uncertainty remains?
• Can routine PV answer it?
• Can existing external or internal data answer it?
• Would another additional PV activity be sufficient?
• Is a PASS legally imposed?
• Is it an RMP commitment?
• Is it voluntary?
• Is the proposed study design capable of answering the question?
• Are the required population and data sources available?
• What regulatory milestones apply?
• How will the result affect the PV system?

If the answer to these questions is documented, the PASS decision is much easier to defend scientifically and regulatorily.

Key Takeaways

The decision to conduct a PASS should be a documented scientific and regulatory decision rather than a default response to uncertainty.

For imposed studies, the regulatory obligation governs the lifecycle even when the scientific context evolves. For voluntary studies, the organisation should periodically reassess whether the study continues to provide meaningful value.

The strongest PASS systems connect the original safety question to protocol design, regulatory milestones, study results and subsequent pharmacovigilance action.

References

1. European Medicines Agency. GVP Module VIII — Post-authorisation safety studies (Rev. 3). citeturn0search18
2. European Medicines Agency. Post-authorisation safety studies (PASS) and current procedural guidance. citeturn0search0
3. European Medicines Agency. Outcomes of imposed non-interventional PASS. citeturn0search5
4. European Medicines Agency. GVP Module V — Risk management systems (Rev. 2). citeturn0search20
5. Directive 2001/83/EC, as amended.
6. Regulation (EC) No 726/2004, as amended.
7. Commission Implementing Regulation (EU) No 520/2012, as amended.

Regulatory Note

This article is an educational explanation of PASS decision-making. It does not replace current EU legislation, GVP guidance, EMA procedural requirements or approved company procedures. Current legal and procedural requirements should be verified before initiating, modifying or closing a PASS.