eduKateSG · Why Science?
Turn one safety case into structured data that another system can read, check and acknowledge
Read FDA’s October 2026 final Level 2 E2B(R3) guidance through ICSR data elements, XML messages, identifiers, terminology, acknowledgements, validation, regional rules and the AEMS transition.
A safety report may begin as a conversation, a clinic note or a laboratory result. To travel reliably between companies and regulators, that story needs a shared digital grammar that preserves who experienced what, when it happened, which product was involved and what the reporter knew.
FDA’s final October 2026 Level 2 ICH E2B(R3) guidance specifies data elements and messages for electronic transmission of individual case safety reports. FDA pairs the international implementation guide with regional technical rules, schemas, reference instances and business rules.
Did you know? Beginning 1 October 2026, FDA says postmarketing ICSRs sent through ESG NextGen must use E2B(R3). The standard is therefore both a data-science lesson and a live example of safety infrastructure.
For students, E2B(R3) shows why careful classification matters. A machine-readable message is useful only when the clinical meaning, provenance, uncertainty and update history remain understandable to people.
Chapter index
Chapters 1–9: ICSR purpose, October 2026 status, data elements, identifiers and controlled terms
- 1. An ICSR is a structured account of one safety case
- 2. The October 2026 FDA guidance is final Level 2
- 3. Data elements give each fact a defined home
- 4. Identifiers preserve case continuity
- 5. Dates create the case chronology
- 6. Controlled terminology makes categories comparable
- 7. Drug-event relationships remain many-to-many
- 8. Narratives carry clinical meaning
- 9. XML provides a machine-readable envelope
Chapters 10–17: XML exchange, acknowledgements, validation, compatibility, FDA regional rules and governance
- 10. Acknowledgements close the exchange loop
- 11. Follow-up changes the case without erasing history
- 12. Nullification is different from correction
- 13. Compatibility rules support controlled migration
- 14. FDA regional rules add local implementation detail
- 15. The 2026 transition changes operational expectations
- 16. Quality control must include the clinical story
- 17. A complete E2B(R3) receipt makes transmission auditable
CHAPTER 1 OF 17 · Understand the message
1. An ICSR is a structured account of one safety case
An individual case safety report organises information about an identifiable patient, an identifiable reporter, one or more suspected products and one or more adverse events or reactions. E2B(R3) gives systems a common way to exchange that information while retaining narrative context and follow-up.
The scientific value of an icsr is a structured account of one safety case depends on preserving meaning while information moves between systems. An individual case safety report organises information about an identifiable patient, an identifiable reporter, one or more suspected products and one or more adverse events or reactions. E2B(R3) gives systems a common way to exchange that information while retaining narrative context and follow-up. Keep this boundary visible: A complete-looking XML file does not prove that the clinical facts are complete or causal.
Try this
Build a data-lineage map for an icsr is a structured account of one safety case. Map a sample report from source documents to every transmitted field and mark unknowns explicitly. Add the owner of every transformation. Keep this limit: A complete-looking XML file does not prove that the clinical facts are complete or causal.
A strong science note separates observation from interpretation. Map a sample report from source documents to every transmitted field and mark unknowns explicitly. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Map a sample report from source documents to every transmitted field and mark unknowns explicitly. Keep this boundary visible: A complete-looking XML file does not prove that the clinical facts are complete or causal.
CHAPTER 2 OF 17 · Understand the message
2. The October 2026 FDA guidance is final Level 2
FDA identifies the combined E2B(R3) implementation guide and compatibility appendix as final Level 2 guidance dated October 2026 under docket FDA-2011-D-0720, issued by CDER and CBER. The page also provides schema files and reference examples.
In the october 2026 fda guidance is final level 2, technical validity and clinical validity are separate checks. FDA identifies the combined E2B(R3) implementation guide and compatibility appendix as final Level 2 guidance dated October 2026 under docket FDA-2011-D-0720, issued by CDER and CBER. The page also provides schema files and reference examples. Preserve this guardrail: Technical guidance should be read with current FDA regional specifications and applicable reporting law.
Try this
Test a complex case in the october 2026 fda guidance is final level 2. Create a source receipt recording title, date, status, docket, issuing centres and linked technical artefacts. Include partial dates and multiple products. Respect this boundary: Technical guidance should be read with current FDA regional specifications and applicable reporting law.
A strong science note separates observation from interpretation. Create a source receipt recording title, date, status, docket, issuing centres and linked technical artefacts. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Create a source receipt recording title, date, status, docket, issuing centres and linked technical artefacts. Keep this boundary visible: Technical guidance should be read with current FDA regional specifications and applicable reporting law.
CHAPTER 3 OF 17 · Understand the message
3. Data elements give each fact a defined home
The standard separates reporter, patient, reaction, drug, medical history, tests, narrative and assessment information into named data elements. Consistent placement reduces ambiguity and supports automated screening, deduplication and downstream review.
A trustworthy exchange for data elements give each fact a defined home leaves a trace from source fact to acknowledgement. The standard separates reporter, patient, reaction, drug, medical history, tests, narrative and assessment information into named data elements. Consistent placement reduces ambiguity and supports automated screening, deduplication and downstream review. Remember: A field should not be populated merely because a system requires a value; unsupported assumptions damage the record.
Try this
Audit one message for data elements give each fact a defined home. Trace each value to its source and distinguish known, unknown, not applicable and masked information. Separate schema, business-rule and clinical checks. Preserve this caution: A field should not be populated merely because a system requires a value; unsupported assumptions damage the record.
A strong science note separates observation from interpretation. Trace each value to its source and distinguish known, unknown, not applicable and masked information. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Trace each value to its source and distinguish known, unknown, not applicable and masked information. Keep this boundary visible: A field should not be populated merely because a system requires a value; unsupported assumptions damage the record.
Message, case and sender identifiers help receiving systems distinguish new reports, follow-up information, duplicates and transmissions. Stable identifiers allow a case to evolve without losing its history when more details arrive.
The scientific value of identifiers preserve case continuity depends on preserving meaning while information moves between systems. Message, case and sender identifiers help receiving systems distinguish new reports, follow-up information, duplicates and transmissions. Stable identifiers allow a case to evolve without losing its history when more details arrive. Keep this boundary visible: Reusing, changing or concatenating identifiers carelessly can split one case or merge unrelated people.
Try this
Build a data-lineage map for identifiers preserve case continuity. Design an identifier lifecycle covering intake, transmission, follow-up, nullification and migration. Add the owner of every transformation. Keep this limit: Reusing, changing or concatenating identifiers carelessly can split one case or merge unrelated people.
A strong science note separates observation from interpretation. Design an identifier lifecycle covering intake, transmission, follow-up, nullification and migration. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Design an identifier lifecycle covering intake, transmission, follow-up, nullification and migration. Keep this boundary visible: Reusing, changing or concatenating identifiers carelessly can split one case or merge unrelated people.
E2B(R3) represents dates for receipt, events, treatment, tests and outcomes at appropriate precision. Partial dates can communicate limited knowledge without inventing a day or month. Chronology supports seriousness, expectedness and causal review.
In dates create the case chronology, technical validity and clinical validity are separate checks. E2B(R3) represents dates for receipt, events, treatment, tests and outcomes at appropriate precision. Partial dates can communicate limited knowledge without inventing a day or month. Chronology supports seriousness, expectedness and causal review. Preserve this guardrail: Imputing precision may reverse event order or create a false exposure relationship.
Try this
Test a complex case in dates create the case chronology. Build a timeline that preserves each source date’s actual precision and timezone where relevant. Include partial dates and multiple products. Respect this boundary: Imputing precision may reverse event order or create a false exposure relationship.
A strong science note separates observation from interpretation. Build a timeline that preserves each source date’s actual precision and timezone where relevant. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Build a timeline that preserves each source date’s actual precision and timezone where relevant. Keep this boundary visible: Imputing precision may reverse event order or create a false exposure relationship.
CHAPTER 6 OF 17 · Structure the case
6. Controlled terminology makes categories comparable
Coded values and controlled vocabularies let different systems express common concepts consistently. MedDRA terminology supports reaction and medical-history coding, while product and route fields follow defined structures and regional terminology.
A trustworthy exchange for controlled terminology makes categories comparable leaves a trace from source fact to acknowledgement. Coded values and controlled vocabularies let different systems express common concepts consistently. MedDRA terminology supports reaction and medical-history coding, while product and route fields follow defined structures and regional terminology. Remember: A code can hide nuance if the verbatim report and coding rationale are lost.
Try this
Audit one message for controlled terminology makes categories comparable. Store the original wording beside the selected code and review important coding changes. Separate schema, business-rule and clinical checks. Preserve this caution: A code can hide nuance if the verbatim report and coding rationale are lost.
A strong science note separates observation from interpretation. Store the original wording beside the selected code and review important coding changes. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Store the original wording beside the selected code and review important coding changes. Keep this boundary visible: A code can hide nuance if the verbatim report and coding rationale are lost.
One case can contain several products, indications and reactions. The message structure lets each drug and event retain relevant timing, action taken, rechallenge or dechallenge information and reporter or sender assessments without flattening the case.
The scientific value of drug-event relationships remain many-to-many depends on preserving meaning while information moves between systems. One case can contain several products, indications and reactions. The message structure lets each drug and event retain relevant timing, action taken, rechallenge or dechallenge information and reporter or sender assessments without flattening the case. Keep this boundary visible: A global causality statement may conceal different relationships for different drug-event pairs.
Try this
Build a data-lineage map for drug-event relationships remain many-to-many. Create a relationship matrix for every product and reaction, including uncertainty and assessor. Add the owner of every transformation. Keep this limit: A global causality statement may conceal different relationships for different drug-event pairs.
A strong science note separates observation from interpretation. Create a relationship matrix for every product and reaction, including uncertainty and assessor. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Create a relationship matrix for every product and reaction, including uncertainty and assessor. Keep this boundary visible: A global causality statement may conceal different relationships for different drug-event pairs.
Structured fields enable computation, while the narrative explains the sequence, differential diagnoses, interventions and unresolved questions. A useful ICSR uses both so reviewers can reconstruct the case rather than reading disconnected codes.
In narratives carry clinical meaning, technical validity and clinical validity are separate checks. Structured fields enable computation, while the narrative explains the sequence, differential diagnoses, interventions and unresolved questions. A useful ICSR uses both so reviewers can reconstruct the case rather than reading disconnected codes. Preserve this guardrail: Narrative text should not contradict structured values or include unnecessary identifying information.
Try this
Test a complex case in narratives carry clinical meaning. Run a structured-to-narrative consistency review before transmission. Include partial dates and multiple products. Respect this boundary: Narrative text should not contradict structured values or include unnecessary identifying information.
A strong science note separates observation from interpretation. Run a structured-to-narrative consistency review before transmission. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Run a structured-to-narrative consistency review before transmission. Keep this boundary visible: Narrative text should not contradict structured values or include unnecessary identifying information.
The ICH ICSR message uses an XML schema that defines permitted structure and relationships. Schema validation can detect malformed messages before submission, and reference instances show expected coding patterns.
A trustworthy exchange for xml provides a machine-readable envelope leaves a trace from source fact to acknowledgement. The ICH ICSR message uses an XML schema that defines permitted structure and relationships. Schema validation can detect malformed messages before submission, and reference instances show expected coding patterns. Remember: Schema-valid means structurally acceptable, not medically correct or compliant with every business rule.
Try this
Audit one message for xml provides a machine-readable envelope. Validate syntax, schema, controlled terms, business rules and clinical coherence as separate layers. Separate schema, business-rule and clinical checks. Preserve this caution: Schema-valid means structurally acceptable, not medically correct or compliant with every business rule.
A strong science note separates observation from interpretation. Validate syntax, schema, controlled terms, business rules and clinical coherence as separate layers. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Validate syntax, schema, controlled terms, business rules and clinical coherence as separate layers. Keep this boundary visible: Schema-valid means structurally acceptable, not medically correct or compliant with every business rule.
Acknowledgement messages tell the sender whether a transmission and its cases were accepted or rejected and identify errors. They support reconciliation between the sending safety database and the regulator’s receiving system.
The scientific value of acknowledgements close the exchange loop depends on preserving meaning while information moves between systems. Acknowledgement messages tell the sender whether a transmission and its cases were accepted or rejected and identify errors. They support reconciliation between the sending safety database and the regulator’s receiving system. Keep this boundary visible: A successful gateway transport message is not necessarily acceptance of every case.
Try this
Build a data-lineage map for acknowledgements close the exchange loop. Reconcile transport receipts, message acknowledgements and case-level results with retry controls. Add the owner of every transformation. Keep this limit: A successful gateway transport message is not necessarily acceptance of every case.
A strong science note separates observation from interpretation. Reconcile transport receipts, message acknowledgements and case-level results with retry controls. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Reconcile transport receipts, message acknowledgements and case-level results with retry controls. Keep this boundary visible: A successful gateway transport message is not necessarily acceptance of every case.
CHAPTER 11 OF 17 · Test the exchange
11. Follow-up changes the case without erasing history
New laboratory results, outcome information or reporter corrections can be sent as follow-up. Version discipline preserves what changed and when, allowing reviewers to understand the cumulative case and the latest assessment.
In follow-up changes the case without erasing history, technical validity and clinical validity are separate checks. New laboratory results, outcome information or reporter corrections can be sent as follow-up. Version discipline preserves what changed and when, allowing reviewers to understand the cumulative case and the latest assessment. Preserve this guardrail: Overwriting the prior state can make regulatory clocks and decisions impossible to audit.
Try this
Test a complex case in follow-up changes the case without erasing history. Generate a before-and-after change report for every follow-up message. Include partial dates and multiple products. Respect this boundary: Overwriting the prior state can make regulatory clocks and decisions impossible to audit.
A strong science note separates observation from interpretation. Generate a before-and-after change report for every follow-up message. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Generate a before-and-after change report for every follow-up message. Keep this boundary visible: Overwriting the prior state can make regulatory clocks and decisions impossible to audit.
A nullification indicates that a previously transmitted case should no longer be considered valid, such as a report sent in error. Ordinary corrections and new information normally use follow-up processes.
A trustworthy exchange for nullification is different from correction leaves a trace from source fact to acknowledgement. A nullification indicates that a previously transmitted case should no longer be considered valid, such as a report sent in error. Ordinary corrections and new information normally use follow-up processes. Remember: Nullification should not be used simply to tidy data or replace a difficult history.
Try this
Audit one message for nullification is different from correction. Define approval and evidence requirements for follow-up, correction and nullification. Separate schema, business-rule and clinical checks. Preserve this caution: Nullification should not be used simply to tidy data or replace a difficult history.
A strong science note separates observation from interpretation. Define approval and evidence requirements for follow-up, correction and nullification. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Define approval and evidence requirements for follow-up, correction and nullification. Keep this boundary visible: Nullification should not be used simply to tidy data or replace a difficult history.
CHAPTER 13 OF 17 · Protect the meaning
13. Compatibility rules support controlled migration
The backwards and forwards compatibility appendix and mapping materials explain conversion between E2B(R2) and E2B(R3). Mapping helps migration but some richer R3 meaning may not survive a round trip to an older model.
The scientific value of compatibility rules support controlled migration depends on preserving meaning while information moves between systems. The backwards and forwards compatibility appendix and mapping materials explain conversion between E2B(R2) and E2B(R3). Mapping helps migration but some richer R3 meaning may not survive a round trip to an older model. Keep this boundary visible: Automated conversion can be syntactically successful while losing specificity or changing meaning.
Try this
Build a data-lineage map for compatibility rules support controlled migration. Test representative cases and document every lossy mapping before migration. Add the owner of every transformation. Keep this limit: Automated conversion can be syntactically successful while losing specificity or changing meaning.
A strong science note separates observation from interpretation. Test representative cases and document every lossy mapping before migration. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Test representative cases and document every lossy mapping before migration. Keep this boundary visible: Automated conversion can be syntactically successful while losing specificity or changing meaning.
CHAPTER 14 OF 17 · Protect the meaning
14. FDA regional rules add local implementation detail
FDA’s August 2024 regional implementation guide describes its submission approach, regionally controlled terminology and AEMS data elements beyond the international guide for covered drug and biological products.
In fda regional rules add local implementation detail, technical validity and clinical validity are separate checks. FDA’s August 2024 regional implementation guide describes its submission approach, regionally controlled terminology and AEMS data elements beyond the international guide for covered drug and biological products. Preserve this guardrail: International conformance alone does not guarantee acceptance by a particular regulator.
Try this
Test a complex case in fda regional rules add local implementation detail. Maintain a versioned rule matrix separating ICH core requirements from FDA regional rules. Include partial dates and multiple products. Respect this boundary: International conformance alone does not guarantee acceptance by a particular regulator.
A strong science note separates observation from interpretation. Maintain a versioned rule matrix separating ICH core requirements from FDA regional rules. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Maintain a versioned rule matrix separating ICH core requirements from FDA regional rules. Keep this boundary visible: International conformance alone does not guarantee acceptance by a particular regulator.
CHAPTER 15 OF 17 · Protect the meaning
15. The 2026 transition changes operational expectations
FDA states that from 1 October 2026, postmarketing ICSRs submitted through ESG NextGen must use E2B(R3), after accepting R2 through 30 September for organisations not yet transitioned. FDA also requires covered electronic IND safety reports in R3 from 1 April 2026.
A trustworthy exchange for the 2026 transition changes operational expectations leaves a trace from source fact to acknowledgement. FDA states that from 1 October 2026, postmarketing ICSRs submitted through ESG NextGen must use E2B(R3), after accepting R2 through 30 September for organisations not yet transitioned. FDA also requires covered electronic IND safety reports in R3 from 1 April 2026. Remember: Submission route and product context affect the applicable instructions; the Safety Reporting Portal has a different workflow.
Try this
Audit one message for the 2026 transition changes operational expectations. Record product, report type, route, standard version and effective date before choosing a channel. Separate schema, business-rule and clinical checks. Preserve this caution: Submission route and product context affect the applicable instructions; the Safety Reporting Portal has a different workflow.
A strong science note separates observation from interpretation. Record product, report type, route, standard version and effective date before choosing a channel. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Record product, report type, route, standard version and effective date before choosing a channel. Keep this boundary visible: Submission route and product context affect the applicable instructions; the Safety Reporting Portal has a different workflow.
CHAPTER 16 OF 17 · Protect the meaning
16. Quality control must include the clinical story
Automated checks can test required elements, formats, code lists and business rules. Human review still needs to assess seriousness, expectedness, chronology, duplicate potential, privacy, internal consistency and whether the report communicates the source faithfully.
The scientific value of quality control must include the clinical story depends on preserving meaning while information moves between systems. Automated checks can test required elements, formats, code lists and business rules. Human review still needs to assess seriousness, expectedness, chronology, duplicate potential, privacy, internal consistency and whether the report communicates the source faithfully. Keep this boundary visible: Passing every technical check cannot repair a misunderstood source document.
Try this
Build a data-lineage map for quality control must include the clinical story. Combine automated validation with independent medical and case-processing review. Add the owner of every transformation. Keep this limit: Passing every technical check cannot repair a misunderstood source document.
A strong science note separates observation from interpretation. Combine automated validation with independent medical and case-processing review. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Combine automated validation with independent medical and case-processing review. Keep this boundary visible: Passing every technical check cannot repair a misunderstood source document.
CHAPTER 17 OF 17 · Protect the meaning
17. A complete E2B(R3) receipt makes transmission auditable
The receipt records source documents, minimum case criteria, identifiers, standard and regional versions, data provenance, terminology, attachments, validation results, acknowledgements, follow-up status, reconciliation and unresolved issues. It connects the clinical case to the exchanged message.
In a complete e2b(r3) receipt makes transmission auditable, technical validity and clinical validity are separate checks. The receipt records source documents, minimum case criteria, identifiers, standard and regional versions, data provenance, terminology, attachments, validation results, acknowledgements, follow-up status, reconciliation and unresolved issues. It connects the clinical case to the exchanged message. Preserve this guardrail: Auditability demonstrates what was sent and accepted; it does not establish causality.
Try this
Test a complex case in a complete e2b(r3) receipt makes transmission auditable. Give the receipt to an independent reviewer and ask whether the case can be reconstructed without guesswork. Include partial dates and multiple products. Respect this boundary: Auditability demonstrates what was sent and accepted; it does not establish causality.
A strong science note separates observation from interpretation. Give the receipt to an independent reviewer and ask whether the case can be reconstructed without guesswork. Record the source, version or date, population or device context, method, result and one uncertainty. If the official record does not show a field, write “not shown” instead of filling the gap from memory.
Evidence checkpoint
Give the receipt to an independent reviewer and ask whether the case can be reconstructed without guesswork. Keep this boundary visible: Auditability demonstrates what was sent and accepted; it does not establish causality.
E2B(R3) message assurance map
| Question | Evidence to capture | Boundary to retain |
|---|---|---|
| Clinical meaning | Source facts, narrative and assessments | A valid file can still be wrong |
| Data structure | Elements, codes, identifiers and XML | Do not invent missing facts |
| Exchange control | Gateway, acknowledgements and reconciliation | Transport is not case acceptance |
| Lifecycle | Follow-up, correction, nullification and migration | Preserve history and provenance |
Science learning becomes useful when a student can show where a claim came from, why it applies and where it stops. Continue with the eduKate Science Learning Hub, compare choices in the Education Hub, and connect these habits to career pathways built from subject capabilities.
E2B(R3) turns pharmacovigilance into dependable information engineering. Its best outcome is not merely an accepted XML file, but a case whose clinical meaning survives structured exchange and remains traceable when new evidence arrives.
Primary sources and verification date
Source status checked 26 October 2026. FDA guidance describes the agency’s current thinking and does not replace applicable statutes, regulations or product-specific communication with FDA.
- FDA: ICH E2B(R3) implementation guide and compatibility appendix — Final Level 2 guidance, October 2026; docket FDA-2011-D-0720.
- FDA: Regional E2B(R3) implementation guide — Final technical specifications, August 2024; docket FDA-2016-D-1280.
- FDA: AEMS electronic submissions — Current routes and transition dates, checked 26 October 2026.