A well-written ICSR narrative helps reviewers understand what happened, when it happened, and how the available evidence fits together. This article focuses on narrative writing within Individual Case Safety Reports (ICSRs), including chronology, source-data consistency, writing conventions, quality checks, templates, and the role of automation and AI.
In Brief
- ICSR narrative writing turns structured case data into a clear clinical account of an adverse event or suspected adverse drug reaction.
- A strong narrative helps pharmacovigilance and medical reviewers understand chronology, clinical context, outcome and possible alternative explanations.
- ICSR narratives are used across clinical trial cases, spontaneous reports and literature cases, with the source context shaping how the case is written.
- Common quality issues include unclear timelines, missing information, inconsistent case fields, unsupported causality statements and poor follow-up integration.
- The key takeaway is that templates, automation and AI can support narrative writing, but clinical judgement and qualified review remain central.
What is ICSR narrative writing?
An ICSR narrative is a clear, chronological account of the clinical course of an adverse drug reaction (ADR) associated with a medicinal product. It may also be referred to as a case narrative. This is different from Clinical Study Report (CSR) narratives, which summarise selected safety events at study or participant level. ICSR narratives are typically required or expected for cases that need a clinical narrative, including deaths, serious adverse events, adverse events of special interest, and suspected unexpected serious adverse events (SUSARs) in clinical trial settings. The ICSR narrative summarises key information, including patient characteristics, medical history, treatment details, progression of events, diagnosis, outcomes, and relevant laboratory data. It should also include any evidence supporting or refuting the suspected ADR. The narrative sits within an Individual Case Safety Report (ICSR), alongside structured case data and gives reviewers a concise clinical account of the event. ICSRs also have their own minimum reporting criteria and case processing requirements.
When is an ICSR narrative usually needed in pharmacovigilance?
ICSR narratives are commonly used for clinical trial cases (such as serious adverse events, adverse events of special interest, exposure during pregnancy), spontaneous reports, and literature cases. The source context affects the narrative. A clinical trial case may draw heavily on CRFs, lab data, hospital reports and/or manually reported information from site, while a spontaneous case may depend more on limited information summarised by a patient or healthcare provider, and a literature case relies on published case details. All sources need careful interpretation without adding unsupported assumptions.
How should an ICSR narrative be structured?
Narratives should be written in the third person and past tense, following the chronological order of events. The details in the narrative should align with the corresponding structured ICSR data elements and must avoid directly identifiable personal information, including names of participants, healthcare professionals, or institutions. Where applicable, autopsy or post-mortem findings should be included. Timelines, medical conditions, treatments, and outcomes should be described clearly and specifically. Jargon and ambiguous terms should be avoided. If detailed supplementary records are important to a case (e.g., an autopsy report), their availability and the relevant details therein should be mentioned in the narrative.
What information should an ICSR narrative include?
A standard case narrative should consist of the following information:
- Source of report and participant demography
- Relevant medical history: pre-existing conditions, prior similar experiences
- Concomitant medications
- Suspect drug information: dosage, route of administration, start/stop dates, timing and conditions surrounding the onset of the reaction(s)
- Adverse event details: the onset and progression of the event(s), relevant diagnostic data, lab values, imaging results, seriousness, outcome
- If outcome is fatal, relevant details
- Action taken with suspect product; dechallenge and rechallenge information, if applicable
- The original reporter’s causality assessment and any alternative etiological considerations
- Sponsor/MAH medical evaluation and comment
Follow-up reports received should be integrated into the ICSR narratives based on the information provided. This may include new clinical information, corrections to earlier details, changes in outcome, additional test results, or updated causality information. Overall, the narrative should serve as a clear, standalone medical account, incorporating clinical evaluation and consideration of alternative explanations for the adverse event.
How to keep an ICSR narrative consistent with source documents and case data
The narrative should be checked against the source documents and structured safety database fields. Source documents may include (but are not limited to) Report Forms, CIOMS or MedWatch reports, CRFs, data clarification forms, medical records, literature articles, and follow-up correspondence. Key checks include dates, seriousness, outcome, coding, causality, suspect and concomitant medication details, and whether follow-up information has been reflected consistently. Where structured ICSR standards apply, the narrative should be consistent with the relevant case data fields.
Writing conventions that improve ICSR narrative quality
Narrative writing conventions should be agreed at the start of a project where possible. These include date format, chronological wording, relevance rules for medical history and concomitant medication, use of generic or trade names, abbreviation handling, lab normal ranges, verbatim terms from the source, and concise medical phrasing, and template structure. Templates can improve consistency by giving writers a standard order and expected content fields, but they should not replace case-specific judgement. The writer or reviewer still needs to decide which information is clinically relevant, whether alternative explanations are sufficiently described, and whether the narrative accurately reflects the source data.
ICSR narrative QC checklist and common quality issues
Quality review should check whether the narrative gives a clear timeline, includes relevant clinical context, matches reported information and structured case fields, uses consistent terminology, reflects follow-up information, avoids unsupported causality statements, and has appropriate medical review where required. Common issues include incomplete or missing key information, misclassification of treatment and concomitant medications, poor organisation, speculative wording, subjective opinion, and failure to update the narrative after follow-up.
Automation and AI in ICSR narrative writing
Artificial Intelligence (AI) and automation are increasingly being used in pharmacovigilance to improve the efficiency, consistency, and quality of ICSR narratives. AI-enabled pharmacovigilance databases can extract relevant information from source documents, organise case information chronologically and generate draft narratives using predefined templates. This can reduce manual effort and allow case processors to focus more on clinical assessment and data quality. However, AI-generated narratives should not be considered a replacement for medical and pharmacovigilance expertise, as AI systems are still evolving and may misinterpret complex clinical scenarios. They may also struggle with nuanced medical judgement, incomplete source data, conflicting information, or case-specific relevance decisions.
Manual narrative writing remains useful where clinical judgement, unusual case detail, or complex chronology requires close review. Template-based automation can suit repeatable case structures where the source data are well organised. AI-assisted drafting and QC may help with summarising information, checking consistency, and identifying missing fields, but outputs need validation, governance, and qualified review, maintenance, and adaptability as source documents, templates, safety systems, and regulatory expectations change.
Conclusion
A strong ICSR narrative gives reviewers a clear, chronological and medically relevant account of the case. It should use appropriate writing conventions, integrate follow-up information, remain consistent with source documents and structured case data, and be reviewed with suitable pharmacovigilance and medical oversight.
About QVigilance
We are a specialised pharmacovigilance service provider, offering comprehensive pharmacovigilance services to sponsors of clinical trials and manufacturers of authorised medicinal products. QVigilance is adept in supporting customers establishing a compliant pharmacovigilance system to support our customers’ products and safeguard their patients.
FAQs
What errors could be identified in the narrative of ICSR?
Common issues include:
- Incomplete or missing key information
- Misinterpretation of the available data
- Lack of logical flow or poor organisation
- Using speculative language (e.g., “might be,” “could be”)
- Subjective opinions instead of factual reporting
- Failure to integrate follow-up information into narratives after
Do ICSR narratives mention participants identifiers, e.g. names or addresses?
No. Participant identifiers should not be mentioned in the narrative.
Why is chronology important in an ICSR narrative?
Chronology helps regulatory authorities understand the temporal relationship between the suspect medicinal product and the adverse event. A clear sequence of treatment initiation, event onset, investigations, interventions, outcomes and confounding factors allows regulators and medical reviewers to better assess causality.
Should AI be used to independently write and finalise an ICSR narrative?
AI can assist in drafting and reviewing ICSR narratives by extracting information and generating structured summaries. However, the final narrative should be reviewed by a qualified pharmacovigilance professional and a medical reviewer to ensure medical accuracy, completeness, and regulatory compliance.