The Note Written in Different Ink
During a monitoring visit at a three-investigator site, the CRA compares source with EDC for subject 07-003. The 4 February source entry reads blood pressure 148/92 mmHg; EDC shows 128/92. The coordinator says it was 'obviously a keystroke error' and that she corrected the EDC herself on 20 February. The source note documenting 148/92 is written in different ink from surrounding entries and carries a small initial but no date and no explanation; it appears inserted between entries dated 5 February. The PI has already signed the case report book for this subject. Eligibility required seated SBP of 150 mmHg or less, so the subject qualified either way. A second check shows HbA1c source 8.4% versus EDC 7.4%, against an inclusion criterion of 8.0% or less.
1. Is this a routine query or something more serious? What does the CRA do first?
Read the worked answer
The blood pressure alone would be a routine transcription query. The pattern changes the classification:
- A source entry added after the visit, in different ink, initialled but undated and unexplained.
- A self-initiated EDC edit by the coordinator with no query trail.
- A second variable (HbA1c) where the source value would have made the subject ineligible.
That combination is a potential data integrity issue, not a clerical error. First actions:
- Do not alter, remove, or 'tidy' any source document. Do not instruct the coordinator to write an explanation onto the original page.
- Preserve copies of the relevant source pages and the EDC audit trail for the two fields, following site document-copy policy.
- Interview factually and separately: when was the note written, by whom, why; who made the EDC change; who authorised it. Record answers verbatim with times.
- Escalate the same day to the sponsor's clinical lead and QA, and in writing to the PI, who owns site oversight. Do not confront the coordinator as though a finding were already established, and do not conduct a fraud investigation yourself.
- Continue SDV of the remaining subjects, prioritising eligibility criteria.
2. What are the rules on correcting source and EDC, and how were they breached?
Read the worked answer
Under ICH E6, corrections to source data must not obscure the original entry, must be dated, initialled, and explained if necessary, and must be traceable. A note added later must carry the date it was actually written, never backdated to the visit date. The inserted, undated note fails on date and explanation.
EDC changes must be made through a documented process that preserves the audit trail: who changed what, when, and why. The coordinator editing 128/92 to 148/92 (or the reverse) without a query, without a reason code, and without PI review is a breach of the site's data-entry controls even if the final value turns out to be correct.
The CRA should review the EDC audit trail for both variables and for all data entered by that coordinator, and check whether the PI's signature on the case report book was obtained before or after the edits. Signing off data whose source disagrees with the database is an oversight failure attributable to the investigator.
3. What are the downstream consequences for eligibility, the subject, and the site?
Read the worked answer
Eligibility: BP is a non-issue - 148/92 satisfies the criterion either way. HbA1c is decisive: 8.4% means the subject did not meet an inclusion criterion. This is a major deviation, not a data-entry footnote.
- The PI must assess and document whether the subject should continue, after discussion with the sponsor medical monitor. A subject enrolled outside criteria may remain on safety grounds, but the decision must be reasoned and recorded, not defaulted.
- The subject's data are analysed both ways (with and without), and the deviation is reported to the IRB per its policy.
The subject: confirm they were actually informed and consented correctly, and check no safety monitoring was skipped because ineligibility went unnoticed.
The site: the finding may undermine confidence in all data entered by that individual. Expect 100% SDV of that coordinator's subjects, a for-cause audit, and possible review by the institution's research integrity office. If falsification is substantiated, the sponsor handles regulatory notification - the CRA's role is to document and escalate accurately, not to pronounce on intent.