9.3 Causality, Expectedness and Reference Safety Information
Key Takeaways
Preserve investigator and sponsor assessments when causality opinions differ.
E2A treats reasonable causal suspicion by either reporting clinician or sponsor as relevant for ADR reporting.
A mild listed reaction does not automatically make a much more severe presentation expected.
Causality, Expectedness and Reference Safety Information
Investigators and medical monitors evaluate several clinical dimensions:
- Temporal Sequence: Did the adverse event emerge after administration of the investigational product? Did it occur within a plausible pharmacokinetic/pharmacodynamic window?
- Dechallenge: Does the adverse event subside, improve, or completely resolve when the investigational product is temporarily stopped or discontinued? (Positive dechallenge).
- Rechallenge: Does the adverse event re-emerge or worsen if the investigational product is re-introduced? (Positive rechallenge).
- Underlying Medical Condition: Can the event be fully explained by the natural history or progression of the participant's pre-existing disease?
- Concomitant Medications: Could a concomitant prescription drug, over-the-counter medicine, herbal supplement, or diagnostic procedure be the primary etiological factor?
- Known Pharmacological Mechanism: Is the event consistent with the receptor binding, metabolic pathways, or toxicology documented in nonclinical animal studies or the Investigator's Brochure?
Investigator vs. Sponsor Causality: The Non-Downgrade Rule
Both the Principal Investigator and the trial sponsor perform independent causality assessments. However, strict international harmonization rules govern how disagreements are handled:
Important
The Non-Downgrade Mandate: Under ICH GCP and international regulatory reporting standards, the sponsor may never downgrade an investigator's causality assessment to avoid expedited reporting.
- If the Principal Investigator determines that an SAE has a "reasonable possibility" of being related to the investigational product (causality = Related), the sponsor must treat the event as an Adverse Reaction for regulatory reporting purposes, even if the sponsor's medical monitor strongly believes the event is completely unrelated.
- If the Principal Investigator evaluates an event as "Unrelated," but the sponsor's medical monitor evaluates it as "Related," the sponsor must report the event as related.
- In summary: If either the investigator OR the sponsor considers the event reasonably possibly related to the IMP, the event is classified as related for expedited regulatory reporting! The sponsor may record its differing opinion in the safety report narrative, but it cannot override the investigator's assessment to suppress expedited submission.
Expectedness and the Reference Safety Information (RSI)
Once an adverse event is determined to be serious and related (a Serious Adverse Reaction, or SAR), the next determination is expectedness. Expectedness dictates whether the event requires expedited notification to health authorities.
What is the Reference Safety Information (RSI)?
The Reference Safety Information (RSI) is the official regulatory baseline document that details all known, expected adverse reactions associated with the investigational medicinal product.
- In clinical trials of investigational drugs, the RSI is located in the designated part of the applicable IB.
- For trials utilizing commercially authorized drugs within their licensed indications, the approved Summary of Product Characteristics (SmPC) or local Package Insert serves as the RSI.
- Use the applicable, version-controlled RSI and relevant requirements; do not assume one fixed IB section number.
Criteria for Determining Unexpectedness
An adverse event is categorized as unexpected if:
- It is not listed in the approved Reference Safety Information at all.
- It is listed, but the severity or specificity of the observed event is greater than what is documented in the RSI (e.g., RSI describes mild reversible skin rash; participant experiences toxic epidermal necrolysis).
- The event has a fatal or life-threatening outcome, whereas the RSI lists only non-fatal occurrences of the reaction.
- It reflects a class effect known for similar drugs, but has not yet been specifically observed or listed for the trial drug in its approved RSI.
Comprehensive Safety Classification Comparison Matrix
To synthesize these regulatory constructs, review the following comparative matrix detailing clinical scenarios and their regulatory classifications:
| Term | Abbreviation | Core Definition | Causal Relationship Required? | Regulatory Action Triggered | Realistic Clinical Scenario Example |
|---|---|---|---|---|---|
| Adverse Event | AE | Any untoward medical occurrence in a study subject; temporal association only. | No. Causality not required; can be unrelated. | Recorded on source documents and routine eCRF AE log; monitored during IMVs. | A participant trips over a garden hose, fracturing their left wrist. Recorded as a moderate AE; with seriousness assessed separately, including medically important events. |
| Serious Adverse Event | SAE | Any AE resulting in death, immediate life threat, hospitalization, disability, congenital anomaly, or medical intervention. | No. Evaluated regardless of causality (related or unrelated). | Immediate, subject to protocol-defined exceptions reporting to sponsor; recorded on SAE form. | A participant is admitted to the hospital overnight for acute bacterial pneumonia. Classified as an SAE due to inpatient hospitalization, regardless of whether pneumonia is drug-related. |
| Adverse Drug Reaction | ADR / AR | Any noxious and unintended response to an investigational product at any dose. | Yes. At least a "reasonable possibility" of causal relationship. | Recorded as an AE/ADR on eCRF; reviewed for cumulative safety signal surveillance. | A participant develops severe nausea and dry mouth 45 minutes after taking the oral investigational product, resolving upon washout. |
| Suspected Unexpected Serious Adverse Reaction | SUSAR | An adverse reaction that is simultaneously Serious, Suspected related, and Unexpected per the RSI. | Yes. Suspected causal relationship evaluated by investigator or sponsor. | Triggers mandatory expedited regulatory reporting (7 calendar days for fatal/life-threatening; 15 days for others). | A participant receiving an investigational kinase inhibitor develops acute Stevens-Johnson Syndrome requiring ICU admission; SJS is not listed in the current RSI. |
Official sources
Checked October 18, 2026: ICH E6(R3), Principles and Annex 1, ICH E2A, WMA Declaration of Helsinki, 2024.
The RSI lists mild rash, but a suspected related reaction is a serious severe skin disorder not consistent with that description. What matters?
The sponsor may erase the investigator’s causal opinion.
Wait for the annual IB review before classifying it.
Assess expectedness against the actual nature and severity in the applicable RSI.
Any mention of skin effects makes every skin disorder expected.
Sections you finish are checked off in the contents.