Treat CPI preparation as building a decision system: classify every event on seriousness, severity, expectedness, and relatedness; determine whether the decision belongs to the PI or a delegated team member; route the notification to the correct recipient — sponsor, IRB, or internal record; and write the documentation that proves each judgment. A one-page trigger-to-action matrix plus daily vignette drills, a documentation exercise, and a readiness rubric cover the core domains without relying on memorized timelines.
PI Accountability Versus Delegation: Who Owns Each Task?
A principal investigator may delegate tasks to qualified, listed team members but retains ultimate responsibility for participant safety, protocol conduct, and data integrity. Scenarios test whether you can tell which decisions stay with the PI.
Start from the two documents that formalize accountability. An investigator agreement or similar commitment lists what the investigator personally certifies: supervising the study, ensuring informed consent, controlling the investigational product, and meeting protocol and reporting obligations. A delegation-of-authority log then assigns specific tasks — eligibility checks, safety assessments, source documentation — to named, credentialed individuals. The pair works only if the log matches what actually happens at the site.
In vignettes, the diagnostic question is who made the call, not who performed the task. A coordinator can draw labs and enter data; a medically qualified individual must make clinical eligibility judgments and causality assessments unless the protocol clearly allows otherwise. Watch for scenario cues such as an unlisted sub-investigator performing assessments, an expired credential, or a signature from someone absent from the delegation log — each signals a supervision problem the PI owns.
Serious Is Not Severe: Building an Event Classification Grid
Event questions turn on three independent judgments: seriousness (a defined threshold), severity (intensity), and expectedness or relatedness. Mixing any two produces the wrong action, so practice classifying before choosing a response.
Define each axis separately. Seriousness follows defined criteria such as death, a life-threatening experience, hospitalization, disability, or congenital anomaly; severity describes how intense a symptom is, so severe nausea that resolves without intervention is not automatically serious. Expectedness compares the event with the protocol and the reference safety information such as an investigator brochure; relatedness asks whether the investigational product plausibly caused it. One event gets scored on every axis, and each combination points to a different pathway.
Drill the grid with borderline vignettes rather than obvious cases, because classification drives everything downstream: who is notified, on what basis, and with which assessment attached. Write one sentence per vignette stating your score on each axis and the document where you recorded it. If you cannot justify expectedness without naming the correct source document — brochure, protocol, or reference safety information — that is the gap to close before moving to reporting pathways.
| Event profile | Key judgment | Commonly confused with | First documented step |
|---|---|---|---|
| Hospitalization for a suspected product reaction | Serious; assess relatedness and expectedness separately | A severe but manageable symptom | Record the assessment and clinical course in source; check protocol notification requirements |
| Severe headache, outpatient, a known effect of the product | Severe but not necessarily serious | A serious event | Document intensity and course; follow protocol adverse-event procedures |
| Expected lab abnormality crossing a protocol threshold | Protocol-defined action | An unanticipated problem | Apply the protocol's stopping or reporting rules as written |
| Missed visit after travel, no impact on safety or data | Protocol deviation | Noncompliance triggering broader review | Document cause, impact assessment, and a corrective plan |
Reporting Pathways: Sponsor, IRB, and Institutional Notifications Differ
Safety information travels on separate tracks: sponsor reporting per protocol and applicable requirements, IRB reporting per institutional policy, and internal site documentation. Scenarios reward identifying the correct recipient before the correct action.
The tracks require different judgments. Sponsor reporting generally concerns serious, product-related, and unexpected events judged against reference safety information. IRB reporting concerns unanticipated problems involving risks to participants or others, noncompliance, and consent issues, judged against participant risk rather than product causality alone. One event can be reportable to one recipient and not another, and institutional policies often add stricter local requirements. Your protocol — not memory alone — defines the operative timelines.
Notice the exam-style trap in the vignette below: expected in the brochure reduces sponsor urgency but does not erase participant risk, and serious does not automatically mean related. Practice restating each notification in its own language — a sponsor safety report emphasizes causality and expectedness; an IRB submission emphasizes risk, corrective actions, and consent implications. When a vignette supplies a protocol deadline, anchor your answer to the document rather than to a number you memorized.
- Worked scenario 1: a participant in a hypertension trial is hospitalized with chest pain three days after a dose increase; the event is listed in the brochure as a possible risk.
- Plausible mistake: filing it as a routine expected adverse event because it appears in the brochure, and deferring notification until the next monitoring visit.
- Better decision: record seriousness, suspected relatedness, and expectedness in the source with dates and clinical course; report to the sponsor per protocol timelines; separately assess whether it meets your IRB's unanticipated-problem criteria and follow that policy.
- Why it matters: expectedness answers a product-safety question for the sponsor, while the IRB asks a participant-risk question. The documentation must show both judgments were made, because one event can satisfy different reportability definitions on each track.
Consent Deviations: What the PI Must Do After the Error
A consent error triggers a sequence: assess what the participant actually knew, determine reportability, correct the consent process, and fix the root cause. Re-consenting alone, without assessment and documentation, leaves the problem open.
Train on deviation-style consent errors by writing the PI's first three actions for each before checking reportability rules: an outdated form version still in use, a signature obtained by someone not approved to consent participants, enrollment before consent was signed, or a re-consent owed after an amendment. Run this drill weekly and rotate the scenarios so the sequence — welfare first, classification second, correction third — becomes automatic instead of improvised at the moment of the error.
For practice, write the sequence as a checklist you could execute under time pressure: welfare assessment, reportability determination against institutional definitions, participant re-consent where appropriate, root-cause analysis, corrective and preventive action, and evidence that prevention worked on the next enrollment. Vignettes that include a second, later enrollment are testing whether your corrective action is durable — a strong answer shows the check that would have caught the outdated version before it was used again.
- Worked scenario 2: a coordinator enrolls a participant using the previous consent version two weeks after the IRB approved a revised one; the participant has already completed the first study visit.
- Plausible mistake: treating it as paperwork housekeeping — re-consenting quietly on the new version and moving on, with nothing more than a note to file.
- Better decision: review with the participant what they understood at enrollment, assess whether any risk changed between versions, report per institutional policy, re-consent with the current version, document the root cause, and add a version-control check before further enrollments.
- Why it matters: the classification depends on whether the information gap exposed the participant to risk. That judgment is PI-level, and the corrective action must address the process failure, not just the single form.
Documentation That Survives Scrutiny: Source, Logs, and Attributable Records
Strong records are attributable, legible, contemporaneous, original, and accurate. Practice reading a scenario for what is missing — an unsigned assessment, an undated correction, a delegation log that stopped being maintained.
Scenarios often embed the answer in the record rather than the event. A causality assessment that exists only in an email, a correction with no initial or date, or a delegation log never updated after a coordinator left is a documentation finding independent of the underlying clinical decision. Train yourself to separate what happened from what the record proves happened, because the remediation differs: one may need a corrected entry, the other a process change.
Practical exercise: draft a one-page source note for a visit that included a dose reduction for toxicity. Include the date, the objective finding, the decision, the person making it and their role, participant communication, and next steps. Then audit your note against the attributability checklist. Expected observations: a first draft typically omits who made the decision or when it was communicated, and adding those two elements turns a narrative into an accountable record.
Ethics in Practice: Vulnerability, Conflicts, and Stopping Decisions
Ethics items test applied judgment: enrolling a vulnerable participant appropriately, managing conflicts, and recognizing when participant safety requires pausing enrollment rather than continuing as written.
Vignettes place ethical principles inside operational decisions. A potential participant with limited literacy asks a friend to interpret the consent discussion; a coordinator has a family relationship with a participant; enrollment pressure mounts near a site target; a participant's lab value sits just below the discontinuation threshold. Each situation asks whether the plan protects the participant first and whether the PI's response is documented as a deliberate judgment rather than an omission, within IRB and institutional guidance.
Compare two defensible-looking answers to sharpen the distinction. Letting a participant's relative interpret a consent discussion may seem supportive, but it compromises the independence of the consent process and usually conflicts with consent requirements; arranging a professional interpreter or IRB-approved translation preserves autonomy and compliance together. Similarly, keeping a borderline participant enrolled because they want to stay ignores that eligibility criteria exist to protect participants, and any exception must run through a documented, protocol-sanctioned process — if one exists at all.
A Six-Week Sequence and Readiness Checks for CPI Preparation
Sequence the domain as decisions, not chapters: two weeks building the classification and notification matrix, two weeks on documentation and ethics scenarios, then two weeks of timed case analysis scored against a rubric.
An adaptable sequence: weeks one and two, draft the trigger-to-action matrix and drill ten short vignettes per study day, writing your first action and a justification sentence each time. Weeks three and four, rotate through documentation exercises — source notes, delegation logs, consent filing checks — plus ethics vignettes where you compare two defensible answers. Weeks five and six, complete full case sets under time limits, grade against the rubric below, and revisit any axis you misclassified.
Keep administrative facts out of your study matrix: eligibility rules, exam format, fees, and scheduling belong to the credential's issuing body, and secondary materials go stale. Confirm those details on official issuer pages before registering. SOCRA's site, which describes its own certification and educational mission for clinical research professionals, is useful field context; treat every study source the same way — verify its scope against the credential you are actually pursuing.
- Classification check: for ten mixed vignettes, score seriousness, severity, expectedness, and relatedness correctly, citing the source document for every expectedness call.
- Notification check: for each reportable event, name the recipient, the judgment that made it reportable, and the document recording the decision — without mixing up sponsor and IRB tracks.
- Accountability check: given a scenario with a delegation gap, identify which decisions stayed with the PI and the record that should have shown it.
- Self-check milestone: scoring eight of ten on each drill is a learning benchmark to guide review, not a prediction of exam performance.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.