Signal Detection in Pharmacovigilance: A Guide for Kenyan MAHs
Signal detection is the process of identifying new or changing patterns in adverse event data that may indicate a previously unrecognized risk associated with a medicine. It's a core, ongoing PV activity — not a one-time review — and increasingly something PPB inspectors expect to see evidence of, not just a policy document referencing it.
The statistical methods behind signal detection Established disproportionality analysis methods include:
- Proportional Reporting Ratio (PRR)
- Reporting Odds Ratio (ROR)
- Chi-Square testing with Yates's correction for small sample sizes
These methods compare the frequency of a specific adverse event for a given product against a background rate, flagging statistically meaningful deviations worth investigating further.
Why manual signal detection is hard to sustain Running these calculations meaningfully requires structured ADR data at scale — something that's straightforward in a database designed for it, and genuinely difficult to do reliably in spreadsheets once a MAH's ADR case volume grows.
What a PPB inspector is actually looking for Evidence that signals are being systematically evaluated against your product portfolio on an ongoing basis, with documented outcomes — not just that a spreadsheet of ADRs exists.
FAQ
How often should signal detection be run? Best practice is continuous or at minimum aligned with your PSUR reporting cycle, not just annually. Does Kenya require a specific signal detection methodology? PPB guidelines don't mandate one specific statistical method, but expect demonstrable, systematic evaluation consistent with international pharmacovigilance practice.