Attribution and What You Do With It

Part of Sub-Investigator (SUBI) Training: Adverse Events and Attribution.

About 2 minutes. Free, in your language, no account needed.

Causality is a judgment you are asked to make

Was there a reasonable possibility that the study intervention caused the event? The categories vary by protocol. The question does not.

It is a clinical judgment, made with incomplete information, and it is expected of you.

You are not being asked to be certain. You are asked one thing. Can a causal link be reasonably excluded?

That threshold is deliberately low, because underreporting is the failure that harms participants.

Timing. Did it follow administration in a plausible interval?

Known profile. Is it in the investigator brochure?

Dechallenge. Did it improve when the intervention stopped?

Rechallenge. Did it return on restarting? Informative and rarely available.

Alternative explanations. Disease progression, concomitant medication, an intercurrent illness.

Another plausible cause does not exclude the intervention. Both can contribute. The question stays the same.

"Probably the infection" is often right and is not the same as "not related".

You assess causality against the study intervention without knowing the arm. That is normal and it is what the assessment is designed for.

Wanting to unblind to decide causality is usually not a valid reason to unblind.

A sponsor may query a causality assessment. Reconsidering on new information is proper. Changing because the query kept coming back is not.

If you change it, record why, on the medical reasoning.

For serious events, most protocols reserve the causality assessment to the PI. Know whether yours does before you complete the field.