Meta-analysis of transcripts
M
Marc Friso
The ability to analyse multiple patient transcripts would be very helpful in order to generate relevant FAQ / patient information sheets. It would also open up the ability for clinicians to perform qualitative research.
Heidi Team
Thanks for sharing this — really thoughtful idea. Being able to analyse themes across multiple transcripts to generate FAQs/patient info sheets (and support qualitative research) makes a lot of sense, especially for recurring conditions.
A couple quick questions so we can pass the right detail to the team:
1) What would “meta-analysis” look like for you — topic clustering, common questions, sentiment, coding/tagging, or something else?
2) How would you want to select the transcripts (date range, diagnosis/tag, clinician, clinic, manual selection), and roughly how many at a time?
3) What should the output be (draft FAQ, patient handout at a chosen reading level, research-style themes with quotes), and do you need de-identification/PHI stripping by default?
If you have an example of a patient info sheet you’d want it to produce (or a screenshot/mockup of the workflow), that’d be super helpful too.