Signal Detection
The process of identifying new or changing safety information from pharmacovigilance data that may require further investigation.
Usage Examples
- Signal detection identified a potential hepatotoxicity risk.
- Disproportionality analysis showed elevated reporting ratio.
- The signal was validated and prioritized for assessment.
What is Signal Detection?
Signal detection is the systematic process of identifying signals of potential safety concerns from pharmacovigilance data. A signal is information suggesting a new potentially causal association, or a new aspect of a known association, between an intervention and an adverse event.
Methods include disproportionality analysis (comparing observed vs. expected reporting rates), clinical review of case series, literature surveillance, and data mining of spontaneous reporting databases. Signals require validation, prioritization, and investigation.
Signal management includes detection, validation, prioritization, assessment, and action. EU GVP Module IX provides detailed guidance on signal detection and management.
Regulatory Context
This term appears most often in pharmacovigilance workflows where submission quality, regulatory evidence, and audit readiness depend on consistent language. It is commonly referenced alongside EU GVP MODULE IX, ICH E2E.
When This Matters
- Signal detection identified a potential hepatotoxicity risk.
- Disproportionality analysis showed elevated reporting ratio.
- The signal was validated and prioritized for assessment.
Common Mistakes
- Treating safety signal reviews as periodic instead of continuous.
- Not linking new enforcement letters to internal CAPA and labeling workflows.
- Using static templates for dynamic benefit-risk communication updates.
Related Regulations
Frequently Asked Questions
A statistical method comparing the observed reporting rate of a drug-event combination to the expected rate based on background reporting. High ratios may indicate potential signals.
After investigation confirms a causal association with sufficient strength of evidence. Not all signals are confirmed as risks; many are explained by confounding factors or reporting bias.
Related Terms
Related Use Cases
Related Regulatory Intelligence
Related Actions
Sources & References

