Evidano is an AI-powered qualitative data analysis platform that accelerates thematic extraction, cross-segment comparisons, PII redaction, and exportable adaptation outputs. This post shows a reproducible, AI-enabled workflow for qualitative analysis of HIV disclosure decision aids, built from Zurashvili et al.’s 16 July 2026 NGT study in Georgia. You will learn how to move from de-identified NGT notes and short rankings (n=22 participants; providers=12, PLWH=10; recruitment Feb–Mar 2026) to a prioritized adaptation checklist and stakeholder brief without losing nuance. Practical takeaways include a 2-week runbook, comparison of provider versus patient priorities, and concrete ways to operationalize adaptation recommendations using Evidano (see Evidano). The goal is faster synthesis, defensible segment comparisons, and repeatable outputs you can share with implementers and funders.
Key Takeaways
This post explains how to convert NGT field notes and ranking scores into a prioritized adaptation plan in days using AI-enabled tools while preserving nuance and confidentiality.
- Zurashvili et al. published a short NGT study on 16 July 2026 (n=22; providers=12, PLWH=10) and the joint group prioritized adapting the Family disclosure decision aid.
- Providers in the study favored digital scalability, while people living with HIV (PLWH) preferred individual or session-based support.
- The practical workflow converts flip-chart items, notes, and rankings into a reproducible adaptation checklist within a 2-week runbook using Evidano.
Findings snapshot
| Item | Value | Source / note |
|---|---|---|
| Published | 16 July 2026 | PLoS One |
| Total participants | 22 (12 providers; 10 PLWH) | NGT sessions, Feb 9–Mar 3, 2026 |
| Preferred by providers | Digital (READY favored) | Providers ranked digital highest |
| Preferred by PLWH | Session-based (Family disclosure DA favored) | PLWH prioritized individualized sessions |
| Final consensus | Adapt Family disclosure DA for Georgia | Joint session prioritized PLWH preference |
What happened, methods & key findings
The authors ran three NGT sessions and used pre-prepared summaries and independent rankings to prioritize an adaptation, publishing their findings in PLoS One on 16 July 2026.
The authors ran three NGT sessions (providers, PLWH, joint) using pre-prepared one-page summaries of nine candidate interventions grouped into session-, paper-, and digital formats. Participants silently generated ideas, shared them round-robin, discussed, and independently ranked formats and interventions. Notes (de-identified) and ranking scores were aggregated and summarized via descriptive content analysis.
- Divergence: providers emphasized accessibility and scalability through digital formats, while PLWH emphasized trust, privacy, and peer involvement in session-based formats.
- Adaptation priorities collected from participants included peer educators, locally relevant legal and treatment content, flexible session length (individual first, group later), and guidance for disclosure to non-HIV healthcare providers.
- Limitation flagged by the authors: younger PLWH were under-represented and discussions were captured in field notes rather than verbatim transcripts.
So what for researchers, UX teams, and program leads
For qualitative researchers
Qualitative researchers should treat NGT outputs as concise but high-value data sources to preserve nuance when adapting interventions.
NGT outputs are short (flip-chart items and rank scores) but high value. To preserve nuance, extract: (a) ranked items, (b) illustrative quotations, and (c) suggested adaptations. Use cross-segment comparisons (providers versus PLWH) to surface tension points to address during adaptation.
Tip: convert flip-chart items into coded themes, then quantify frequency and rank concordance between stakeholder groups.
For UX / intervention designers
UX and intervention designers should offer a phased pathway that balances immediate individualized support with later group or digital options.
Design implication: offer a phased pathway, individual sessions immediately post-diagnosis, optional peer groups later, and a digital companion for privacy and scalability.
Test criteria: acceptability (trust and peer role), feasibility (staff time), and fidelity (core decision-aid components retained).
For policy & program teams
Policy and program teams should balance PLWH-preferred individualized care with provider concerns about scale through pragmatic hybrid models.
Programs should balance PLWH-preferred individualized care with provider concerns about scale. Hybrid models (peer-led initial session plus digital follow-ups) are a pragmatic compromise.
Measurement: track uptake, decisional conflict, and treatment adherence post-adaptation as early indicators.
Do more, faster with Evidano
From NGT notes → themes (minutes to hours)
Evidano converts de-identified flip-chart images, rapporteur notes, and ranking spreadsheets into candidate themes and frequency tables in minutes to hours.
Import de-identified flip-chart images, rapporteur notes, and ranking spreadsheets into Evidano. Use automated thematic extraction to produce candidate codes (trust, privacy, peer-support, legal info) and a frequency table showing which stakeholder raised each theme and how often.
Compare segments reliably
Evidano enables cross-segment analysis that quantifies preference divergence and surfaces phrases unique to each group.
Run cross-segment analysis (providers versus PLWH) to quantify preference divergence (for example, digital score versus session score) and surface phrases unique to each group. Evidano produces side-by-side theme counts and salient quotes you can paste into adaptation briefs.
Preserve verbatim nuance and confidentiality
Evidano provides transcription with custom dictionaries and PII redaction so teams can preserve verbatim nuance while protecting participant confidentiality.
Use Evidano transcription with custom dictionaries and PII redaction to convert interviews or session recordings into safe, shareable text for secondary analysis and for inclusion in supporting files.
Produce decision-ready outputs
Evidano exports adaptation checklists, stakeholder memos, co-occurrence networks, and slide-ready summaries that map recommendations to implementation tasks.
Generate exportable adaptation checklists, stakeholder memos, co-occurrence networks, and a short slide-ready summary that maps recommendations (peer role, legal content, session sequencing) to implementation tasks and costs.
Security note: data is encrypted and not used to train third-party models, see Evidano.
2‑week AI-enabled runbook to reproduce these results
The runbook is a day-by-day checklist to turn de-identified NGT notes and rankings into an adaptation plan within 14 days.
- Day 0–1: Gather inputs: de-identified NGT notes, ranking spreadsheets, one-page intervention summaries, and any audio. Confirm consent and IRB constraints.
- Day 1–2: Ingest and preprocess: upload files to Evidano, run PII redaction, and (if audio) auto-transcribe with a custom dictionary for local terms.
- Day 2–4: Thematic extraction: auto-generate themes, import an initial codebook, and refine with one to two human-pass edits.
- Day 4–6: Cross-segment analysis: compute theme frequencies, rank concordance, and extract representative quotes per segment (providers versus PLWH).
- Day 6–9: Co-occurrence and network checks: map which themes cluster (for example, 'peer support' plus 'trust' plus 'session-based').
- Day 9–11: Synthesis and adaptation checklist: produce prioritized changes (peer involvement, legal content, session sequencing).
- Day 11–14: Stakeholder brief and shares: export visuals, a one-page memo, and run a short AI chat over the corpus to answer anticipated funder or IRB questions.
Ethics & safeguards
Ethics and participant safety require separating consent forms, scrubbing identifiers, and following IRB guidance before sharing outputs.
- The study authors removed direct and indirect identifiers before data sharing in the study’s supporting information.
FAQ: HIV disclosure decision aids
What did Zurashvili et al. find in their NGT study?
Zurashvili et al. found that providers favored digital delivery while PLWH favored session-based support, and the joint group prioritized the Family disclosure decision aid for adaptation.
The study ran three NGT sessions (providers, PLWH, joint), used one-page summaries of nine candidate interventions, and aggregated de-identified notes and ranking scores to reach consensus on adapting the Family disclosure decision aid for Georgia.
How can researchers convert NGT notes into an adaptation plan quickly?
Researchers can convert NGT notes into an adaptation plan by extracting ranked items, coding themes, quantifying cross-segment concordance, and synthesizing a prioritized checklist within a 2-week runbook.
Follow the runbook steps: gather inputs, ingest and redact PII, auto-transcribe if needed, generate themes, run cross-segment analyses, map co-occurrences, and synthesize the adaptation checklist.
What limitations did the study report that teams should watch for?
The study reported under-representation of younger PLWH and reliance on field notes rather than verbatim transcripts as key limitations.
Teams adapting the intervention should consider targeted recruitment to cover under-represented subgroups and capture verbatim recordings when consent and IRB permit.
How does Evidano handle confidentiality and PII?
Evidano supports PII redaction and encrypted storage to protect participant confidentiality during transcription and analysis.
Use PII redaction at ingest, keep consent and identifying forms separate, and follow IRB guidance before sharing outputs or secondary analyses.
Conclusion, next steps & CTA
Zurashvili et al. (16 July 2026) show that a short NGT study yields clear priorities for adaptation, and the operational gap is turning notes and rankings into reproducible adaptation plans quickly.
- Ready to try this on your NGT notes or interview transcripts? Run the 2-week workflow above and produce an adaptation checklist and stakeholder brief in days using Try Evidano for free.
- If you need a secure pilot: ingest de-identified notes, run theme and cross-segment analyses, and export visuals that decision-makers can act on immediately (see Evidano).