AI-enabled qualitative analysis of menstrual health helps researchers and program teams move from interview transcripts to prioritized, evidence-backed recommendations. The primary audience for this post is qualitative researchers, WASH program designers, and NGO monitoring teams who need reproducible thematic syntheses of sensitive interview data. The PLOS One study published 24 July 2026 provides a compact, high-value dataset (18 in-depth interviews and 5 key informant interviews collected 01/09/2025–31/10/2025) that illustrates common coding challenges in menstruation research: short interviews, overlapping structural and cultural themes, and refusal or silence in transcripts. This post shows how AI-enabled qualitative research workflows can extract themes, frequency patterns, and quotable evidence from that PLOS One material to speed policy translation and program design.
Key Takeaways
According to the PLOS One study published 24 July 2026, women and adolescent girls in Khulna Railway Slum described menstrual experiences shaped by three interrelated domains: structural-environmental constraints, socio-cultural stigma, and gendered inequities. The PLOS One dataset includes 18 in-depth interviews and 5 key informant interviews collected between 01/09/2025 and 31/10/2025, which the authors analyzed thematically to saturation. Direct quotations in the PLOS One paper include participants saying, "When water stops coming from the tap, I wait for hours" (IDI-08) and "I wrap the used cloth in paper and throw it away at night" (IDI-10).
- PLOS One reports data collection from 01/09/2025 to 31/10/2025 with 18 IDIs and 5 KIIs, demonstrating thematic saturation by interview 15 (authors: Alam & Al-Mamun, 2026).
- PLOS One reports participant characteristics: ages 15–45, 39% aged 20–30, 39% non-literate, 33% secondary education, and 67% married (PLOS One, published 24 July 2026).
- PLOS One documents concrete WASH-related practices and risks: communal toilets without doors, hidden drying of cloths, and ad hoc disposal into drains, which the study links to infection risks and psychosocial harm (PLOS One, 24 July 2026).
What happened and how the study was done
What happened: the PLOS One team conducted a qualitative study in Khulna Railway Slum to explore lived menstruation experiences and drivers of menstrual hygiene management. The study area and publication are described in PLOS One (Alam & Al-Mamun, PLOS One, 24 July 2026).
How it was done: according to PLOS One, fieldwork ran from 01/09/2025 to 31/10/2025, using purposive sampling for 18 in-depth interviews (IDIs) with women aged 15–45 and 5 key informant interviews (KIIs). The authors used a combined inductive-deductive thematic analysis guided by a Feminist Political Ecology framework and coded data in NVivo v12 (PLOS One, 24 July 2026).
Constraints and measures: PLOS One reports ethical approval (IRB Approval No. 473478-FY 2024–2025; Approval Date: 10 August 2025) and intentional anonymization; the authors state that full transcripts are restricted for confidentiality, while a minimal de-identified dataset and codebook are available on request via the institutional data access committee (PLOS One, 24 July 2026).
Findings snapshot
| Date | Metric | Value | Implication |
|---|---|---|---|
| 01/09/2025–31/10/2025 | Interviews | 18 IDIs, 5 KIIs | Sufficient for thematic saturation as reported by authors (PLOS One, 24 July 2026) |
| 24 July 2026 | Publication | PLOS One article (Alam & Al-Mamun, 2026) | Peer-reviewed qualitative evidence about menstruation in Khulna informal settlements |
| Study sample | Age and education | Ages 15–45; 39% aged 20–30; 39% non-literate; 33% secondary education | Shows educational and age heterogeneity relevant for segmented thematic analysis (PLOS One, 24 July 2026) |
| Settlement data | Residency duration | 72% lived in slum >8 years | Long-term residency signals entrenched infrastructure and social patterns (PLOS One, 24 July 2026) |
Implications for qualitative researchers and program teams
How should researchers adapt interview guides for menstruation in informal settlements?
Answer: Researchers should embed WASH, privacy, disposal, and stigma probes in interview guides from the first question. The PLOS One study (24 July 2026) shows that structural themes (water scarcity, toilet privacy, waste disposal) and socio-cultural themes (shame, silence) consistently co-occur, so interviewers should capture both material and normative detail.
Supporting detail: PLOS One reports participants saying things like, "The toilet near our room has no door. Boys and men walk by all the time. I feel scared to change my clothes there" (IDI-04), which suggests an interviewer prompt on toilet design, lighting, and perceived safety is essential during data collection (PLOS One, 24 July 2026).
What should WASH program designers measure when using qualitative data?
Answer: Programs should map qualitative codes to measurable WASH indicators such as private toilet access, continuous water supply, and safe disposal options. The PLOS One authors directly link inadequate WASH with risky practices and psychosocial harms in their 24 July 2026 paper.
Supporting detail: PLOS One documents actions like hidden disposal into drains and drying cloths under beds, which programs can convert into monitoring indicators (e.g., percent of households with lockable toilets, percent with disposal bins) and measure alongside qualitative narratives (PLOS One, 24 July 2026). For technical guidance on integrating MHH into WASH programs, see the UNICEF guidance on menstrual health and hygiene.
How Evidano helps: mapping PLOS One needs to AI features
Definition and first-use
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Practical fit: PLOS One required thematic coding across three cross-cutting domains (structural-environmental, socio-cultural stigma, gendered inequities); Evidence synthesis via Evidano can automate initial code suggestions, surface co-occurrence patterns, and produce frequency and cross-segment summaries to speed analytic consensus.
Problem: small sensitive transcripts and confidentiality → Solution: secure ingestion and redaction
Problem statement: the PLOS One team restricted full transcripts for confidentiality (PLOS One, 24 July 2026).
Evidano feature: Evidano supports encrypted uploads and PII redaction during transcription, plus controlled access to de-identified excerpts for auditors, which helps teams reproduce coding while protecting participants.
Problem: mixed inductive-deductive coding → Solution: hybrid thematic workflows
Problem statement: PLOS One used a combined inductive and deductive approach and reconciled coder differences in consensus meetings (PLOS One, 24 July 2026).
Evidano feature: Evidano provides AI-assisted code suggestions anchored to seed codes, automated co-occurrence matrices, and visual hierarchical code→subcode maps, reducing the time spent on initial codebook building and inter-coder calibration. See Evidano features for specifics.
Problem: extracting quotable evidence and segment comparisons → Solution: searchable quotes and cross-segment frequency
Problem statement: PLOS One includes many participant quotations but manual extraction is time consuming (PLOS One, 24 July 2026).
Evidano feature: Evidano indexes transcripts for rapid retrieval of verbatim quotes by code, demographic tag, or date and generates tables that show code frequency by subgroup (e.g., adolescents vs adults), so teams can produce policy briefs faster.
Problem: audio-to-text in local languages → Solution: transcription and translation
Problem statement: the PLOS One team recorded interviews in Bengali and used back-checked translations for rigor (PLOS One, 24 July 2026).
Evidano feature: Evidano provides transcription with custom dictionaries, bilingual translation workflows, and back-translation checks to preserve participant meaning while enabling English analysis. See Evidano speech-to-text.
FAQ: AI-enabled qualitative analysis of menstrual health
How many interviews are enough to reach thematic saturation in MHH studies?
Direct answer: Many qualitative MHH studies report saturation around 12–20 IDIs, but context matters and iterative analysis is key.
Supporting detail: PLOS One collected 18 IDIs and 5 KIIs between 01/09/2025 and 31/10/2025 and reported saturation by interview 15, illustrating that 15–20 in-depth interviews often suffice for dense topical domains like menstrual stigma when complemented by key informants (PLOS One, 24 July 2026). AI-assisted coding can help detect saturation earlier by flagging repeated code patterns.
Can AI tools preserve confidentiality for sensitive interviews about menstruation?
Direct answer: Yes, when tools support encrypted storage, PII redaction, and access controls, AI workflows can preserve confidentiality.
Supporting detail: The PLOS One authors restricted full transcripts for ethical reasons (PLOS One, 24 July 2026). Platforms that offer PII redaction and role-based access help researchers comply with IRB safeguards while enabling reproducible analysis across teams.
How do I ensure AI-generated codes do not overwrite local cultural meanings?
Direct answer: Combine AI suggestions with human-led codebook validation and reflexive team checks.
Supporting detail: PLOS One used reflexivity and team consensus meetings to refine themes (PLOS One, 24 July 2026). The recommended workflow is: run AI-assisted coding, review suggested codes against raw quotes, and resolve differences in documented consensus meetings.
Can AI extract quotable evidence and generate a codebook from a small dataset?
Direct answer: Yes, AI can accelerate quote extraction and initial codebook drafts, but human verification remains essential.
Supporting detail: In the PLOS One study the authors produced a codebook and selected quotations (PLOS One, 24 July 2026); AI tools can replicate that step by surfacing candidate quotations and code frequencies for reviewer approval.
Conclusion & Next Steps
The PLOS One study (published 24 July 2026) demonstrates how tightly interwoven structural WASH issues, stigma, and resource inequality shape menstrual experiences in urban informal settlements. AI-enabled qualitative analysis can accelerate converting small, ethically sensitive datasets like the PLOS One interviews (18 IDIs, 5 KIIs collected 01/09/2025–31/10/2025) into actionable program recommendations and monitoring indicators. Evidano combines encrypted ingestion, transcription and translation, AI-assisted thematic coding, and quotable evidence extraction to help teams reproduce the PLOS One analytic workflow faster and more transparently; see Evidano features to learn more. Ready to move from transcripts to policy briefs? Try Evidano for free.