Important Potential Risks

An important potential risk is a suspected adverse clinical outcome for which there is a scientific basis to consider a causal relationship with a medicinal product possible, but the available evidence is not yet sufficient to conclude that the relationship is causal. The word important adds a second requirement: if the suspected relationship were later confirmed, it would matter sufficiently to the medicinal product's benefit-risk balance to justify focused risk-management planning.

This makes the category fundamentally different from both a signal and an identified risk. A signal is information that suggests a new potentially causal association, or a new aspect of a known association, and requires further investigation. An important potential risk is a risk-management classification within the RMP. An identified risk is one for which the evidence is already sufficient to support causality.

Purpose and Regulatory Context

EU legislation requires a risk-management system for medicinal products and establishes requirements for risk-management plans. The detailed scientific framework used to decide which safety concerns belong in an RMP is provided principally by Good Pharmacovigilance Practices (GVP) Module V — Risk management systems and the EMA guidance on the EU RMP format.

GVP Module V deliberately narrows the safety specification. The RMP is not intended to contain every safety signal, every theoretical concern, every adverse event observed in development or every uncertainty surrounding a medicine. For potential risks, the RMP should address adverse clinical outcomes for which there is scientific evidence to suspect a possible causal relationship with the medicinal product, while the available evidence remains insufficient to establish causality.

Among those potential risks, only those that are important for risk-management planning should be listed as important potential risks. GVP Module V explains that these are risks which, if further characterised and confirmed, would have an impact on the benefit-risk balance of the medicinal product.

This distinction separates three levels of regulatory reasoning:

1. EU legislation requires the marketing-authorisation holder to operate an appropriate risk-management system and maintain the RMP where applicable.
2. GVP Module V and the EU RMP format guidance explain how safety concerns should be selected, characterised and connected to pharmacovigilance and risk minimisation.
3. Product-specific scientific judgement determines whether a particular suspected adverse outcome meets the criteria for inclusion.

The third step cannot be replaced by a universal numerical threshold or a company checklist.

What Makes a Risk Potential?

The defining feature of a potential risk is uncertain causality.

There must be more than the abstract possibility that harm could occur. GVP Module V requires a scientific basis for suspecting an association between the medicinal product and an undesirable clinical outcome. At the same time, the available evidence must remain insufficient to conclude that the association is causal.

Evidence supporting a potential risk can arise from many sources:

None of these sources carries a fixed regulatory weight. A single observation may be highly informative in one setting and weak in another. Conversely, several individually weak observations do not automatically become persuasive simply because they are numerous.

The assessment therefore depends on the totality of evidence: clinical coherence, biological plausibility, temporality, consistency, alternative explanations, exposure, background incidence, bias, confounding and the quality of the available data.

Potential risk does not mean unproven signal

A potential risk and a signal overlap conceptually but are not interchangeable.

Signal management asks whether available information suggests a new potentially causal association, or a new aspect of a known association, that warrants further investigation. The RMP asks a different question: does this suspected adverse clinical outcome require continued focused risk-management planning because it could materially affect benefit-risk if confirmed?

A signal may therefore be investigated without ever becoming an important potential risk. Equally, an important potential risk may remain in an RMP after the original signal evaluation has moved into a different stage, because a specific uncertainty still requires further characterisation.

The two processes should communicate, but they should not be collapsed into one administrative workflow.

What Makes a Potential Risk Important?

Once there is a scientific basis for suspicion, a second judgement is required: would confirmation of the risk matter to benefit-risk?

Importance is product- and context-specific. Relevant considerations include:

• seriousness and severity of the possible outcome;
• the plausible magnitude of the risk;
• expected or actual population exposure;
• whether particular populations could be disproportionately affected;
• reversibility, preventability and treatability;
• consequences for treatment choice or clinical monitoring;
• the importance of the medicinal product's benefits in the indication;
• the availability of alternative treatments; and
• whether confirmation could require a meaningful change in product information, pharmacovigilance or risk minimisation.

A rare but catastrophic suspected outcome may be important even when the evidence is sparse. A more common but mild suspected outcome may not materially affect benefit-risk. Importance cannot therefore be reduced to seriousness alone, nor to a specific reporting frequency, relative risk or number of cases.

Why over-inclusion is a problem

An RMP overloaded with speculative concerns becomes less useful. It can obscure the risks that genuinely require active management, generate pharmacovigilance activities that do not answer meaningful questions and make lifecycle review difficult.

The correct objective is not to maximise the number of safety concerns. It is to maintain a safety specification that remains proportionate to the medicinal product's current uncertainties and risk-management needs.

Relationship With Other RMP Safety Concerns

The three principal safety-concern categories answer different scientific questions.

The distinction between an important potential risk and missing information is especially important. Missing information is not a suspected adverse reaction. It is a relevant gap in knowledge about safety in a particular use or population. A pregnancy-data gap, for example, is not automatically a potential teratogenic risk. A specific suspected adverse developmental outcome would require its own scientific basis before being classified as a potential risk.

From Evidence to RMP Classification

The decision to include an important potential risk should follow a transparent scientific argument rather than a mechanical rule. A useful way to structure that argument is to separate four questions.

1. What is the suspected clinical outcome?

The concern should be defined precisely enough to support evaluation. Broad labels such as “cardiac toxicity” or “neurological effects” may be too vague to guide evidence collection. The clinical concept, diagnostic criteria, relevant MedDRA terms and clinically meaningful subtypes should be clear enough that future data can be interpreted consistently.

2. What is the scientific basis for suspecting causality?

The RMP should summarise the evidence supporting suspicion and the major limitations. This may include clinical cases, trial imbalances, epidemiological findings, mechanism, class evidence or non-clinical observations. Equally important are competing explanations such as underlying disease, concomitant treatment, ascertainment bias, confounding or background incidence.

A strong RMP rationale does not merely list evidence sources. It explains how they fit together and why the current state of knowledge supports suspicion without yet supporting causal confirmation.

3. Why would confirmation matter?

The importance assessment should explain what would change if the risk were confirmed. Would the benefit-risk balance be materially affected? Would clinical use, monitoring, patient selection or product information need to change? Would a vulnerable population face disproportionate harm?

This prevents speculative concerns from remaining in the RMP simply because they are scientifically interesting.

4. What uncertainty remains, and how will it be reduced?

An important potential risk should normally be associated with a meaningful unanswered question. Examples include whether an association is real, how large the risk may be, whether certain populations are especially susceptible or whether exposure duration changes risk.

The pharmacovigilance plan should then be designed around the question rather than around the label “important potential risk.”

Relationship With the Pharmacovigilance Plan

GVP Module V states that important potential risks included in the RMP would usually require further evaluation as part of the pharmacovigilance plan. This does not mean that every important potential risk automatically requires a new post-authorisation safety study.

The appropriate activity depends on what evidence is missing and which method can realistically reduce the uncertainty.

Routine pharmacovigilance may be sufficient when spontaneous-report follow-up, cumulative review, literature surveillance and signal management can reasonably address the remaining question. Additional pharmacovigilance may be justified when the uncertainty requires more structured data collection or an analytical design.

Possible approaches include:

The study or activity should have a clear objective. “Monitor the risk” is not an adequate scientific question by itself.

Additional pharmacovigilance is not a badge of importance

A common operational error is to assume that the presence of an important potential risk obliges the company to add a study, registry or enhanced-surveillance programme. That reverses the logic of risk management.

The correct sequence is:

important uncertainty → question that must be answered → method capable of answering it → proportionate pharmacovigilance activity.

Where routine activities can answer the question, adding a study may create burden without improving knowledge. Where routine activities cannot answer it, failing to design an appropriate additional activity leaves the RMP scientifically incomplete.

Relationship With Risk Minimisation

Risk minimisation for an unconfirmed risk requires careful judgement. The fact that a risk is potential means the causal relationship remains uncertain, but uncertainty does not automatically justify inaction.

Routine product information may already contain precautionary language where the available evidence supports it. In some circumstances, the seriousness of the possible outcome, the plausibility of the association and the consequences of waiting for stronger evidence may justify precautionary measures before causality is fully established.

However, additional risk-minimisation measures should not be added merely because a concern appears in the RMP. GVP Module XVI requires risk-minimisation measures to be justified by the risk-management objective. Their burden, feasibility and effectiveness should be considered in proportion to the evidence and the potential harm.

A useful distinction is between managing uncertainty and managing an established risk. For an important potential risk, the principal need may be evidence generation. If interim precautions are also justified, the RMP should explain why they are proportionate despite residual causal uncertainty.

Interface With Signal Management

Signal management and RMP lifecycle management should exchange information continuously, but they serve different regulatory purposes.

A newly validated signal may prompt review of the RMP, but it does not automatically require immediate inclusion as an important potential risk. The organisation should first determine whether the scientific evidence supports a suspected causal relationship and whether confirmation would be important to benefit-risk.

Conversely, evidence generated through signal assessment may alter an existing important potential risk. A signal evaluation can:

• strengthen the causal hypothesis while leaving uncertainty unresolved;
• provide enough evidence to support reclassification as an important identified risk;
• weaken the association sufficiently to support removal from the RMP; or
• reveal that the original concern was too broadly defined and should be refined.

The RMP should therefore reflect the current scientific state rather than the administrative status of a signal in an internal tracking system.

Lifecycle Management

An important potential risk should remain in the RMP only while it continues to meet both criteria: there is a scientifically credible suspected association, and confirmation would still be important to risk management.

Retaining the risk

Retention is appropriate when meaningful uncertainty remains and the concern still has potential benefit-risk importance. The rationale should explain what has been learned since the previous RMP and what remains unresolved.

Simply repeating the same text from one RMP version to the next is not evidence of active lifecycle management.

Reclassification as an important identified risk

Reclassification becomes appropriate when the accumulated evidence is sufficient to support causality and the risk remains important for risk management.

No universal statistical threshold defines this transition. A persuasive epidemiological association, coherent clinical cases, mechanistic evidence and consistency across sources may contribute, but the conclusion depends on the totality of evidence.

The practical consequences should also be reconsidered. Once causality is established, the key pharmacovigilance questions may shift from “is the association real?” to “how frequent is it?”, “who is at greatest risk?”, “what predicts severe outcomes?” or “are the risk-minimisation measures effective?”

Removal from the safety specification

An important potential risk may be removed when the evidence no longer supports continued focused risk-management planning. This may occur when adequate data make a causal relationship sufficiently unlikely, when the original concern is explained by alternative factors, or when the available evidence and exposure make the residual uncertainty no longer important to benefit-risk.

Removal should not depend on a preselected number of negative studies, person-years of exposure or a statistical confidence interval. The evidence required depends on the original concern and the ability of available data to test it.

The rationale should be reconstructable: what was previously suspected, what new evidence changed the interpretation and why focused RMP management is no longer proportionate.

Special Situations

Class effects

A class effect can provide a legitimate scientific basis for an important potential risk, particularly when a new medicinal product shares the pharmacological mechanism thought to produce the adverse outcome. But class membership alone is not enough.

The assessment should examine whether the relevant structural, pharmacodynamic, pharmacokinetic or biological features are shared to an extent that makes extrapolation plausible. Product-specific data that contradict the class hypothesis should also be considered.

Non-clinical findings

A non-clinical finding can support a potential risk when there is a credible pathway to human relevance. Species-specific toxicity, exposure margins, mechanism and absence or presence of corresponding clinical observations all matter.

The RMP should avoid translating every toxicology finding into a human safety concern. The question is whether the finding supports a clinically meaningful suspected adverse outcome in humans.

Off-label use and populations not studied

GVP Module V recognises that an adverse clinical outcome may be considered a potential risk where there is a scientific rationale that it could be associated with off-label use, use in an unstudied population or long-term exposure. If the potential outcome would be important if confirmed, it may warrant inclusion as an important potential risk.

This differs from merely lacking data in that situation. Absence of data alone may support missing information only when the knowledge gap is relevant to risk-management planning; it does not create a potential risk without a specific suspected adverse outcome.

Rare catastrophic outcomes

For very serious outcomes, limited evidence may still justify close attention when the event is unusual, temporally coherent and mechanistically plausible. That does not mean causality is established, nor does it create a universal “single-case rule.”

The severity of the possible harm affects the importance and precautionary response, while the causal classification must still reflect the quality of the evidence.

Practical Implementation

Good implementation depends on keeping the scientific question visible throughout the process. The RMP entry, signal documentation, pharmacovigilance plan, study records and product information should not become separate narratives that drift apart over time.

A practical operating model is to maintain a concise decision record for each important potential risk that identifies:

• the suspected adverse clinical outcome;
• the evidence supporting the causal hypothesis;
• the principal evidence against or limiting that hypothesis;
• why confirmation would be important to benefit-risk;
• the specific uncertainties that remain;
• the pharmacovigilance activities intended to address those uncertainties;
• any justified risk-minimisation actions;
• major new evidence since the previous assessment; and
• the current decision to retain, refine, reclassify or remove the concern.

This is recommended operational practice, not a prescribed EU document format. The value of such a record is traceability: an experienced reviewer should be able to reconstruct how the organisation's understanding changed over time.

Writing the RMP entry

The RMP should be concise enough to remain usable while containing enough evidence to make the reasoning intelligible. A strong description normally answers four questions in sequence:

1. What is suspected? Define the clinical outcome.
2. Why is it suspected? Summarise the evidence and its limitations.
3. Why is it important? Explain the possible benefit-risk consequence if confirmed.
4. What will resolve the uncertainty? Link the concern to proportionate pharmacovigilance activities.

An illustrative formulation might read:

Illustrative example: A possible association between Product X and acute interstitial nephritis is supported by a small number of well-documented cases with compatible timing and by a biologically plausible mechanism. The current evidence is insufficient to establish causality because of limited case numbers and important alternative causes. The outcome is considered potentially important because severe cases can result in acute kidney injury and treatment interruption. Targeted case follow-up and cumulative review are intended to better characterise the clinical pattern and assess whether further study is warranted.

This example is not template language required by EMA. Its purpose is to show the reasoning structure.

Governance and QPPV Oversight

EU pharmacovigilance legislation and GVP establish responsibilities for the marketing-authorisation holder and the EU qualified person responsible for pharmacovigilance (QPPV), but they do not prescribe a particular internal committee structure for every RMP classification decision.

Organisations commonly use multidisciplinary safety governance involving pharmacovigilance physicians, epidemiologists, clinical development, regulatory affairs, statistics and other specialists. That can be good operational practice because classification often requires several types of evidence. However, the regulatory test is not whether a named committee signed a checklist. It is whether the pharmacovigilance system can make, implement and document scientifically sound decisions and whether the QPPV can exercise effective oversight.

For QPPV oversight, useful evidence may include access to the current RMP and supporting assessments, visibility of significant changes in safety concerns, escalation routes for emerging benefit-risk issues, awareness of important pharmacovigilance activities and the ability to obtain the evidence supporting major risk-management decisions.

Potential Failure Modes

The following are illustrative failure modes, not reported inspection findings.

Inspection Considerations

A pharmacovigilance inspection may test whether the RMP safety specification reflects an effective underlying process rather than merely reviewing the wording of the document.

Relevant inspection questions could include:

• Can the organisation explain the scientific basis for each important potential risk?
• Is it clear why the concern is important enough for RMP inclusion rather than routine signal monitoring alone?
• Are the principal uncertainties explicitly defined?
• Do planned pharmacovigilance activities address those uncertainties?
• Are conclusions from signal assessments, aggregate reports, studies and regulatory evaluations reflected consistently in the RMP?
• Can significant decisions to retain, redefine, reclassify or remove concerns be reconstructed from records?
• Are risk-minimisation measures proportionate to the evidence and implemented as described?
• Does the QPPV have sufficient access and oversight to understand important changes in the safety specification?

The underlying inspection themes are scientific coherence, traceability, execution and governance. A sophisticated template cannot compensate for weak reasoning or activities that do not answer the questions stated in the RMP.

Practical Review Checklist

The following checklist is a quality aid, not a regulatory requirement.

1. Is the suspected adverse clinical outcome defined precisely?
2. Is there a genuine scientific basis for suspecting causality rather than only theoretical possibility?
3. Are the strongest supporting and opposing evidence both represented?
4. Is it clear why causality remains unconfirmed?
5. If the risk were confirmed, would it materially affect benefit-risk or the way the medicine is used safely?
6. Is the concern correctly distinguished from a signal, identified risk and missing information?
7. Are the remaining uncertainties explicit?
8. Can the planned pharmacovigilance activities realistically reduce those uncertainties?
9. Are any risk-minimisation measures proportionate to the current evidence?
10. Has important new evidence since the previous RMP been incorporated?
11. Is the rationale for retaining, reclassifying, refining or removing the concern reconstructable?
12. Is the RMP consistent with relevant signal assessments, aggregate reports, product information and regulatory commitments?

Key Takeaways

An important potential risk requires both scientific suspicion and risk-management importance. The causal relationship remains unconfirmed, but there must be a credible evidentiary basis for suspecting it; abstract possibility is not enough.

Signals and important potential risks are related but not interchangeable. Signal management investigates potential associations, whereas the RMP prioritises those suspected risks that require continuing focused risk-management planning because confirmation could affect benefit-risk.

There is no universal case-count, statistical threshold, review frequency or evidence matrix that determines inclusion, reclassification or removal. The decision is based on the totality of evidence and the clinical context.

Important potential risks usually warrant further evaluation, but the pharmacovigilance activity should follow the unanswered scientific question. Additional studies and additional risk-minimisation measures are not automatic consequences of classification.

The category is dynamic. A concern may remain potential, become identified, be refined or be removed as evidence develops. High-quality lifecycle management depends on scientific coherence, traceability and proportionate action.

References

1. European Medicines Agency. Guideline on good pharmacovigilance practices (GVP) Module V — Risk management systems (Rev. 2). EMA/838713/2011 Rev. 2. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-good-pharmacovigilance-practices-module-v-risk-management-systems-rev-2_en.pdf
2. European Medicines Agency. Guidance on the format of the risk management plan (RMP) in the EU — in integrated format (Rev. 2.0.1). EMA/164014/2018 Rev. 2.0.1. https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/guidance-format-risk-management-plan-rmp-eu-integrated-format-rev-201_en.pdf
3. European Medicines Agency. Risk management plans. Current EMA landing page for RMP guidance, templates and related post-authorisation guidance. https://www.ema.europa.eu/en/human-regulatory-overview/marketing-authorisation/pharmacovigilance-marketing-authorisation/risk-management/risk-management-plans
4. European Medicines Agency. Risk management plans (RMP) in post-authorisation phase: questions and answers. EMEA-H-19984/03 Rev. 118, updated 13 July 2026. https://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/risk-management-plans-rmp-post-authorisation-phase-questions-answers
5. European Union. Commission Implementing Regulation (EU) No 520/2012, consolidated version current at 12 February 2026. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:02012R0520-20260212
6. European Union. Directive 2001/83/EC on the Community code relating to medicinal products for human use, as amended.
7. European Union. Regulation (EC) No 726/2004, as amended.
8. International Council for Harmonisation. ICH E2E Pharmacovigilance Planning. https://www.ich.org/page/efficacy-guidelines

Regulatory Note

This article distinguishes binding EU requirements from GVP and EMA procedural guidance and from recommended operational practice. As of 7 September 2026, EMA's published RMP page continues to direct applicants to GVP Module V Rev. 2 and the integrated EU RMP format Rev. 2.0.1; EMA's pre-authorisation guidance, revised in October 2025, likewise states that Revision 2 of the RMP format should be used, including for generics. The current consolidated version of Commission Implementing Regulation (EU) No 520/2012 is dated 12 February 2026. Future revisions of GVP Module V or the RMP format should be checked before using this article for a regulatory submission.