Pharmacovigilance: Monitoring Medicine Safety After Approval
Follow medicine-safety information from an initial report through case assessment, jurisdiction-specific reporting, signal review and documented action.

On this page · 11 sections
- 1What pharmacovigilance covers
- 2Clinical safety, adverse events, signals and quality complaints
- 3US drug reporting: 21 CFR 314.80, FAERS and AEMS
- 4EU and EEA monitoring: EudraVigilance and PRAC
- 5From incoming information to a reviewed safety case
- 6The evidence model behind a defensible case
- 7Assessing signals without mistaking report counts for risk
- 8Worked cases: duplicates, follow-up and a lot-related concern
- 9Integrations, acknowledgements and exception recovery
- 10Failure modes and the evidence that controls are working
- 11Implementation and review checklist
How does Pharmacovigilance apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What pharmacovigilance covers
Pharmacovigilance is the ongoing collection, assessment and use of medicine-safety information to identify, understand and reduce medicine-related harm. Common aliases are PV and drug safety monitoring. The term covers safety work across medicine development and marketed use; this article concentrates on post-approval operations rather than defining pharmacovigilance as exclusively post-approval.
For human medicines in the EU, EMA's pharmacovigilance overview explains that use after authorisation may involve large patient populations, long periods of treatment and other medicines. This is the practical setting for ongoing safety monitoring: understanding a medicine does not stop when it reaches the market.
Operationally, PV commonly connects incoming safety information with clinical assessment, follow-up, reporting decisions and action to manage risk. Its purpose is not simply to accumulate reports. Teams need to turn information into questions that can be assessed and decisions that can be explained. The case-handling and signal-assessment sections below describe those activities separately.
The scope here is medicines. The US drug provisions and EU medicine-safety overview discussed below are not a common reporting rule for every product made at a regulated site. Do not extend this article's medicine-specific discussion to food, supplements, devices or chemicals. For product classification and jurisdiction questions, start with the US and EU regulatory mapping sections below; the European Medicines Agency overview provides related agency context. Keep binding duties, agency explanations and practical workflow choices distinct when applying that mapping.
03US drug reporting: 21 CFR 314.80, FAERS and AEMS
Binding duties under the drug regulation
For US human drugs under 21 CFR 314.80(b)), covered applicants must promptly review all adverse drug experience information obtained or received from any source, foreign or domestic. For US human drugs under that provision, persons subject to its reporting requirements must also develop written procedures for surveillance, receipt, evaluation and reporting of postmarketing adverse drug experiences to FDA. For US human drugs under the same provision, applicants must submit all followup information on adverse drug experience reports that FDA forwarded to them.
21 CFR 314.80(a)) defines a serious adverse drug experience as one "that results in any of the following outcomes: Death, a life-threatening adverse drug experience, inpatient hospitalization or prolongation of existing hospitalization, a persistent or significant disability/incapacity, or a congenital anomaly/birth defect." The same paragraph defines an unexpected adverse drug experience as "any adverse drug experience that is not listed in the current labeling for the drug product." These two definitions set the trigger used throughout the reporting duties below.
15-day Alert reports and periodic reports
Under 21 CFR 314.80(c)(1)(i)(1)(i)), "the applicant must report each adverse drug experience that is both serious and unexpected, whether foreign or domestic, as soon as possible but no later than 15 calendar days from initial receipt of the information by the applicant." Under 314.80(c)(1)(ii)(1)(ii)), "the applicant must promptly investigate all adverse drug experiences that are the subject of these postmarketing 15-day Alert reports and must submit followup reports within 15 calendar days of receipt of new information or as requested by FDA."
Under 314.80(c)(1)(iii)(1)(iii)), the 15-day Alert report duties in (c)(1)(i) and (c)(1)(ii) "also apply to any person other than the applicant whose name appears on the label of an approved drug product as a manufacturer, packer, or distributor (nonapplicant)." A label-named nonapplicant may instead meet its obligation by submitting serious reports to the applicant, in which case it "must submit, by any appropriate means, each report to the applicant within 5 calendar days of initial receipt of the information by the nonapplicant." That route is specific to a nonapplicant named on the label; it is not a general allowance for every contract manufacturer, and a contract manufacturer that is not label-named should confirm its own reporting role with the applicant rather than assume this provision applies.
Under 314.80(c)(2)(i)(2)(i)), periodic adverse drug experience reports cover "each adverse drug experience not reported under paragraph (c)(1)(i) of this section," submitted "at quarterly intervals, for 3 years from the date of approval of the application, and then at annual intervals," with each quarterly report due "within 30 days of the close of the quarter" and each annual report due "within 60 days of the anniversary date of approval of the application." Under 314.80(c)(2)(iii)(2)(iii)), "periodic reporting, except for information regarding 15-day Alert reports, does not apply to adverse drug experience information obtained from postmarketing studies ..., from reports in the scientific literature, and from foreign marketing experience."
Under 314.80(d)), a 15-day Alert report based on information in the scientific literature must be accompanied by a copy of the published article; this attachment requirement concerns that literature-based report, not every safety submission. Under 314.80(e)), "an applicant is not required to submit a 15-day Alert report under paragraph (c) of this section for an adverse drug experience obtained from a postmarketing study ... unless the applicant concludes that there is a reasonable possibility that the drug caused the adverse experience."
FDA's surveillance-system descriptions
For US drug and therapeutic biologic products, FDA describes the FDA Adverse Event Reporting System (FAERS) as a database designed to support its postmarketing safety surveillance program; FAERS is FDA's description of a surveillance system, separate from determining which reporting duties above apply to a given case.
For US FDA-regulated products, FDA states that it is implementing AEMS, the Adverse Event Monitoring System, to consolidate disparate reporting systems across product categories. Treat this as an implementation announcement, not confirmation that every legacy submission route has changed; before configuring or changing a route, consult current FDA product-specific submission instructions rather than relying on a familiar acronym or a system-renaming announcement.
A practical configuration decision records the product category, report type, responsible reporting organisation and instructions used to select the route, with a named human reviewer accountable for approving that decision. This is an implementation practice, not an additional requirement attributed to 21 CFR 314.80. For related terminology, see adverse event reporting and MedWatch.
04EU and EEA monitoring: EudraVigilance and PRAC
For EU human medicines, the European Medicines Agency (EMA), in its pharmacovigilance overview, describes responsibilities involving marketing authorisation holders, national competent authorities and EMA: each operates a pharmacovigilance system. This is EMA’s explanation of the legal framework, rather than the text of EU legislation or a product-specific submission instruction.
For medicines authorised in the European Economic Area (EEA), EMA describes EudraVigilance as a system for managing and analysing information on suspected adverse reactions. Its role therefore encompasses analysis of safety information, not simply receipt of submissions.
For EU/EEA medicine-safety monitoring, EMA explains that EudraVigilance is a single repository containing suspected adverse reaction reports from both healthcare practice and clinical trials. The repository is not described as exclusively post-authorisation, even though this article focuses on monitoring after approval.
For EU human medicines, EMA describes the role of the Pharmacovigilance Risk Assessment Committee (PRAC) as evaluating safety signals from EudraVigilance and potentially recommending regulatory action. Evaluation and a possible recommendation are the functions described; regulatory action is not presented as an inevitable result of every signal.
05From incoming information to a reviewed safety case
A practical pharmacovigilance workflow assigns ownership as information moves from receipt to assessment and action. The model below describes common operating practice, not a universal legal sequence; activities may run in parallel depending on the product, jurisdiction and organisational procedures.
Receipt → source preservation → triage → duplicate assessment → medical review → reporting-route assessment → transmission reconciliation → signal referral and follow-through.
- Receive and assign. Route information from customer-facing teams, healthcare professionals, commercial partners, literature monitoring and quality investigations to an accountable safety intake owner. At every handoff, identify the receiving owner, source reference and receipt timestamp. These traceability records do not themselves establish a universal reporting clock.
- Preserve the source. The intake owner retains the original message, document or interview record and links it to the working case. Keep later clarification distinguishable from the initial account so reviewers can reconstruct what was known at each assessment.
- Triage the information. A safety case processor identifies urgent concerns, missing information and the people needed for assessment. Assign follow-up questions to named owners rather than leaving incomplete information in an unowned queue.
- Reconcile possible matches. The case processor compares patient, medicine, event and source details with existing records. Distinguish a new case from additional information about an existing case and from a possible duplicate. Preserve original reports and document the reasoning behind any linkage or merge; resemblance alone is not confirmation.
- Obtain medical review. A medical reviewer assesses the clinical account, records uncertainties and specifies questions that could change the assessment. Pass that assessment and its unresolved issues to the reporting-route reviewer, with the source reference and handoff timestamp.
- Assess the reporting route. The responsible safety reporting reviewer uses the medical assessment and applicable procedures to determine the next adverse event reporting action. Record assigned actions and the approved rationale, including decisions to seek further information.
- Reconcile transmission. The submissions owner links the transmitted case version to the recipient response and assigns rejected or unresolved submissions for recovery. Keep submission status separate from medical review status.
- Refer and follow through. Route potential safety patterns to the signal assessment owner. The case owner tracks follow-up, reassessment and outstanding actions; audit trail review can help reconstruct changes and their recorded reasons.
- Receive informationPreserve the original message and receipt metadata.
- Review the caseAssess missing details and possible duplicates.
- Assess medicallyRecord clinical reasoning and uncertainty.
- Check reporting routeConfirm product, jurisdiction and applicable instructions.
- Submit and reconcileTrack acknowledgements, errors and corrected versions.
- Evaluate signalsReview patterns alongside other relevant evidence.
- Approve and follow throughAssign accountable decisions, actions and monitoring.
06The evidence model behind a defensible case
A practical evidence model links records without collapsing their different meanings. The following is an illustrative information-design model, not a prescribed regulatory schema or a software capability.
Source reports → reconciled safety case → versioned assessments; case versions → submissions; assessed case collections → aggregate safety signals.
| Entity | Role in reconstruction |
|---|---|
| Source report | Preserves incoming wording, attachments, origin and receipt information, including subsequent correspondence. |
| Reconciled safety case | Groups related reports into a coherent patient–product–event record while retaining links to each source. |
| Versioned assessment | Records the clinical interpretation and decisions made using the information available at that point. |
| Submission | Identifies the transmitted case version, destination, payload and associated transmission outcome. |
| Aggregate safety signal | Links a safety question to the assessed case collection and other supporting evidence used in its evaluation. |
Useful case fields include a pseudonymous patient reference, product identifier, reported event wording, assessed event term, onset information, exposure information, relevant history, outcome and available lot identifiers. Collection is tailored to relevance and reporting context. Lot identifiers can connect the case to production context without replacing the clinical record. Distinguish missing information from information explicitly reported as absent: an unreported history is not a negative history.
For decision provenance, link each assessment to its reviewer, version, supporting information and rationale. Link each submission to the case version actually transmitted, not merely the latest case view. This preserves the basis of earlier decisions when follow-up changes the record.
Keeping original information, clinical interpretation and transmission outcomes distinguishable supports data integrity by design: reviewers can separate what was reported, what was concluded and what reached a destination. A later interpretation does not overwrite the source account, and a transmission outcome does not stand in for clinical review.
Safety Case
Clinical record for an assessed patient–medicine–event episode.
- case_id
- product_id
- event_term
- onset_at
- lot_number
- seriousness_assessment
Source Report
Incoming information retained separately from the reconciled case.
- source_report_id
- case_id
- receipt_at
- sender_reference
- source_channel
Case Assessment
Human review of evidence, uncertainty and reporting route.
- assessment_id
- case_id
- reviewer_id
- assessment_version
- assessment_rationale
Submission
Traceable transmission outcome for a specific case version.
- submission_id
- case_id
- case_version
- destination
- acknowledgement_status
Safety Signal
Aggregate question with an owner and documented decision.
- signal_id
- linked_case_ids
- review_owner
- evidence_summary
- decision_status
07Assessing signals without mistaking report counts for risk
Using the terminology distinctions above, signal assessment commonly follows a structured clinical review. The sequence below is practical assessment reasoning, not a jurisdiction-specific reporting rule.
- Formulate the question: specify the medicine, event and circumstances being investigated, rather than asking only whether reports increased.
- Examine case quality: assess diagnostic support, missing information and whether apparently separate reports describe the same experience.
- Consider chronology: examine treatment start, event onset, dose changes and what happened after stopping or restarting treatment, where known.
- Test alternative explanations: consider underlying illness, other medicines and clinical circumstances that could explain the event or its timing.
- Review additional evidence: consider relevant literature, trials, pharmacology and suitable observational data. Assess whether available real-world evidence addresses the question and its potential biases.
- Document uncertainty: distinguish observations from interpretations, explain conflicting findings, and decide what further evidence, specialist review or action is appropriate.
A spontaneous-report count is not an incidence estimate. Reports alone do not establish the exposure denominator, and reporting behaviour influences which experiences become reports: publicity, awareness and reporting access can change counts independently of risk. Possible denominator units include patients exposed, treatment courses or patient-time; these are alternatives, not interchangeable measures. Even with exposure data, incomplete or selective adverse event reporting can prevent a valid frequency estimate.
Aggregate review commonly asks whether event patterns differ across medicine combinations or patient subgroups, and whether apparent lot associations warrant investigation. Dashboard patterns generate questions; they do not establish causation or a manufacturing defect. Interpretation depends on clinical context, data quality and relevant exposure differences.
Completing an individual case does not necessarily conclude the broader signal assessment. A practical governance record identifies the signal owner, assessment rationale, remaining uncertainties, next steps and conditions for reconsideration, so case-handling status does not substitute for a safety conclusion.
08Worked cases: duplicates, follow-up and a lot-related concern
These hypothetical examples illustrate operating practice, not legal reporting rules. Each identifies the unit being assessed, the decision recorded and the evidence still needed.
Overlapping reports
Initial situation and assumptions: A healthcare professional and a commercial partner describe the same patient–medicine–event episode. Review confirms matching information. The unit of analysis is the episode, not the incoming report.
Assessment path: Preserve each source account and its receipt history, then link both to the consolidated case. If onset dates conflict, retain both reported dates while seeking clarification; record why a reconciled date was selected rather than silently overwriting either account.
Documented decision: Record the confirmed duplicate relationship and reviewer rationale. Aggregate review treats the linked material as the same episode while retaining the distinct source contributions. This avoids interpreting repeated descriptions as independent clinical occurrences.
Unresolved question: If the identity match cannot be confirmed, the duplicate relationship remains provisional. Keep the records distinguishable, document the missing matching evidence and make the uncertainty visible in aggregate interpretation.
Additional information changes the assessment
Initial situation and assumptions: An existing case initially describes a localized skin complaint. A later clinical account adds mucosal involvement and treatment details, changing the medical reviewer's understanding. The unit of analysis is the case version.
Assessment path: Preserve the original assessment and its evidence basis. Attach the later account to a new version, distinguish newly reported facts from reviewer interpretation, and explain why the event now warrants a different medical assessment.
Documented decision: Reassess the adverse event reporting route against the applicable product and jurisdiction framework. Record the responsible reviewer's reasoning and link any subsequent transmission to the exact reviewed version, preserving its relationship to earlier submissions.
Unresolved question: Does the later account establish the diagnosis or leave competing explanations? Document the targeted clinical follow-up needed; the example implies neither a universal deadline nor a predetermined reporting outcome.
Suspected harms with a possible lot association
Initial situation and assumptions: Incoming complaints describe adverse events and identify a medicine lot, but an association is unestablished. The unit of analysis is the case–lot association.
Assessment path: Link safety review to the quality investigation. Check product identity, lot evidence and distribution context against batch records. Verified identifiers and a relevant quality finding could strengthen the hypothesis; incorrect identifiers, incompatible distribution histories or plausible alternative clinical explanations could weaken it.
Documented decision: Record the current association assessment and supporting evidence. Keep medical assessment, quality disposition and any recall consideration as distinct decisions with accountable reviewers.
Unresolved question: Is the apparent pattern specific to the lot, or does product use and reporting context explain it? Identify the evidence needed without inferring causation from shared lot references.
09Integrations, acknowledgements and exception recovery
As an integration-design practice, map customer-contact channels and partner reports into the safety-case repository, with links to medicine and lot master data, quality investigations and reporting destinations. Document what crosses each interface, who owns it and how its origin remains traceable.
Name the authoritative source for each identifier: the contact system for its intake reference, the partner for its report reference, the designated master-data system for medicine and lot identifiers, quality for investigation references, and the safety repository for its case identifier. Maintain explicit mappings rather than assuming shared keys. This is also a practical concern in manufacturing–quality integration.
Where the destination supports them, distinguish prepared, transmitted, acknowledged, rejected and reconciled states. Prepared means ready to send; transmitted records dispatch; acknowledged records the destination’s response; rejected identifies a processing failure; reconciled means the response has been matched and any discrepancy resolved. Retain the destination reference and response with the exact case version transmitted. An acknowledgement is not, by itself, evidence of safety-case completion.
For an unavailable interface, rejected transmission, unmatched product identifier or partner correction, preserve the failed attempt or incoming correction and assign an owner. Restore connectivity, address the rejection, resolve the mapping or review the correction before controlled resubmission. Reconcile the eventual outcome against the existing case, preserving the relationship between attempts rather than creating an unlinked replacement.
10Failure modes and the evidence that controls are working
A practical pharmacovigilance review tests whether controls leave inspectable evidence, rather than merely appearing in procedures. These are illustrative operational controls, not an exhaustive regulatory checklist.
| Failure mode | Control | Review evidence |
|---|---|---|
| Unowned intake | Named intake owner and explicit handover | Ownership history showing assignment, acceptance and reassignment |
| Outdated routing instructions | Controlled routing instructions maintained through change control | Approved procedure versions and records showing which version governed routing |
| Unexplained case closure | Documented closure rationale with accountable review | Recorded decision, reviewer identity and disposition of outstanding actions |
| Unnecessary exposure of personal information | Role-appropriate access to case information | Access approvals, permission reviews and records of access or disclosure |
| Missed quality-team referral | Traceable referral with a named receiving owner | Linked quality record, receipt confirmation and recorded referral disposition |
For duplicate handling, denominator interpretation and rejected submissions, use the earlier worked-case, signal-assessment and transmission sections respectively.
Reviewers can challenge a completed status by tracing the underlying evidence: who acted, which instruction applied, what supported the decision and which actions remain unresolved. A closure entry alone does not explain whether a referral was received or a pending question was addressed. Record recurring weaknesses in a quality risk register, with an accountable owner and evidence needed to demonstrate improvement.
11Implementation and review checklist
Use this practical readiness check alongside the reporting-framework, case-handling and signal-assessment sections above. It does not certify legal compliance or replace product-specific assessment.
- Scope: identify covered medicines, jurisdictions and responsible organisations.
- Ownership: name intake, medical review and reporting owners, with cover for absences.
- Procedures: confirm current instructions and manage workflow revisions through change control.
- Handoffs: document partner responsibilities, transfer evidence and unresolved-query ownership.
- Traceability: connect source information, case versions and decisions; use an audit trail review workflow to examine changes.
- Routing: verify the current reporting route for each applicable product and jurisdiction, as covered above.
- Recovery: assign rejected or missing acknowledgements to an owner and retain evidence of resolution.
- Aggregate review: confirm how case updates reach the signal-assessment process.
Walk through a representative case, assuming its initial transmission fails. Ask the responsible teams to reconstruct receipt, assessment, routing, recovery and final disposition from records—not memory. Record gaps and test corrective actions before closing the review.
Treat software as support for the chosen workflow, not compliance certification or a replacement for medical judgement. Assign approval of safety decisions and any quality release to authorised people.
Frequently asked questions
Q.Can an incomplete report still be useful when the patient's identity or treatment dates are unclear?+
Yes. In practical safety triage, incomplete information can still identify a concern worth following up; usefulness is separate from whether a case meets applicable reporting criteria. Preserve the original account, distinguish unknown information from contradictory information, and record attempts to clarify the patient, medicine, event and treatment sequence. Avoid guessing dates or treating missing details as evidence that nothing happened. An adverse event reporting workflow should distinguish initial intake, follow-up and the reporting assessment.
Q.How should a contract manufacturer route a safety report when it is not the marketing authorisation holder?+
A practical approach is to use the agreed safety contact and escalation procedure rather than leave the report solely in the customer complaint queue. Forward the original account securely, preserve when and how it arrived, and obtain acknowledgement from the receiving safety team. Include available product and lot pedigree information without delaying transfer to complete a manufacturing investigation. Define responsibility for follow-up and reporting assessment in the relevant agreement, with a backup contact for failed handoffs.
Q.Does FDA's introduction of AEMS mean every legacy FAERS submission route has already changed?+
Not on that basis alone. For US FDA-regulated products, FDA describes AEMS as being implemented to consolidate reporting systems. A practical submission-control approach is to check current FDA instructions for the specific product, reporter and submission type before changing an established route. Distinguish MedWatch reporting instructions from system branding. Manage operational changes through change control, checking destination, access, acknowledgement handling and fallback arrangements rather than assuming a renamed platform establishes a universal cutover.
Q.Should a supplement, food or device complaint be routed using the drug-reporting provisions discussed here?+
Do not use the drug-reporting discussion as a default routing instruction. Practical triage starts with the product category, market, event and organisation's role, then checks the applicable reporting pathway. A supplement concern belongs in a category-specific assessment, such as supplement adverse event reporting; a device concern needs device-specific assessment, with quality-system context addressed separately through resources such as FDA QMSR. For food complaints, identify the relevant food-safety route. Escalate uncertain classification to regulatory specialists rather than guessing.
Q.How should the record handle disagreement between the reporter's account and the medical reviewer's assessment?+
Common documentation practice is to preserve both accounts with clear attribution. Record what the reporter described separately from the medical reviewer's interpretation, including the reasoning and information supporting that interpretation. Do not overwrite the original narrative to make it agree with the assessment. Use data integrity by design to distinguish source information from later evaluation, and an audit trail review workflow to examine consequential edits. Record unresolved differences and the follow-up needed to clarify them.
Q.When should a possible medicine-safety concern also be referred for a manufacturing-quality investigation?+
A practical trigger is information suggesting a product defect or process contribution: unusual appearance, suspected contamination, packaging damage, a strength concern or similar reports associated with a lot. Refer the concern to quality while the safety assessment continues; neither team needs to establish causation before sharing a credible concern. Link the case to the relevant batch record and complaint investigation where available. Keep clinical assessment and manufacturing findings distinguishable, with a named owner responsible for exchanging updates.
Q.Who should reconsider an earlier decision not to escalate when new clinical or literature information arrives?+
A practical governance model assigns reconsideration to the designated safety assessment owner, involving a medical reviewer and other specialists according to the concern. The person receiving new information should route it back to that owner rather than assume the earlier decision still applies. Preserve the previous rationale, identify what changed, and document the revised assessment and actions. Where the concern is tracked in a quality risk register, update its status without substituting that register for the clinical evaluation.
Q.What should management review when no new safety signal has been accepted but case-handling problems remain?+
Operational management review should examine the reliability of the safety process independently of whether a signal was accepted. Useful topics commonly include unacknowledged transfers, unresolved follow-up, duplicate handling, inconsistent assessments, ageing work, interface failures and unclear ownership. Examine recurring causes, resource constraints and whether previous corrective actions actually improved handling. Assign accountable owners and follow-up actions. The practical question is whether information reaches a defensible assessment reliably, not simply whether the current signal list has changed.
Primary sources
Want to see how Pharmacovigilance could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
