Faster Qualitative Analysis of Health Literacy: HELIA-TB

Evidano is an AI-powered qualitative data analysis platform that ingests transcripts, supports multilingual transcription and translation, runs thematic and cross-segment analyses, and produces visual reports and audit-ready appendices. The PLOS ONE study adapted the Health Literacy Instrument for Adults to TB (HELIA‑TB) for Gujarat, India and validated it with 393 adults, showing strong reliability (overall α=0.86; domain α range 0.82–0.89; ICC=0.88) and predictive validity for treatment adherence (Cohen's d=0.89). Researchers and program teams can cut manual coding and synthesis time by applying AI-enabled workflows that preserve cultural nuance and auditability. This post explains a practical, reproducible workflow (grounded in the HELIA‑TB adaptation) and shows how Evidano ingests transcripts, supports multilingual transcription/translation, runs thematic and cross-segment analyses, and produces visual reports you can share with stakeholders.

Key Takeaways

Chauhan et al. adapted the Health Literacy Instrument for Adults to TB (HELIA‑TB), validated it in Gujarat with 393 adults, and published the validation on 30 June 2026.

HELIA‑TB retained five domains (access, reading, understanding, appraisal, decision-making) and showed strong internal consistency (overall α=0.86; domain α 0.82–0.89), content validity (S‑CVI/Ave=0.92), and predictive validity for adherence (Cohen's d=0.89).
The adaptation followed an exploratory sequential mixed-methods design (expert review, forward/back translation, cognitive interviews, pilot n=30, full psychometrics n=393; Mar 2024–Mar 2025).
Teams can reproduce HELIA‑TB style adaptations with a 7-step workflow combining cognitive interviews, pilot testing, and AI-assisted coding and cross-segment checks to move from interviews to program-ready recommendations quickly.

Findings snapshot

Published	Sample (n)	Key metrics	Primary result	Source
30 June 2026	393 adults with TB (validation sample)	Overall α = 0.86; domain α = 0.82–0.89; ICC = 0.88; S‑CVI/Ave = 0.92; Cohen's d (adherence) = 0.89	HELIA‑TB is reliable, valid, and predictive of adherence	PLOS ONE

Fast take + source

Chauhan et al. culturally adapted HELIA to TB, ran forward and back translations, conducted cognitive interviews and pilot testing, then validated the instrument with 393 adults and published the results on 30 June 2026.

Why researchers care: HELIA‑TB retained five domains (access, reading, understanding, appraisal, decision-making) and showed high internal consistency (α range 0.82–0.89) and content validity (S‑CVI/Ave = 0.92).
Payoff for your team: AI-assisted workflows speed item drafting, codebook alignment, and cross-segment checks so teams can move from interviews to program-ready recommendations in days rather than weeks.
Source: PLOS ONE.

How the adaptation worked (plain English)

The adaptation used an exploratory sequential mixed-methods design: expert review, forward translation into Gujarati, cognitive interviews (5 patients and 8 frontline workers), back-translation, pilot testing (n=30), then full psychometric testing (n=393) from March 2024 to March 2025. Item edits preserved original domains but replaced general items with TB-specific wording (for example, “I can find health information about TB”).

The qualitative phase tuned language and stigma sensitivity; the quantitative phase ran confirmatory factor analysis (CFI = 0.946; RMSEA = 0.051) confirming the five-factor structure.
Predictive checks linked higher HELIA‑TB scores to better treatment adherence and self-rated health.

Implications for researchers: qualitative analysis of health literacy

For qualitative researchers

Qualitative researchers should preserve cognitive interview transcripts and field notes because thematic patterns that explain why items were changed (terminology, stigma) are as important as psychometrics.

Preserve cognitive interview transcripts and field notes: thematic patterns that explain why items were changed (terminology, stigma) are as important as psychometrics.

Use cross-segment comparisons (for example, low vs. no formal education) to ensure items function across literacy strata; HELIA‑TB reported α=0.904 for educated and α=0.848 for no formal education.

For program evaluators & policymakers

Program evaluators and policymakers can use validated, context-adapted measures like HELIA‑TB to identify patients at high risk of non-adherence and trigger counselling or social support.

Validated, context-adapted measures (like HELIA‑TB) identify patients at high risk of non-adherence, actionable screening can trigger counselling or social support.

Quantify impact: the observed Cohen's d = 0.89 for adherence implies large practical effects when health literacy is addressed.

For UX/design teams

UX and design teams should run iterative rapid cycles with frontline workers to refine phrasing before large-scale piloting because cognitive interviews exposed confusing terms.

Cognitive interviews exposed confusing terms (for example, 'physician' → 'doctor'); use iterative rapid cycles with frontline workers to refine phrasing before large-scale piloting.

Design interfaces that surface easy examples and local phrasing to improve comprehension during digital surveys.

Do more, faster with Evidano

Ingest & clean

Upload transcripts, interview notes, reports, and survey spreadsheets to ingest mixed inputs while preserving metadata such as date, location, and respondent type.

Evidano ingests mixed inputs and preserves metadata (date, location, respondent type).

Transcription & translation

Transcribe audio with a custom dictionary and apply PII redaction, then translate between Gujarati and English while maintaining developer-approved phrasing.

Automatically transcribe audio with a custom dictionary (local terms like 'ASHAs', 'DOT') and PII redaction; translate Gujarati↔English while maintaining developer-approved phrasing.

AI-assisted coding & thematic analysis

Generate initial codebooks from cognitive interview themes, run thematic, frequency, and co-occurrence analyses, and refine codes interactively to reduce manual coding hours.

Generate initial codebooks from cognitive interview themes, run thematic, frequency, and co-occurrence analyses, and refine codes interactively, reducing manual coding hours.

Cross-segment and predictive checks

Run cross-segment comparisons by education, facility type, and adherence, and link qualitative themes to quantitative scores to surface explanatory mechanisms.

Run cross-segment comparisons (education, facility type, adherence) and link qualitative themes to quantitative scores (for example, HELIA‑TB scores vs. adherence) to surface explanatory mechanisms.

Audit trail & stakeholder-ready outputs

Produce traceable quotes tied to codes, shareable dashboards, and a downloadable methods appendix that auditors and programs can review.

Produce traceable quotes tied to codes, shareable dashboards, and a downloadable methods appendix for ethics review, useful when integrating measures into national programs.

This week’s 7-step workflow to reproduce HELIA‑TB style adaptation

This 7-step workflow lists the practical steps to run a rigorous, reproducible adaptation with AI support.

Follow these steps to run a rigorous, reproducible adaptation with AI support:

1) Collect: Gather the original instrument, existing translations, and program documents.
2) Expert review: Upload documents to Evidano; use AI chat to summarize domain alignment and risky terminology.
3) Forward translate: Keep bilingual drafts and store both versions in Evidano for side-by-side comparison.
4) Cognitive interviews: Record audio, transcribe with a custom dictionary, and tag confusing items automatically.
5) Revise items: Use extracted themes and example quotes to justify item edits in a shared report.
6) Pilot & validate: Import pilot responses (spreadsheets) and run reliability, content-validity, and cross-segment analyses.
7) Package: Export psychometric tables, codebook, and a stakeholder brief for program uptake.

Wrapping up & next steps

Combine rigorous field methods with AI-assisted workflows to speed synthesis and maintain auditability when adapting health literacy tools.

If you are adapting health literacy tools or running cognitive interviews at scale, combine rigorous field methods with AI-assisted workflows to speed synthesis and maintain auditability.

Next move: try the 7-step workflow on your instrument with a small pilot, upload transcripts and pilot spreadsheets to Evidano and generate thematic and cross-segment reports in hours, not weeks (Try Evidano for free).

Learn more or request a demo: visit Evidano and reference HELIA‑TB as a use case.

FAQ: HELIA‑TB

What is HELIA‑TB and who validated it?

HELIA‑TB is the Health Literacy Instrument for Adults adapted specifically for tuberculosis, and Chauhan et al. validated it in Gujarat, India with 393 adults, publishing the validation on 30 June 2026.

Chauhan et al. culturally adapted HELIA to TB and completed psychometric testing with a validation sample of 393 adults.

What domains does HELIA‑TB include and how reliable are they?

HELIA‑TB includes five domains: access, reading, understanding, appraisal, and decision-making, and the domains showed high internal consistency with α values between 0.82 and 0.89.

HELIA‑TB retained five domains and reported domain α values in the 0.82–0.89 range and an overall α of 0.86.

How was HELIA‑TB adapted in the field?

HELIA‑TB was adapted using expert review, forward translation into Gujarati, cognitive interviews with patients and frontline workers, back-translation, pilot testing (n=30), then full psychometric testing (n=393) from March 2024 to March 2025.

The adaptation process included cognitive interviews that tuned language and stigma sensitivity and item edits that replaced general wording with TB-specific examples.

What psychometric and predictive metrics did HELIA‑TB achieve?

HELIA‑TB achieved confirmatory fit indices (CFI = 0.946; RMSEA = 0.051), ICC = 0.88, S‑CVI/Ave = 0.92, and a predictive effect for adherence of Cohen's d = 0.89.

The instrument showed strong model fit, inter-rater reliability, content validity, and a large practical effect size for adherence.

How can teams reproduce a HELIA‑TB style adaptation quickly?

Teams can reproduce HELIA‑TB style adaptations quickly by following a 7-step workflow that combines document collection, expert review, forward translation, cognitive interviews, pilot testing, psychometric analysis, and packaging for stakeholders.

The 7-step workflow includes collecting source materials, using AI-assisted summaries and codebook generation, running cognitive interviews and pilot analyses, and exporting psychometric tables and stakeholder briefs.

Topics

HELIA-TB
health literacy adaptation
qualitative analysis TB India
cognitive interviews
Evidano

Keep reading

Browse all articles

Faster Qualitative Analysis of Health Literacy: HELIA-TB

Key Takeaways

Findings snapshot

Fast take + source

How the adaptation worked (plain English)

Implications for researchers: qualitative analysis of health literacy

For qualitative researchers

For program evaluators & policymakers

For UX/design teams

Do more, faster with Evidano

Ingest & clean

Transcription & translation

AI-assisted coding & thematic analysis

Cross-segment and predictive checks

Audit trail & stakeholder-ready outputs

This week’s 7-step workflow to reproduce HELIA‑TB style adaptation

Ethics & a short caveat

Wrapping up & next steps

FAQ: HELIA‑TB

What is HELIA‑TB and who validated it?

What domains does HELIA‑TB include and how reliable are they?

How was HELIA‑TB adapted in the field?

What psychometric and predictive metrics did HELIA‑TB achieve?

How can teams reproduce a HELIA‑TB style adaptation quickly?

Topics

Keep reading

Company

About

Preview Evidano

Account

Newsletter