42 stakeholder interviews conducted in 2026 explore perspectives on a machine learning model for glioblastoma prognosis. The dataset contains transcripts from patients, caregivers, and neurosurgeons discussing risks, benefits, and ethical challenges of AI-assisted surgical decision-making. The project was authored by Tristan McIntosh and harvested by QDR.
Use Cases
- Analyzing perceived risks of AI in clinical settings based on stakeholder interview themes
- Comparing ethical concerns between patients, caregivers, and neurosurgeons mentioned in the description
- Studying informational needs for AI adoption in surgery based on described stakeholder feedback
- Investigating required new skills for clinicians using AI tools as discussed in interviews
Strengths
- Includes perspectives from three distinct stakeholder groups: patients (N=13), caregivers (N=14), and neurosurgeons (N=15)
- Interviews cover four main sections: personal experiences, prognosis prediction, surgical decision-making, and advice
- The underlying ML model is described with specific performance metrics (90% and 94% accuracy)
Limitations
- Row count and column-level documentation are absent; field semantics must be inferred after download
- Data collection relies on referrals and snowball sampling, which may introduce selection bias
- The dataset is described but the actual interview transcripts or structured data are not detailed
Provenance
- Source
- Washington University in St. Louis and QDR Harvested Dataverse
- Collection Method
- One-hour interviews with patients and caregivers, and 30-minute interviews with neurosurgeons conducted via Zoom.
- Freshness
- Last updated 2026-05-04 07:10:09; freshness should be verified
- Geography
- United States (for neurosurgeon recruitment)