
Documenting the investigator's medical judgment: creating traceable records of causality and expectedness assessments
Design documentation templates that capture the investigator's causality, expectedness, and seriousness assessments as traceable regulatory records, manage initial versus revised assessments without overwriting originals, and evaluate the regulatory consequences of documentation failures.
"Possibly related" -- and nothing else
A regulatory auditor opens a study file. She is reviewing safety reporting documentation for a Phase III oncology trial at a site in Nashville, Tennessee. She locates the SAE report for a participant who experienced febrile neutropenia during Cycle 4 of treatment. The investigator's causality assessment reads: "possibly related." The expectedness determination: a checkmark next to "expected." No date of assessment. No signature. No indication of which IB version the investigator used to determine expectedness. No record of what safety data were available at the time of the assessment. No written rationale -- just two words and a checkmark on a pre-printed form.
The auditor cannot determine when this assessment was made. She cannot verify that the investigator used the current Reference Safety Information. She cannot confirm that the investigator had reviewed the sponsor's IND Safety Report from six weeks earlier, which described emerging cases of treatment-related sepsis across other sites. She cannot distinguish whether this assessment represents the investigator's initial determination or a revised judgment made after follow-up data arrived. The medical judgment may have been excellent. The documentation makes it invisible.
This is the problem this lesson addresses. Not the quality of the investigator's medical reasoning -- that is the investigator's domain, and it is not ours to evaluate or influence. The problem is the documentation infrastructure that captures that reasoning as a regulatory record. When the assessment exists only as a checkmark on a form, the investigator's judgment is untraceable, unverifiable, and indefensible under audit. And the RC is the person positioned to ensure that does not happen -- not by making the assessment, but by designing the templates and workflows that make the assessment documentable.
What you will learn
By the end of this lesson, you will be able to: