GVP Module VI: Literature Cases and ICSR Management

Introduction

Scientific literature is an important source of individual case safety information.

Patients, healthcare professionals and researchers may publish information about suspected adverse reactions without the report first entering a company's spontaneous-reporting channels. A pharmacovigilance organisation must therefore maintain an appropriate literature-monitoring process capable of identifying relevant publications, assessing whether they contain individual cases and managing reportable information within the applicable regulatory framework.

The basic workflow is:

Literature monitoring
        ↓
Screening
        ↓
Relevant publication?
        ↓
Individual case information?
        ↓
Case extraction
        ↓
Duplicate assessment
        ↓
ICSR processing
        ↓
Regulatory reporting where required
        ↓
Follow-up / signal assessment

Literature surveillance is therefore not simply a search activity. It is a pharmacovigilance process with defined inputs, decisions, outputs and quality controls.

1. Why Literature Monitoring Matters

Scientific publications can contain safety information that is not available through spontaneous reporting channels.

A publication may describe:

• an individual suspected adverse reaction;
• a series of patients;
• a case report;
• a clinical study;
• an observational study;
• a review identifying previously reported cases;
• or a safety signal.

The organisation must distinguish information that requires ICSR management from information relevant to other pharmacovigilance activities.

2. Literature Monitoring Is Broader Than Case Finding

A literature-monitoring process should support several activities:

• identification of potential ICSRs;
• identification of new safety information;
• signal detection;
• risk assessment;
• product-specific safety review;
• and awareness of emerging clinical evidence.

A search strategy designed only to find obvious case reports may therefore miss information relevant to the broader safety profile.

3. Defining the Search Scope

The search strategy should reflect the medicinal products and responsibilities covered by the organisation.

Relevant parameters can include:

• active substance;
• product name;
• relevant synonyms;
• developmental or historical names;
• combination products;
• and other identifiers needed to retrieve relevant literature.

The scope should be documented and periodically reviewed.

4. Search Strategy

A literature search should be sufficiently sensitive to identify relevant safety information while remaining manageable for systematic review.

Search strategies can use combinations of:

• product terms;
• active-substance terms;
• adverse-reaction terms;
• disease terms;
• and relevant database-specific search fields.

The precise strategy depends on the product and the literature sources being monitored.

5. Screening

Search results normally contain many publications that are not relevant to the product or pharmacovigilance question.

Screening should therefore apply defined criteria.

A practical sequence is:

Search result
     ↓
Product relevant?
     ↓
Human safety information?
     ↓
Potential individual case?
     ↓
Potentially reportable?

Screening decisions should be reproducible and appropriately documented.

6. Identifying an Individual Case

A publication can contain information about an individual patient even when it is presented in a scientific rather than spontaneous-reporting format.

The reviewer should assess whether the publication provides sufficient information to constitute an individual case under the applicable framework.

The publication should not be dismissed merely because it is a journal article rather than a conventional safety report.

7. Published Case Reports

Case reports are often the clearest literature source for individual safety cases.

They may provide detailed information about:

• patient characteristics;
• medicinal-product exposure;
• timing;
• adverse reaction;
• investigations;
• treatment;
• outcome;
• and clinical reasoning.

The reviewer should extract relevant information accurately and distinguish the author's interpretation from the underlying clinical facts.

8. Case Series

A publication may describe several patients rather than a single case.

The reviewer should determine whether individual patients can be distinguished and whether the available information supports separate ICSR processing.

A case series should not automatically become one database case simply because it appears in one publication.

9. Published Clinical Studies

Clinical studies can contain individual adverse-event information as well as aggregate safety findings.

The organisation should assess the publication in the context of the applicable pharmacovigilance processes and reporting requirements.

Not every adverse event described in a study publication should automatically be converted into an individual spontaneous ICSR.

The source, study design and applicable reporting framework matter.

10. Reviews and Secondary Literature

Systematic reviews, meta-analyses and narrative reviews can identify important safety information but may not contain sufficient original information to constitute new individual cases.

They can nevertheless be valuable for:

• identifying primary publications;
• detecting emerging safety concerns;
• understanding the broader evidence base;
• and supporting signal evaluation.

The reviewer should avoid treating a secondary citation as a new patient case without appropriate source assessment.

11. Literature and Duplicate Management

A literature case may already exist in the safety database because the same clinical event was reported through another source.

Duplicate assessment is therefore particularly important.

The process should compare the publication with existing cases using relevant patient, product, clinical and temporal information.

The literature source may add important information even when it is ultimately determined to describe an existing case.

12. Source Documentation

The organisation should maintain sufficient evidence to demonstrate how the literature source was identified, assessed and processed.

Relevant records may include:

• search date;
• database or source searched;
• search strategy;
• screening decision;
• publication citation;
• case extraction;
• duplicate assessment;
• and reporting decision.

The next chunk will cover publication authors as reporters, receipt dates, regulatory timelines, literature follow-up, duplicate cases, automation, vendor oversight and practical examples.

13. Publication Authors and Reporters

A literature publication may identify one or more authors, but the author list should not automatically be treated as equivalent to the reporter information for an ICSR.

The pharmacovigilance assessment should determine what information about the source and reporter is actually available and relevant under the applicable reporting framework.

The publication should remain traceable as the source of the safety information.

14. Receipt Date for Literature Cases

Determining when a literature case enters the pharmacovigilance system is important because regulatory reporting timelines depend on the applicable receipt rules.

The organisation should have a documented process for determining when relevant information has been identified and received for pharmacovigilance purposes.

This is particularly important where literature searches are performed by vendors or affiliates.

15. Search Frequency

The frequency of literature searches should be appropriate to the applicable regulatory requirements, product portfolio and risk profile.

The organisation should be able to demonstrate that required searches were performed according to the established schedule and that exceptions were identified and managed.

16. Search Documentation

A defensible literature-monitoring record should show what was searched and what happened to the results.

Relevant information may include:

• search date;
• database;
• search terms;
• search strategy or version;
• number of results;
• screening outcome;
• reviewer;
• and disposition of potentially relevant publications.

The exact documentation requirements depend on the organisation's procedures and applicable guidance.

17. Literature Case Extraction

When a publication contains a potential individual case, relevant clinical information should be extracted accurately.

Important information can include:

• patient characteristics;
• medicinal product;
• indication;
• dose;
• exposure dates;
• adverse reaction;
• medical history;
• concomitant medicines;
• investigations;
• treatment;
• outcome;
• and author's assessment.

The processor should distinguish reported facts from interpretations made by the publication authors.

18. Narrative Construction From Literature

The ICSR narrative should provide a coherent representation of the published case.

It should not simply reproduce the publication verbatim.

The narrative should preserve the clinically relevant sequence and identify the literature source so that a reviewer can understand where the information originated.

19. Causality in Published Cases

Authors may express an opinion about causality.

That opinion is relevant information, but it does not necessarily determine the MAH's own medical assessment.

The case should distinguish:

Author's assessment
       ↓
Source information
       ↓
MAH medical assessment

The available evidence should be assessed according to the organisation's applicable procedures.

20. Literature and Seriousness

The seriousness of a reported event should be assessed from the available clinical information and the applicable criteria.

A publication may contain enough detail to establish seriousness even when the original source did not explicitly use regulatory seriousness terminology.

The processor should therefore assess the clinical facts rather than relying only on the author's wording.

21. Literature and Expectedness

Expectedness, where relevant to the reporting decision, should be assessed using the applicable reference information and regulatory framework.

The fact that an event has appeared in the scientific literature does not by itself determine whether it is expected.

22. Duplicate Assessment With Existing Cases

Before creating a new literature case, the organisation should search its existing safety database for potential matches.

Relevant comparison points may include:

• patient characteristics;
• product;
• event;
• exposure dates;
• clinical chronology;
• hospital or treatment setting;
• and reporter or publication information.

A literature publication may describe an existing case in substantially greater detail than the original report.

23. Literature Follow-Up

Direct follow-up of a published case may not always be possible or appropriate.

The organisation should nevertheless assess whether additional information can reasonably be obtained, including through the publication authors or other available sources, where appropriate.

Follow-up decisions should be documented according to the applicable procedure.

24. Multiple Products

A publication may describe several medicinal products, including concomitant treatments.

The processor should identify the suspected product and distinguish it from concomitant medicines where the available information allows.

The case should not automatically attribute every reported reaction to every medicinal product mentioned in the publication.

25. Multiple Patients

When a publication describes several patients, the organisation should determine whether the patients can be distinguished sufficiently for individual case processing.

Where patient-level information is available, separate cases may be appropriate.

Where the publication only provides aggregate information, the applicable regulatory and pharmacovigilance process should determine the appropriate handling.

26. Literature and Signal Management

A publication can be important even when it does not generate a new ICSR.

For example, a systematic review may identify a pattern across multiple studies that is relevant to signal detection.

Literature monitoring should therefore maintain an appropriate pathway from screening to broader safety evaluation.

27. Automation

Literature surveillance can use automation to reduce the volume of manual screening.

Potential tools include:

• automated database searches;
• keyword filtering;
• relevance scoring;
• duplicate article detection;
• and machine-assisted classification.

Automation should support, not obscure, the pharmacovigilance decision-making process.

28. Human Oversight of Automated Screening

An automated system can incorrectly classify an important publication as irrelevant.

The organisation should therefore validate and monitor automated screening where it is used for regulated pharmacovigilance activities.

Controls should address:

• sensitivity;
• false-negative risk;
• changes to algorithms;
• search-term changes;
• system performance;
• and escalation of uncertain results.

29. Vendor Literature Monitoring

Literature monitoring is often outsourced.

The MAH should maintain appropriate oversight of vendor performance.

Relevant controls can include:

• search strategy approval;
• defined responsibilities;
• quality metrics;
• sample review;
• reconciliation;
• audit;
• deviation management;
• and escalation.

The use of a specialist vendor does not transfer ultimate pharmacovigilance oversight away from the MAH.

30. Reconciliation

Reconciliation should confirm that publications identified as potentially reportable have progressed through the appropriate workflow.

A useful model is:

Search results
      ↓
Relevant publications
      ↓
Potential cases
      ↓
Cases created
      ↓
Cases processed
      ↓
Cases submitted where required

Discrepancies should be investigated and resolved.

The next chunk will cover practical literature-case examples, inspection considerations, governance, metrics, final principles, References and the Regulatory Note.

31. Practical Example: Published Case Report

A journal article describes one patient who developed a suspected adverse reaction after exposure to a medicinal product. The publication contains sufficient information to understand the patient, exposure, reaction and outcome.

The literature process identifies the article, the case is extracted, and the existing safety database is searched for potential duplicates.

If no matching case exists and the applicable criteria are met, the literature case is processed through the normal ICSR workflow.

32. Practical Example: Literature Case Already in the Database

A publication describes a patient whose reaction was previously reported directly to the MAH.

The literature source adds laboratory information and a more detailed clinical course.

The publication should not automatically create a second case. The existing case should be assessed against the publication and the additional information incorporated where appropriate, with the literature source retained for traceability.

33. Practical Example: Case Series

A publication describes six patients receiving the same medicinal product. Individual patient characteristics and outcomes are provided.

The reviewer should determine whether the available information supports separate case processing rather than assuming that the six patients represent one case simply because they appear in one publication.

34. Practical Example: Systematic Review

A systematic review summarises several previously published studies and reports an association between a medicinal product and an adverse event.

The review may be important for signal management even if it does not provide sufficient new patient-level information to create an ICSR.

The reviewer should assess the underlying primary publications where appropriate rather than treating the review itself as a new patient report.

35. Practical Example: Literature Case With Multiple Products

A case report describes a patient taking several medicines who develops an adverse reaction.

The processor should identify the suspected product using the available clinical evidence and should distinguish suspected medicines from concomitant therapies.

The case should not automatically attribute the reaction to every medicine mentioned in the publication.

36. Practical Example: Author Causality Assessment

The publication authors conclude that the medicinal product was likely responsible for the reaction.

That conclusion is part of the source information and should be represented accurately where relevant. It does not prevent the MAH from conducting its own assessment using its established pharmacovigilance process.

37. Practical Example: Automated Screening

A literature-monitoring vendor uses an automated relevance classifier. The system excludes a paper because the adverse event is described using terminology not included in the classifier's training or search configuration.

A mature quality system should be capable of detecting and addressing such failures through validation, performance monitoring, periodic review and appropriate human oversight.

38. Inspection Considerations

An inspector may ask:

• What literature databases are monitored?
• How often are searches performed?
• Who approves the search strategy?
• How are search results screened?
• How are potential ICSRs identified?
• How are duplicates assessed?
• How is the literature receipt date determined?
• How are reportable cases tracked through submission?
• How is vendor performance monitored?
• How do you demonstrate that searches were completed as required?

The organisation should be able to produce objective evidence rather than relying on a description of the process.

39. Inspection Risk: Missed Publications

A significant literature-monitoring weakness is failure to identify relevant publications.

Potential causes include:

• incomplete search terms;
• inadequate database coverage;
• incorrect product synonyms;
• search-schedule failures;
• system outages;
• or ineffective automated screening.

The investigation should address the underlying control rather than treating each missed article as an isolated operator error.

40. Inspection Risk: Poor Traceability

Another weakness occurs when an organisation can show that a literature search occurred but cannot reconstruct what happened to individual results.

A defensible process should allow a reviewer to trace a publication from:

Search
  ↓
Screening decision
  ↓
Case identification
  ↓
Duplicate assessment
  ↓
Case processing
  ↓
Submission / disposition

41. Vendor Oversight

Where literature monitoring is outsourced, the MAH should retain appropriate oversight of the complete process.

Oversight should be proportionate to the risk and may include:

• qualification;
• defined responsibilities;
• search-strategy governance;
• performance metrics;
• sample review;
• reconciliation;
• deviation management;
• audit;
• and CAPA.

The MAH should understand how vendor systems and processes affect the pharmacovigilance database.

42. Metrics

Useful literature-monitoring metrics can include:

• searches completed on time;
• search-result volume;
• relevant-publication rate;
• potential-ICSR rate;
• confirmed literature-case rate;
• duplicate rate;
• time from identification to case creation;
• time from case creation to submission;
• missed-publication findings;
• and vendor performance.

Metrics should be interpreted carefully. A low number of cases does not necessarily demonstrate that the literature process is effective.

43. Governance

Literature monitoring should have defined ownership within the pharmacovigilance quality system.

Governance should cover:

• search scope;
• strategy approval;
• database selection;
• screening;
• case processing;
• quality control;
• vendor management;
• change control;
• and escalation.

The process should also connect with signal management and other safety activities.

44. What Good Looks Like

A mature literature-case process has:

• a documented search strategy;
• appropriate database coverage;
• defined search frequency;
• reproducible screening criteria;
• trained reviewers;
• accurate case extraction;
• robust duplicate detection;
• appropriate follow-up;
• controlled reporting;
• vendor oversight where applicable;
• reconciliation;
• meaningful metrics;
• and inspection-ready records.

The organisation should be able to demonstrate that literature surveillance is a functioning pharmacovigilance control rather than a periodic search exercise.

45. Final Principles

1. Literature is an important source of individual case safety information.
2. Literature monitoring is broader than identifying case reports and also supports signal and risk assessment.
3. Search scope and frequency should be defined and controlled.
4. Screening should be reproducible and appropriately documented.
5. Published case reports should be assessed for individual-case information rather than dismissed because they are scientific publications.
6. Literature cases require appropriate duplicate assessment.
7. Additional information from a publication should be incorporated into an existing case where appropriate rather than creating unnecessary duplicate cases.
8. Author causality assessments are source information and do not automatically determine the MAH's assessment.
9. Automated literature screening requires appropriate validation and human oversight.
10. Outsourcing does not remove the MAH's responsibility for appropriate oversight.
11. Literature monitoring should connect with signal management and other pharmacovigilance processes.
12. The complete literature-to-ICSR pathway should be traceable and inspection-ready.

Key Takeaways

• Literature surveillance is a core pharmacovigilance source and should be managed as a controlled process.
• A publication may contain a reportable individual case even though it did not arrive through a conventional spontaneous-reporting channel.
• Case extraction requires careful separation of reported facts, author interpretation and the MAH's own assessment.
• Duplicate assessment is essential because literature cases may describe cases already received through other sources.
• Case series, clinical studies and reviews require context-specific assessment rather than automatic conversion into ICSRs.
• Automated screening can improve efficiency but must be validated and monitored.
• Vendor literature monitoring requires effective MAH oversight.
• Inspection readiness depends on traceability from search through screening, case processing and final disposition.

References

1. European Medicines Agency. Good Pharmacovigilance Practices (GVP), Module VI — Collection, management and submission of reports of suspected adverse reactions to medicinal products. Current version should be consulted for literature monitoring and ICSR requirements.
2. European Medicines Agency. GVP Module VI — literature monitoring and related guidance. Current EMA documentation should be consulted for applicable responsibilities and requirements.
3. European Medicines Agency. GVP Module IX — Signal management. Relevant to the use of literature information in signal detection and evaluation.
4. European Medicines Agency. EudraVigilance guidance and electronic reporting requirements. Relevant when literature findings result in reportable ICSRs.
5. European Parliament and Council. Directive 2001/83/EC, as amended. EU legal framework for medicinal products for human use and pharmacovigilance.
6. European Parliament and Council. Regulation (EC) No 726/2004, as amended. Union framework for authorisation and supervision of medicinal products and relevant pharmacovigilance obligations.

Regulatory Note

This article is an educational and practical explanation of literature-case management under the EU pharmacovigilance framework. It does not replace the current GVP Module VI guidance, applicable EU legislation, EudraVigilance requirements or organisation-specific procedures.

Literature-monitoring requirements, technical expectations and regulatory responsibilities may change. Before applying this article to a live pharmacovigilance process, verify the current EMA guidance, applicable legislation, technical requirements and effective dates.

The practical examples are illustrative and are not descriptions of specific regulatory inspection cases unless an authoritative source is explicitly identified.