Pilot Data, IRB, and Authorship in a Debriefing Study

Problems.

1. Human subjects data were collected and shared before IRB review. Even if the study might qualify for exemption, exemption is a determination made by the IRB, not by the investigator.
2. The data include identifiers (names) and clinical judgment scores, creating privacy and confidentiality risk. Emailing a file with identifiers may violate institutional data security policy.
3. The consent process was misleading: students were told the survey was for course improvement, not for research. That undermines voluntary consent and may create coercion because the novice is their instructor.
4. Using the pilot data for a conference presentation would be a new use not covered by the original purpose. The data may not be usable for research.
5. Authorship was assigned before contributions were discussed, and the novice's intellectual contribution may be undervalued.

Corrective steps, in order.

1. Stop all data sharing and analysis immediately. Retrieve or delete the emailed file per IT guidance, and document the action.
2. Contact the IRB office for a preliminary consult. Disclose that data were collected before review and ask what is permitted. Do not submit a protocol that hides the timing.
3. Secure the data in an approved storage location, remove identifiers where possible, and limit access.
4. Determine whether the pilot data can be used at all. If the IRB says no, do not use them for research or publication.
5. Notify the program director and, if required, the research integrity officer. Follow institutional policy.
6. Redesign the study with proper recruitment, consent, and a plan to avoid coercion (for example, a neutral recruiter and anonymous surveys).
7. Discuss authorship after the study design is settled and contributions are documented.

Authorship. Use a written authorship agreement based on substantive contributions (conception, design, analysis, drafting, approval). The tenured professor's claim to first author is not automatic. If the novice conceived the study, drafted the protocol, and collected the data, she may be first author. The statistician should be included only if they make a substantive analytical contribution, not for being available. Discuss order before submission.

Use of pilot data. Do not use the pilot data in a conference submission unless the IRB explicitly approves that use. Even with approval, the consent was for course improvement, so the data may not be ethically usable for research. Plan to collect new data under an approved protocol. If the IRB permits use of de-identified pilot data as preliminary data, describe it transparently and do not present it as research findings.

Faculty meeting next month. The program director can present a description of the innovation and the plan for evaluation, not student data. Present the educational problem, the design of the debriefing model, and the evaluation questions. Do not show names, scores, or preliminary results from the pilot. If the IRB has not yet determined the status, say so. This protects students and the institution.

Conference deadline. The 6-week deadline is secondary to compliance. Submit an abstract only after IRB approval and with data that were collected ethically, or submit a methods or innovation abstract without student outcomes.

Redesign.

• Question and design. A quasi-experimental pre-post design with a comparison group if feasible, or a mixed-methods design. Define clinical judgment with a validated instrument and a clear primary outcome.
• Ethics. Submit to the IRB before any data collection. Request a waiver of documentation of consent only if justified. Avoid instructor-researcher coercion: a neutral colleague recruits, surveys are anonymous, and participation is not linked to grades.
• Participants. 40 students is a small sample; plan for attrition and avoid overinterpreting subgroup analyses. Consider collecting across two cohorts.
• Data security. Use the institution's approved survey and storage platform. No identifiers in email. A data management plan with access controls.
• Analysis. Pre-register the analysis plan or at least document it before data collection. Include a statistician from the design stage if the analysis is complex.
• Dissemination. Agree on authorship early with a contributions table. Report limitations, including the single-site design and the instructor effect.
• Timeline. Build in IRB review time, recruitment, data collection across a semester, analysis, and writing. The 6-week conference deadline likely requires a methods abstract, not an outcomes paper.