AI-enabled qualitative analysis can shorten the time from data collection to policy-ready insight for community-led cohort studies. Researchers and program leads working on Indigenous harm reduction studies need reproducible thematic synthesis that respects Indigenous data governance. This post uses the WHiSE 2.0 cohort profile to show concrete, ethical ways AI can help analyze open-ended interviews, questionnaire comments, and governance-linked datasets using ai-enabled qualitative analysis methods.
Key Takeaways
According to the PLOS One cohort profile (published July 24, 2026), WHiSE 2.0 enrolled 356 Indigenous participants across Thunder Bay, Sault Ste. Marie, and Sudbury; AI-enabled qualitative analysis can accelerate the planned mixed-methods synthesis while preserving OCAP and CARE governance. Verma et al., PLOS One (2026) emphasize that "WHiSE 2.0 is the first prospective cohort study examining the harm reduction needs of Indigenous people using substances in northern Ontario."
- WHiSE 2.0 enrolled 356 participants, with site counts 173 in Thunder Bay, 101 in Sault Ste. Marie, and 82 in Sudbury, according to PLOS One (published July 24, 2026).
- According to PLOS One (data collection Feb 2023–Dec 2024), 84.0% of participants reported knowing what harm reduction is and 94.8% knew where to access harm reduction services.
- According to PLOS One (2026), 43.3% of participants reported injecting drugs in the prior three months and 68.0% reported a lifetime overdose, showing both high need and high engagement with services.
- According to PLOS One (2026), 71.3% of participants engaged in Indigenous ceremonies and the authors note that the study is "a culturally grounded, community-led study" that requires Indigenous data governance in analysis.
What happened and how the WHiSE 2.0 data were collected
What happened: WHiSE 2.0 used community-based recruitment and an interviewer-administered questionnaire to enroll 356 Indigenous participants to describe substance use and harm reduction needs, as reported in PLOS One (Verma et al., published July 24, 2026).
According to PLOS One (2026), the study began in November 2022 and collected quantitative questionnaires in Thunder Bay from February 2023 to June 2024, in Sault Ste. Marie from May to November 2024, and in Sudbury from May to December 2024.
According to PLOS One (2026), the questionnaire included 119 items covering demographics, cultural connection, substance use history, harm reduction knowledge, STBBI testing, and overdose experiences; the team used REDCap for secure data capture as documented in PLOS One (2026).
According to PLOS One (2026), study governance applied OCAP and CARE principles and the authors describe community advisory review and an Indigenous advisory committee to govern secondary uses of data.
Findings snapshot
| Date / Source | Metric | Value (from PLOS One) | Implication |
|---|---|---|---|
| Published July 24, 2026; PLOS One | Total participants enrolled | 356 (173 Thunder Bay, 101 Sault Ste. Marie, 82 Sudbury) | Sufficient for cross-site descriptive comparisons and local planning |
| Data collection Feb 2023–Dec 2024; PLOS One | Knew what harm reduction is | 84.0% | High awareness but persistent access and cultural barriers |
| Baseline survey (round 1); PLOS One | Injected drugs in past 3 months | 43.3% overall (70.7% Sudbury, 40.5% Thunder Bay, 25.7% SSM) | Local variation requires tailored harm reduction supplies and services |
| Baseline survey; PLOS One | Ever had an overdose (lifetime) | 68.0% overall (83.2% SSM, 73.2% Sudbury, 56.6% Thunder Bay) | Indicates urgent need for naloxone, safe supply, and culturally safe response |
| Baseline survey; PLOS One | HIV and HCV testing (ever) | 82.3% HIV tested, 84.3% HCV tested | High testing uptake but diagnoses remain (9.6% HIV, 44.0% HCV) and need linkage-to-care |
Implications for mixed-methods and qualitative researchers
For researchers, WHiSE 2.0 demonstrates that community co-designed instruments and Indigenous governance require analysis workflows that both respect data sovereignty and scale thematic coding, as reported in PLOS One (2026).
According to PLOS One (2026), OCAP and CARE principles were embedded in governance; researchers should therefore build analytic pipelines that can export de-identified summaries for community review and restrict raw-text access as required by Indigenous partners.
According to PLOS One (2026), the study used interviewer-administered surveys plus planned qualitative work; combining those open-text data with structured variables benefits from reproducible AI-assisted thematic mapping that preserves audit trails and supports local interpretation.
- Design for local variation: PLOS One (2026) reports site-level differences (e.g., higher injecting in Sudbury); researchers should run cross-site code-frequency and co-occurrence analyses to detect divergent themes.
- Plan governance-first: PLOS One (2026) required Indigenous advisory review for any secondary use; researchers should implement gated access, summary-only exports, and community review steps in their analytic workflow.
- Expect mixed modalities: PLOS One (2026) collects interviewer notes and open responses; prepare for transcription, bilingual translation, and coded quotes with provenance metadata.
How Evidano helps apply AI-enabled qualitative analysis to WHiSE-style studies
What Evidano is
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Evidano can ingest interview transcripts, REDCap exports, and documents to produce thematic, frequency, and cross-segment analyses while preserving provenance and audit trails.
Problem: Large open-text volumes with governance constraints → Solution
Problem: WHiSE 2.0 combines 119-question surveys with planned qualitative interviews and OCAP/CARE governance, as described in PLOS One (2026).
Solution: Evidano supports secure ingestion of REDCap exports and document uploads, automated de-identification, and role-based access so community partners can review summaries without exposing raw identifiers; see Evidano features.
Problem: Manual coding is slow and inconsistent → Solution
Problem: Manual thematic coding across 356 participants and follow-up qualitative rounds is time-consuming, as implied by the scale reported in PLOS One (2026).
Solution: Evidano applies LLM-tuned thematic extraction and hierarchical coding with co-occurrence visualizations and frequency tables, enabling rapid codebooks that teams can review and lock before final analysis.
Problem: Audio transcription and PII handling → Solution
Problem: WHiSE 2.0 used interviewer-administered questionnaires and plans qualitative interviews, which will require transcription and PII safeguards as noted in PLOS One (2026).
Solution: Evidano offers transcription with custom dictionaries and PII redaction and integrates with secure pipelines for human review; see Evidano speech-to-text.
Problem: Community review and interpretive sovereignty → Solution
Problem: PLOS One (2026) emphasizes that Indigenous partners must control how results are used and shared.
Solution: Evidano exports community-ready summaries and visualizations (word clouds, co-occurrence networks, hierarchical codes→subcodes) and preserves an export log for advisory review, enabling community-led interpretation workflows.
FAQ: ai-enabled qualitative analysis
How can AI-enabled qualitative analysis help cohort studies like WHiSE 2.0?
Answer: AI-enabled qualitative analysis speeds coding, surfaces patterns, and produces reproducible summaries that researchers can verify and share with community partners.
Supporting detail: According to PLOS One (2026), WHiSE 2.0 has 356 participants and planned qualitative waves; AI-assisted thematic extraction reduces initial coding time and produces frequency and co-occurrence matrices for local review.
Is it ethical to use AI on Indigenous research data?
Answer: It can be ethical only if AI workflows are governed by community principles like OCAP and CARE and if community partners control access and outputs.
Supporting detail: According to PLOS One (2026), WHiSE 2.0 embedded OCAP and CARE in its governance; any AI analysis should implement gated access, de-identified exports, and advisory approval before dissemination.
What outputs should researchers expect from AI-enabled qualitative analysis?
Answer: Expect reproducible codebooks, ranked theme frequencies, cross-segment comparisons, quote extracts with provenance, and visualizations for community review.
Supporting detail: According to PLOS One (2026), WHiSE 2.0 will report mixed-methods findings; AI outputs that link codes to original text and to participant strata (site, age, housing) speed interpretation.
How does Evidano protect sensitive cohort data?
Answer: Evidano enforces role-based access, automated de-identification, encrypted storage, and audit logs to align with community governance needs.
Supporting detail: For WHiSE-style projects that follow OCAP/CARE as documented in PLOS One (2026), Evidano preserves export logs and enables summary-only sharing to advisory committees prior to any wider release.
Conclusion & Next Steps
WHiSE 2.0 (PLOS One, published July 24, 2026) provides a concrete case where ai-enabled qualitative analysis can accelerate culturally grounded insight without compromising governance.
Researchers should pair community-reviewed analytic plans with AI-assisted thematic tools to turn 356 participant records and forthcoming qualitative interviews into actionable, locally relevant recommendations, as highlighted by PLOS One (2026).
If you are preparing a mixed-methods, community-governed study and want a workflow that supports transcription, thematic coding, and gated community review, see how Evidano can fit into your pipeline via Evidano features and try an evaluation of your own data.
Next step: Try Evidano for free to upload a small de-identified sample and generate a thematic preview you can share with community partners.