Qualitative Analysis of Rehab Feedback with AI

Evidano is an AI-powered qualitative data analysis platform that ingests transcripts and assessment spreadsheets, aligns text and numeric signals to shared codes, and produces thematic and cross-segment analyses and exportable visuals. A June 7, 2026 co-design study with 17 patients and 15 health care practitioners found consistent UI preferences for progress bars, color coding, and combined absolute and normative values.

• Study snapshot: the co-design process had three steps and included prototype usability testing with patient SUS mean 93.6 (SD 6.4) and HCP SUS mean 80.9 (SD 8.1).
• Design tradeoff: participants wanted quick visual signals (progress bars, color codes) plus access to absolute numbers and normative context (VPIT normalized 0-100).
• Operational playbook: a compact two-week sprint can reproduce these insights by combining auto-transcription, AI-assisted coding, and one-click theme-to-metric reports.

The study ran a three-step user-centered co-design process at Kliniken Valens to identify information needs, test representations of VPIT metrics, and validate a tablet app prototype. The authors collected feedback using Likert scales, rankings, and open responses and found consistent preferences for progress bars, color coding, and combined absolute and normative values, with health care practitioners initially favoring simpler visuals and patients wanting more numeric detail.

• Methods: structured feedback sessions (30–45 minutes), mockups → Figma prototype, SUS, directed content analysis.
• Data types: conventional assessments (grouped into six domains) plus ten VPIT metrics grouped into five lay domains and normalized to 0-100.
• Design tradeoff: participants wanted both quick signals and on-demand access to absolute and normative numbers.

This quick workflow shows how to reproduce the study insights in a two-week sprint using AI-enabled qualitative analysis.

• Day 1: Import transcripts, survey spreadsheets, and assessment CSVs into Evidano; set a custom dictionary (VPIT, MDC/MCID, clinical terms).
• Day 2–3: Auto-transcribe (with PII redaction) and translate non-English responses; run initial topic extraction.
• Day 4–6: Apply or refine a codebook (UX preferences, emotional tone, data needs) using AI-assisted coding; review with a clinician coder.
• Day 7–9: Generate frequency tables, co-occurrence networks, and segment comparisons (patients vs HCPs; age; role).
• Day 10: Produce one-page executive brief and visualization pack (progress bar mockups, normative vs absolute views) for design decisions.
• Day 11–14: Iterate prototype language and UI with stakeholders and re-test a small sample.

Use AI-enabled qualitative analysis to convert co-designed neurorehab feedback into prioritized product changes and measurable implementation steps. If you study or build patient-facing feedback like RehaLink (study details at rehab.jmir.org/2026/1/e85072), apply a two-week sprint of transcript ingestion, AI-assisted coding, and theme-to-metric reporting to surface tradeoffs and quantify stakeholder divergences.

Start a pilot: Try Evidano for free to upload a sample session and see a thematic report mapped to your metrics.