Thursday, June 25, 2026
🇺🇸 FDA

Pharmacovigilance · WHO-UMC · ADR Case Review

WHO-UMC Causality Assessment

Free WHO-UMC causality assessment helper for adverse drug reaction categories. Map temporal relationship, dechallenge, rechallenge, alternatives, and data completeness to a likely ADR category—with links to seriousness, SUSAR, signal detection, and ICSR workflows.

Quick Answer

The WHO-UMC causality assessment system classifies whether a suspected medicine caused an adverse event using six categories: Certain, Probable/Likely, Possible, Unlikely, Conditional/Unclassified, and Unassessable/Unclassifiable. Assessment weighs temporal relationship, alternative explanations, dechallenge, rechallenge when ethical, pharmacological plausibility, and data completeness—not a numeric score. This free interactive helper supports ICSR medical review and pharmacovigilance documentation; final category assignment requires qualified medical judgment and company SOPs.

WHO-UMC causality categories (ordinal confidence)

Certain → Probable/Likely → Possible → Unlikely · Conditional/Unclassified · Unassessable/Unclassifiable

Qualitative global introspection—not a numeric score. All category criteria should be reasonably complied with per the official WHO-UMC document.

Interactive helper

Assess Likely WHO-UMC Category

Answer each field from the case narrative, source documents, medical history, concomitant medication record, and follow-up data. The output is a category helper, not a final regulatory or medical determination.

Case chronology

Consider start date, dose changes, latency, expected onset window, and event resolution timing.

Alternative explanations

Review underlying disease, comorbidities, concomitant medicines, lab artifacts, procedures, and infection.

Withdrawal and rechallenge
Clinical evidence and data quality
Definitive pharmacological or phenomenological event?

How to Use This Helper

1
Extract dates, dose history, AE onset, seriousness, outcome, and follow-up from the source record or trial database.
2
Check whether event timing is pharmacologically and clinically plausible for each suspect product.
3
Assess non-drug causes before assigning a higher-confidence category; document competing diagnoses and concomitant medicines.
4
Record dechallenge and rechallenge only when clinically meaningful and ethically available—rechallenge is not required for Probable/Likely.
5
Use the generated rationale as a reviewer prompt, then finalize according to SOP, E2B(R3) relatedness fields, and medical judgment.

WHO-UMC Causality Categories

Category Core WHO-UMC decision logic
Certain Plausible time relationship; cannot be explained by disease or other drugs; plausible withdrawal response; definitive pharmacological or phenomenological event; rechallenge satisfactory if necessary.
Probable/Likely Reasonable time relationship; unlikely to be attributed to disease or other drugs; clinically reasonable response to withdrawal; rechallenge not required.
Possible Reasonable time relationship, but could also be explained by disease or other medicines, or withdrawal information is unclear or absent.
Unlikely Temporal relationship makes a causal association improbable, or another drug, disease, procedure, or condition provides a more likely explanation.
Conditional/Unclassified Event needs more data for proper assessment, or additional information is under examination.
Unassessable/Unclassifiable Report suggests an adverse reaction but cannot be judged because information is insufficient, contradictory, or cannot be supplemented or verified.

WHO-UMC notes that few adverse reactions are truly Certain or Unlikely; most cases fall between Possible and Probable/Likely depending on chronology, alternative causes, and withdrawal information quality.

WHO-UMC vs Naranjo vs Bradford Hill

WHO-UMC

Qualitative six-category scale for individual case review. No numeric score. Standard in global spontaneous reporting and many ICSR workflows.

Naranjo algorithm

Weighted questionnaire producing a numeric score mapped to definite, probable, possible, or doubtful. More rigid; useful for training but may not align one-to-one with WHO-UMC categories.

Bradford Hill

Epidemiologic causality criteria (strength, consistency, temporality, etc.) for population-level inference—not designed for single ICSR triage in pharmacovigilance databases.

PV Workflow for Pharma Professionals

Causality assessment sits after case intake and seriousness triage in individual case safety report (ICSR) processing. Safety teams confirm minimum validity, code events in MedDRA, apply ICH E2A seriousness criteria, then assign WHO-UMC or company-equivalent causality before evaluating expectedness and expedited reportability.

A typical workflow: (1) intake and duplicate check; (2) Seriousness Checker for ICH E2A SAE criteria; (3) WHO-UMC causality with this helper; (4) expectedness against IB, SmPC, or label; (5) SUSAR Assessment for investigational studies; (6) E2B(R3) submission with drug–reaction relatedness fields; (7) aggregate review via Signal Detection when disproportionality screening follows case-level work.

Case-level causality is distinct from signal validation: PRR/ROR screens are hypothesis-generating and do not establish that a product caused an event in any individual report. Use ICSR Case Processing guidance for validity, follow-up, and narrative quality alongside this tool.

Limitations

Not a regulator-specific rule engine

Local reporting obligations, causality conventions, and study-specific rules can differ by authority, sponsor, and protocol.

Not a clinical diagnosis

The result depends on the accuracy of the case record and does not establish medical causation or product liability.

Not a substitute for follow-up

Limited or contradictory data should prompt targeted follow-up rather than forced certainty—Conditional/Unclassified is often appropriate during active case work.

Evidence & Sources

Frequently Asked Questions

The WHO-UMC system is a structured global introspection method for classifying the likelihood that a suspected drug caused an adverse event. It uses clinical chronology, alternative explanations, dechallenge, rechallenge when ethical, pharmacological plausibility, and data quality to place a case into one of six categories: Certain, Probable/Likely, Possible, Unlikely, Conditional/Unclassified, or Unassessable/Unclassifiable. It is widely used in spontaneous reporting and ICSR medical review.
No. This helper summarizes WHO-UMC logic for professional reference and documentation support only. Final causality must be assigned by qualified safety or medical reviewers according to current local SOPs, product knowledge, protocol rules, and regulator-specific requirements. Automated or checklist outputs cannot substitute for narrative review, source verification, or sign-off by the responsible medical assessor.
WHO-UMC is a qualitative global introspection scale with six ordinal categories—no numeric score. The Naranjo algorithm assigns weighted yes/no/unknown answers to produce a total score (definite, probable, possible, doubtful). Bradford Hill criteria support epidemiologic causality inference at the population level, not individual ICSR triage. Studies show moderate agreement between WHO-UMC and Naranjo but category boundaries differ; organizations should use one validated method consistently in their SOP rather than mixing scales on the same case.
Certain requires compatible timing, no reasonable alternative cause, supportive dechallenge where applicable, a definitive pharmacological or phenomenological event, and satisfactory rechallenge when rechallenge is necessary and ethical. Probable/Likely requires compatible timing, unlikely alternative attribution, and clinically reasonable dechallenge—but rechallenge is not required. Possible still has reasonable timing but competing disease or drug explanations remain, or withdrawal information is missing or unclear. Most spontaneous reports fall in Possible or Probable/Likely because rechallenge is rarely ethical.
A Certain classification generally requires a plausible time relationship, no other likely explanation, clinically plausible improvement on withdrawal where applicable, a definitive pharmacological or phenomenological event, and satisfactory rechallenge evidence when rechallenge is necessary and ethical. In practice, Certain is rare in spontaneous pharmacovigilance because rechallenge is seldom performed and alternative causes are difficult to exclude completely.
No. Rechallenge is explicitly not required for Probable/Likely in the WHO-UMC framework. A compatible chronology, reasonable exclusion of other causes, and a clinically reasonable dechallenge response are usually sufficient. Positive rechallenge strengthens confidence toward Certain when ethically justified; negative rechallenge weighs against high-confidence causality.
In clinical trials, dechallenge may be documented through protocol-mandated drug interruption, dose reduction, or unblinded treatment decisions; rechallenge may occur only when ethically permitted by protocol and investigator judgment. In spontaneous post-marketing reports, dechallenge depends on reporter narrative and follow-up; rechallenge is rarely performed for serious events. WHO-UMC applies the same category logic in both settings, but trial cases often have richer timing and concomitant medication data, while spontaneous reports may lack withdrawal details entirely.
Missing information lowers confidence. If more information is needed before a reasonable judgment can be made, WHO-UMC uses Conditional/Unclassified. If the report is contradictory or too incomplete to evaluate even with follow-up, Unassessable/Unclassifiable is more appropriate. Targeted follow-up on dates, dose, concomitant medicines, dechallenge, and alternative diagnoses often moves a case from Conditional to Possible or Probable/Likely.
Conditional/Unclassified means the case suggests an adverse reaction but additional data are needed or already under examination—follow-up may resolve the category. Unassessable/Unclassifiable means the available information is insufficient, contradictory, or cannot be verified or supplemented despite reasonable efforts. The latter is a dead-end for structured assessment until new valid information arrives; the former is a temporary holding category during active case work.
Individual case safety reports (ICSRs) transmitted under ICH E2B(R3) include drug–reaction relatedness fields (G.k.9.i) documenting assessment method, source, and result for each suspect product. Many safety databases map WHO-UMC categories or company equivalents into these E2B fields for EudraVigilance, FDA FAERS, and regional gateways. Causality at case level is distinct from signal-level causality conclusions in aggregate safety evaluation.
Case-level WHO-UMC causality supports ICSR medical review, while disproportionality screening (PRR/ROR) in signal detection is hypothesis-generating and does not establish causality. In clinical trials, seriousness (ICH E2A) is assessed first, then causality and expectedness determine SUSAR status and 7-day or 15-day expedited reporting. Use our Seriousness Checker, SUSAR Assessment, and Signal Detection tools alongside this helper—not as substitutes for integrated medical review.
Yes. Causality should reflect the evidence available at the time of assessment and be updated when new follow-up, laboratory results, autopsy findings, or corrected concomitant medication data materially change the chronology or alternative-cause analysis. Safety databases should retain audit trails when categories change because late-arriving information is common in both trial and post-marketing pharmacovigilance.

Related Pharmacovigilance Tools

PRR Signal Detection SAE Seriousness Checker SUSAR SUSAR Assessment ICSR Case Processing PT MedDRA Lookup

Drug approvals, clinical trials, M&A, event intelligence, and company updates — weekly.

Customize topics & region
Topics
Region

No spam. Unsubscribe anytime. See our privacy policy.