Preparing your experience...
Preparing your experience...
Capability Domain
Evaluate whether a drug caused an adverse event using the Naranjo algorithm, WHO-UMC system, and Bradford Hill criteria. Each method approaches the same question from a different analytical lens.
Causality assessment answers: did this drug cause this event? No single algorithm is definitive. NexVigilant runs all three established methods simultaneously so you can compare their verdicts and document the rationale required for regulatory submissions.
The Naranjo scale asks 10 structured questions about the case. Each answer contributes +2, +1, 0, or −1 to a running score. The total score places the case into one of four causality bands.
Definite
Score >= 9
Very strong evidence of drug causation
Probable
Score 5 – 8
Likely caused by the drug
Possible
Score 1 – 4
Drug may have contributed
Doubtful
Score <= 0
Unlikely to be drug-related
Key questions: previous exposure, improvement on dechallenge, reappearance on rechallenge, alternative causes, placebo effect, drug levels, dose-response, prior experience, objective confirmation.
The World Health Organization Uppsala Monitoring Centre system classifies causality using a structured set of criteria for plausibility, timing, and clinical pattern.
| Category | Key Criteria |
|---|---|
| Certain | Plausible time sequence, known reaction type, confirmed on rechallenge, no alternative explanation |
| Probable/Likely | Plausible time, known reaction, improved on dechallenge, alternative explanation unlikely |
| Possible | Plausible time, could be drug or disease, information on dechallenge lacking |
| Unlikely | Time sequence improbable, other explanation plausible |
| Conditional/Unclassified | More data needed for proper assessment |
| Unassessable/Unclassifiable | Insufficient information, cannot be supplemented or verified |
Bradford Hill criteria are used when evaluating population-level causation rather than individual case attribution. Particularly useful for signal confirmation and regulatory submissions.
Strength
How large is the association (e.g., PRR, ROR)?
Consistency
Replicated across different studies and populations?
Specificity
Association limited to specific drug-event pair?
Temporality
Drug exposure precedes the event?
Biological Gradient
Higher dose → stronger effect?
Plausibility
Biologically or mechanistically coherent?
Coherence
Consistent with known natural history?
Experiment
Reverses on drug removal?
Analogy
Similar drug-event pairs known?
4 Causality Methods
Naranjo scoring, WHO-UMC classification, RUCAM hepatotoxicity scale, and Bradford Hill multi-criterion analysis
FAERS Case Narratives
Real-world case data from FDA's adverse event reporting system including dechallenge and rechallenge information
3 Reference Sources
DailyMed drug labeling, PubMed case reports, and FAERS outcomes — all queried live
Data Sources
Connect your AI agent to mcp.nexvigilant.com and score any case using all three methods simultaneously.