Causality is the investigator's judgement of whether the study treatment caused an event. WHO-UMC and Naranjo give structure to that judgement. Capture requires it on serious events, records it under investigator sign-off and keeps every change on the audit trail.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Assessment
Causality (study treatment)
AERELInvestigator judgement
Possibly relatedAction taken with study treatment
Outcome
Rationale
Key points
The method
Every adverse event in a trial is an untoward medical occurrence in a participant. It is not necessarily caused by the treatment. Causality assessment is the step where the investigator decides how likely it is that the investigational product caused the event. That decision drives what happens next. Under FDA's IND safety reporting rule, 21 CFR 312.32, a "suspected adverse reaction" is an adverse event for which there is a reasonable possibility that the drug caused it, and expedited reporting depends on that call together with seriousness and expectedness.
Convention in regulatory guidance is that the investigator makes the first assessment for each serious event, as the clinician who knows the participant. The sponsor reviews it and may add its own assessment, but sponsors are generally expected not to downgrade the investigator's causality without saying so, and to report both views when they differ. Check the safety reporting rules of each region where you run the trial.
Seriousness, severity and causality are three separate questions. Seriousness is a regulatory classification (death, life-threatening, hospitalisation, disability, congenital anomaly, important medical event) and is covered on the AE seriousness decision tree. Severity is intensity, often a CTCAE grade, and can be looked up in the CTCAE grade lookup. Causality is the link to the study treatment. A severe event can be unrelated, and a mild one can be clearly related.
WHO-UMC system
A standard category system from the WHO Uppsala Monitoring Centre, widely used in pharmacovigilance. Wording varies slightly between sources.
| Category | What it means |
|---|---|
| Certain | Plausible time relationship to drug intake; cannot be explained by disease or other drugs; response to withdrawal plausible; pharmacologically or phenomenologically definitive; rechallenge satisfactory where necessary |
| Probable / likely | Reasonable time relationship; unlikely to be explained by disease or other drugs; response to withdrawal clinically reasonable; rechallenge not required |
| Possible | Reasonable time relationship; could also be explained by disease or other drugs; information on withdrawal may be lacking or unclear |
| Unlikely | Time relationship makes a link improbable but not impossible; disease or other drugs provide plausible explanations |
| Conditional / unclassified | More data are needed for a proper assessment, or are being examined |
| Unassessable / unclassifiable | The report cannot be judged because information is insufficient or contradictory and cannot be supplemented or verified |
Naranjo algorithm
Ten questions, each answered yes, no or do not know, scored and summed. The total ranges from -4 to +13.
| Total score | Category |
|---|---|
| 9 or more | Definite |
| 5 to 8 | Probable |
| 1 to 4 | Possible |
| 0 or less | Doubtful |
The ten questions cover previous reports of the reaction, whether it appeared after the drug was given, improvement on stopping, recurrence on rechallenge, alternative causes, placebo response, drug levels, dose-response, similar reaction on earlier exposure, and objective confirmation. Naranjo was designed for single-drug adverse reactions in routine practice.
Choosing a scale
Neither scale is mandated for clinical trials. Many protocols use a simple five-level list (not related, unlikely, possibly, probably, definitely related) and, for reporting, treat possibly, probably and definitely as "related". That is easy for sites to apply consistently and maps cleanly to the reasonable-possibility standard. WHO-UMC adds the "unassessable" and "conditional" categories that matter when information is incomplete, and it is the common choice in pharmacovigilance.
Naranjo gives a reproducible score and is common in hospital drug-reaction reviews, but its questions assume a single suspected drug, which fits less well to trials with combination regimens, devices or vaccines, and agreement between the two scales is often only moderate in published comparisons. Whichever you choose, write it into the protocol and safety management plan, train sites on it, and use the same option list on every form.
The point of a structured scale is consistency between investigators. Give sites a short rationale field and a prompt for timing, dechallenge, alternative causes and relevant concomitant medication, so the assessment is justified rather than a click. The SAE report form template shows one layout, with causality as a dropdown and a narrative field for the reasoning.
In Capture
The AE form has a causality field. Submission checks the minimum dataset, and causality is required whenever the event is serious. The PI or a sub-investigator signs off the record under an electronic signature with a meaning statement, including under delegation.
Site records the event
Minimum dataset checked on submit
Investigator assesses causality
Required because the event is serious
Safety alert acknowledged
Recorded with user and time
PI or sub-I signs off
E-signature with meaning statement
Sponsor safety review
Safety overview across sites
Create a sample serious event, record causality and sign it off in the free sandbox. Then correct the assessment and look at the audit trail.
Workflow
Pick WHO-UMC, Naranjo or a simple related scale, define what counts as related for reporting, and write it into the safety plan.
Use the same option list for AE and SAE forms, add a rationale field and let skip logic show serious-event fields only when seriousness is yes. Start from the adverse event eCRF template.
Cover timing, dechallenge and rechallenge, alternative causes and concomitant medication, so assessments are comparable across sites.
Causality feeds reporting. The SAE reporting deadline calculator and the SUSAR reporting page cover timelines and expedited reports.
Follow-up results can change the judgement. Record the new assessment with the reason; the audit trail keeps the earlier one.
What the software does and does not do
Software should make an assessment complete, consistent and traceable. It should not make the assessment. Capture enforces that causality is present on serious events, records who assessed it and who signed it, and keeps the history. It does not calculate a WHO-UMC category or a Naranjo score for the investigator. If your sponsor uses a second, sponsor-side assessment, add it as its own field on the form so both views are stored, rather than overwriting the investigator's.
For oversight, the safety overview shows total, open and serious events across every site, and the site risk monitor flags sites with SAEs waiting for review. MedDRA coding is available with dictionaries scoped to your sponsor organisation, so coded events and causality sit in the same record. The 21 CFR Part 11 compliant EDC page explains the electronic signature and audit controls behind sign-off.
It is the investigator's judgement of whether the study treatment caused an adverse event, based on timing, response to stopping or restarting the treatment, alternative explanations and pharmacology. It determines whether an event counts as a suspected adverse reaction.
Certain, probable or likely, possible, unlikely, conditional or unclassified, and unassessable or unclassifiable. They are defined by time relationship, alternative causes, response to withdrawal and rechallenge.
Ten questions are scored and summed to a total from -4 to +13. A score of 9 or more is definite, 5 to 8 probable, 1 to 4 possible, and 0 or less doubtful.
No. Capture records the investigator's assessment, requires it on serious events, collects sign-off and audits changes. The judgement stays with the investigator.
On submission, the AE term, onset date, seriousness and severity or CTCAE grade are required, and causality is required whenever the event is serious. Many protocols collect it for every event, so build the form to match your protocol.
The change is recorded in the audit trail with the earlier value, the new value, who changed it and the reason, so the history is visible and nothing is overwritten.
Yes. The free sandbox includes AE and SAE forms with sample data, with no credit card and no time limit. You pay only once you go live with real participants.
Keep exploring
Adverse event reporting software
Minimum dataset, SAE alerts and sign-off.
AE seriousness decision tree
Serious or not, step by step.
SUSAR reporting software
Expedited reporting of serious unexpected reactions.
SAE report form template
A site-completed SAE eCRF.
Adverse event eCRF template
The AE log with causality and outcome.
Build the AE and SAE forms, record causality and sign off a test event in the free sandbox. No credit card, no time limit.