KONSİLShort answer: YES. Never meeting the patient does not mean never using the patient's data. Retrospective chart reviews need approval from a non-interventional / minimal-risk board (often expedited review), plus institutional permission from whichever unit owns the records. Nearly every journal asks for the approval number at submission — "it was retrospective, so we skipped it" is a desk rejection.
Decision tree: which path does your study take?
Why does even a "chart review" need review?
- Ethics: the Declaration of Helsinki treats research on identifiable human data as human research — data is an extension of the person.
- Law: health data is special-category personal data (GDPR and national equivalents); the legal basis and safeguards must be written down in the application.
- Publication: ICMJE-following journals require the approval; boards cannot approve retroactively, and journals compare the approval date against your data-collection dates.
What boards specifically ask of retrospective applications
| Field | The precision expected |
|---|---|
| Date range | "Records from Jan 2020 – Dec 2024" — never open-ended |
| Estimated record count | "~850 patient charts" — get a rough count from the archive first |
| Consent-waiver justification | Why individual consent is impracticable + how the data will be protected |
| Data-protection section | Who de-identifies, where data lives, retention period, who holds the key list |
| Data-owner permission | Written permission from the unit that owns the archive/PACS/lab system |
The consent waiver: requested, never assumed
Boards typically grant a waiver when three things hold together: the research cannot practicably be done with individual consent; there is no added risk to the subjects; and the data is processed de-identified. Argue all three separately — the single sentence "consent will not be obtained because the study is retrospective" is not a justification, it is a revision request.
FAQ
Public anonymized dataset — still need approval?
Many institutions issue an exemption determination instead of full review. The key point: the board makes that determination, not you — and the journal will want to see it in writing.
Can I reuse data from my previously approved study for a new question?
Not if the original approval's scope doesn't cover it — you need an amendment or a new application. If the original consent form included future-use language, this gets much easier.
What about case reports?
Most institutions require the patient's written publication consent rather than board review for a single case — and journals demand that consent. A case series (institution-dependent, often 3+) triggers board review.
Sources and standards
- WMA Declaration of Helsinki — identifiable human data and material
- ICMJE Recommendations — ethics approval and consent at submission
- GDPR special-category data provisions and national equivalents
- Institutional non-interventional / expedited review board guidelines
- Journal author guidelines — "ethics approval & consent to participate" sections
Let KONSİL draft your retrospective file
Describe your idea and KONSİL classifies the study, names the right board and drafts the 17-section application — explicitly marking the retrospective-specific fields (date range, record count, waiver justification) as "to be completed" instead of inventing them.
See the ethics file in the sample report → Apply for the closed beta →