A healthcare video repurposing brief should tell the editor exactly which source segment is approved, which viewer question the new clip answers, which claim and limitation must remain intact, and which action follows. It should also specify format, captions, visual evidence, prohibited edits, reviewers, and the acceptance test.
Healthcare marketing services should not send an editor a two-hour recording with “make five reels.” The number of outputs is a planning preference, not evidence that five useful clips exist.
Identify the source so nobody has to guess
Record the file ID, recording date, speaker, role, session or event, approved-use status, transcript version, and exact timecode range. Include several seconds of context before and after the candidate selection.
A source row can look like this:
**File:** PHYSIO-QA-2026-08-10-CAM-A
**Speaker:** Named clinician and approved public role
**Candidate:** 00:18:42 to 00:19:38
**Context start:** 00:18:18
**Transcript:** Review 2
**Rights record:** Internal reference only
**Excluded passage:** 00:19:10 to 00:19:17 pending clinical clarification
The details are fictional. The structure prevents an editor from using an earlier recording, an unapproved answer, or a segment whose rights are unclear.
Mark any patient question, audience member, screen, or background element that cannot appear. A timecode does not establish permission.
Define one viewer question for the derivative
Write the question in the viewer’s language. “Promote physiotherapy” is a campaign goal. “What happens during a first physiotherapy appointment?” is a question a clip can answer.
Add:
- who is likely to ask it;
- what they probably know already;
- the safe takeaway after one viewing;
- what remains unanswered;
- the page that owns the full answer.
This keeps the clip from accumulating unrelated lines. If the selected answer discusses both first-visit preparation and treatment outcomes, choose one or create two independently reviewed units. Do not squeeze both into one montage merely to use the footage.
The question should also differ from other planned derivatives. Three clips with different hooks but the same answer are distribution variants, not three content assets.
Bind the claim to its evidence and boundary
Copy the exact approved claim into the brief. Add its source, reviewer, limitation, and words that must not be removed.
For example:
**Approved meaning:** The first appointment includes a discussion of the person’s goals and an assessment appropriate to the service.
**Boundary:** The exact assessment varies; the clip must not predict diagnosis, treatment, or result.
**Source:** Approved service description and clinician review.
**Protected words:** “varies” and “appropriate to the service.”
This is sample wording and requires real approval before use. The editor should understand why the boundary matters. Otherwise a fast cut may remove the sentence that keeps the claim accurate.
If the source answer is wrong or outdated, repair and re-record it. Editing should not be used to manufacture a claim the speaker did not make.
Choose an action the clip can honestly support
The action should follow the viewer’s question. A first-visit clip might lead to the complete appointment guide. A location clip might lead to directions. A service-orientation clip might lead to current appointment options.
Write the exact action label and destination URL in the brief. Check that the destination repeats the same service, location, and boundary. “Book now” is not appropriate when the page creates only a request.
Avoid stacking actions such as follow, share, comment, call, book, and download. Choose the next useful step. Secondary distribution goals belong in the publishing plan, not in the spoken clinical answer.
Give the editor constraints and room to edit
Specify what must be preserved, then let the editor solve pacing and composition.
Useful controls include:
- target duration range based on placement;
- required context before the first claim;
- sentences or qualifications that cannot be separated;
- allowed removal of filler words and restarts;
- prohibited rearrangement that changes chronology or causation;
- approved B-roll and screen sources;
- on-screen text hierarchy and safe zones;
- caption style and reviewed transcript;
- synthetic or archival disclosure where applicable;
- aspect ratios and destination-specific end cards.
Do not prescribe every cut unless the sequence itself carries clinical meaning. A frame-by-frame script can prevent an editor from fixing an awkward pause; a vague brief forces the editor to make content decisions outside their role.
Use a one-page working template
Copy this structure into the production record:
**Derivative ID and owner** Stable ID, editor, due date, status.
**Source** File, speaker, rights reference, transcript, candidate and context timecodes.
**Audience job** One viewer question, assumed knowledge, safe takeaway, remaining question.
**Approved content** Exact claim, source, limitation, protected wording, prohibited implication.
**Transformation** Duration, crop, order, B-roll, text, captions, disclosure, audio and visual rules.
**Action** Exact spoken and on-screen action, destination, expected next step.
**Review** Clinical, editorial, legal, privacy, advertising, rights, brand, and accessibility applicability.
**Acceptance** Observable conditions for meaning, sound, image, captions, disclosure, and destination.
**Version record** Source hash, edit version, final render ID, reviewer decisions.
A brief can stay one page by linking to source and rights records rather than copying them.
Compare the finished cut with its source
Play the source segment and final edit back to back. Compare the transcript word for word, then compare overall meaning. Check whether cuts change who, when, how often, or how certain the statement is.
Watch muted to inspect captions and visual implications. Listen without picture to catch audio edits and mispronunciations. Open the destination on a phone. Verify that the final export, thumbnail, title, description, and captions all point to the same answer.
Automatic captions require human review. W3C’s guidance on captions notes that captions include speech and relevant non-speech information. Add meaningful sound cues where they help understanding; do not transcribe decorative music as though it carries content.
Decide whether the clip earns publication
Use a complete acceptance statement tied to observable conditions:
The derivative uses the approved source range, answers the named first-visit question, preserves the “varies” boundary, contains reviewed captions, makes no treatment or outcome prediction, and links to the current appointment guide.
If one condition fails, repair the affected unit and rerun the complete review. Do not approve the title while the caption file remains unreviewed.
Our video repurposing work uses this brief to protect meaning while giving editors creative room. The next time you open a long recording, choose one source segment and complete the source, audience, claim, and action fields before requesting a cut.