Study Conduct, Visits, and Data Collection
This domain covers the day-to-day conduct of a visit: doing only what you are trained and delegated to do, performing assessments inside the protocol window, capturing source contemporaneously, and handling anything that departs from plan. Questions usually present a common friction point - a late visit, a missed assessment, a task performed by an undelegated staff member.
Before the visit
- Confirm the participant is on the current protocol version and has current consent.
- Confirm delegated and trained staff for each procedure in the Delegation of Authority Log.
- Prepare source worksheets from the schedule of assessments; schedule within the protocol window.
During the visit
- Perform only delegated, trained procedures; the PI remains responsible for all trial conduct.
- Record source contemporaneously and legibly: who, what, when, values, units.
- Review concomitant medications and adverse events at every visit; reconcile the investigational product accountability against the kit dispensed.
- Administer questionnaires and diaries per protocol; check diary completeness while the participant is present.
Deviations and how to handle them
- A deviation is any departure from the protocol, GCP or regulation. Document what happened, why, the impact on safety and data, and the corrective/preventive action.
- Classify as major (affects safety, rights or primary endpoint validity) or minor, using the sponsor's definitions, and report within the required timeframe.
- Prospective approval is needed for planned changes; only an immediate hazard justifies acting first and reporting afterwards (ICH E6 4.5.2–4.5.4).
Missed visits and early termination
A missed visit is still documented - attempts made, reason, safety follow-up. Early termination requires an end-of-study assessment where possible, IP reconciliation, and notification per protocol.
Worked example
A Week 4 visit window is ±3 days; the participant attends on Day 35. Reasoning: the visit is outside the window and is a protocol deviation. Do not change the date to fit; record the actual date, notify the sponsor, and let the pre-specified analysis rule decide whether the data are used (many protocols flag out-of-window assessments rather than discarding them). Root cause - travel - drives a preventive action such as reminder calls and flexible scheduling.
Traps that cost marks
- Backdating or "correcting" a date to fall inside the window.
- Treating windows as elastic when the endpoint depends on timing.
- Entering data directly into the CRF when the protocol designates a separate source.
- Skipping the concomitant medication and AE review because the visit is running late.