Health researchers need qualitative analysis that can preserve cultural meaning, not just count repeated words. The November 11, 2026 study by Aurore Dupont, Wim Saakie, Marc-Alexandre Tareau, and Romain Blaizot examined how people in French Guiana understand leprosy, including local terms, spiritual explanations, stigma, and treatment practices. The study shows why AI qualitative analysis can help researchers organize multilingual interviews and compare themes, while human and community expertise remain essential for interpretation. This article explains the study’s method and findings, then translates them into a practical workflow for public health, anthropology, and health communication teams. The analysis is research-focused and non-diagnostic.
Key Takeaways
According to the November 11, 2026 PLoS Neglected Tropical Diseases study, culturally specific explanations of leprosy in French Guiana shaped how participants understood symptoms, transmission, stigma, and care. AI qualitative analysis can help researchers identify these patterns across interviews, but culturally grounded interpretation must remain central.
- The study conducted between April and September 2025 included 20 patients and 37 community members across 37 interviews.
- The study found that Maroon and Haitian participants often described spiritual or taboo-related causes, while only 3 participants explicitly referred to bacteria as a cause.
- The study identified “kokobe” and “gwasi” as local terms associated with deformities and skin patches, although the terms did not exactly match the biomedical category of leprosy.
- The study reported that French Guiana had an annual leprosy prevalence of about 1 case per 10, 000 inhabitants between 2015 and 2021.
- The study found that plant-based remedies were widely used, but reported use did not appear to reduce adherence to multidrug therapy.
How the French Guiana Leprosy Study Used Qualitative Research
The 2026 study used semi-structured interviews and thematic analysis to examine how patients and communities explained leprosy. The authors recruited 20 adult patients from health centres across French Guiana and 37 community members in western French Guiana between April and September 2025.
The researchers used purposive sampling for patients and snowball recruitment for community participants. Patient interviews explored illness history, explanatory models, stigma, socioeconomic effects, and care pathways, while community interviews examined collective representations and historical memories.
The research team manually transcribed audio-recorded patient interviews, recorded community interviews through detailed field notes when cultural mediation was required, and organized quotations and codes in an Excel database. The authors combined deductive coding from the interview guide with inductive coding for emergent themes.
The authors wrote that “recruitment and preliminary thematic assessment were conducted concurrently” and continued recruitment until thematic saturation. The study’s design was not intended to statistically compare cultural groups, and the authors cautioned that small group sizes limited generalization.
The study also demonstrates why multilingual and culturally sensitive analysis matters. Twelve of the 19 patient interviews required mediation or a relative’s presence, involving five different languages, while mediation was required in 17 of 18 community interviews.
Findings Snapshot: AI Qualitative Analysis Needs Cultural Context
| Date or period | Metric | Value | Implication for qualitative researchers |
|---|---|---|---|
| April to September 2025 | Patients interviewed | 20 | Patient narratives covered symptoms, beliefs, diagnosis, treatment, stigma, and care access. |
| April to September 2025 | Community members interviewed | 37 | Community narratives added historical and collective interpretations of leprosy. |
| April to September 2025 | Total interviews | 37 | The dataset combined individual, group, patient, and community interviews. |
| 2015 to 2021 | Annual leprosy prevalence in French Guiana | About 1 case per 10, 000 inhabitants | The persistent local burden made culturally appropriate communication relevant to public health planning. |
| Study analysis reported in 2026 | Participants explicitly mentioning bacteria as a cause | 3 participants | The finding indicates a gap between biomedical explanations and local explanatory models. |
| 2024 source cited by the study | Peripheral neuropathy attributed to diabetes versus leprosy in French Guiana | 58.2% versus 0.53% | Confusion between conditions may contribute to delayed recognition and requires careful thematic interpretation. |
Implications for Public Health and Health Communication Teams
Why should researchers code local terminology separately from biomedical diagnoses?
Researchers should code local terminology separately because the 2026 French Guiana study found that “kokobe” and “gwasi” described visible manifestations and disability but did not exactly correspond to the biomedical category of leprosy.
A qualitative codebook can therefore distinguish terminology, symptoms, causal explanations, stigma, and treatment beliefs. That separation prevents an analysis from treating every mention of a local term as an equivalent clinical diagnosis.
How can qualitative analysis reveal communication gaps about disease transmission?
Qualitative analysis can reveal communication gaps by comparing participant explanations with biomedical information discussed in clinical care. The study found that person-to-person transmission was rarely recognized, while environmental, familial, dietary, and spiritual explanations appeared frequently.
One participant told the researchers, “I don’t think it’s a disease that can be transmitted.” The quotation shows why frequency counts alone are insufficient: researchers must examine the meaning, context, and consequences of each belief.
What can researchers learn from the study’s findings about stigma and support?
Researchers can learn that stigma and social support may coexist. The 2026 study documented fear, nondisclosure, discrimination, and self-isolation, but it also found that many participants described family, religious, and community support.
A strong thematic analysis should therefore code both harmful experiences and protective social mechanisms. The study quoted one community member saying, “They are not separated from the community... we live together normally, we talk and eat together.”
How should AI support culturally sensitive health research?
AI should support culturally sensitive health research by accelerating organization, comparison, and retrieval while leaving interpretation and ethical judgment with the research team. Researchers can use AI to group related excerpts, compare patient and community segments, identify co-occurring themes, and locate contradictory narratives.
The study’s authors concluded that culturally sensitive communication should clarify person-to-person transmission while taking local practices into account. AI can make that evidence easier to review, but community mediators and subject-matter experts must validate translations, codes, and interpretations.
How Evidano Supports AI Qualitative Analysis of Health Interviews
Problem: Manual thematic coding slows multilingual health research
Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.
Evidano can ingest interview transcripts and research documents, then generate thematic, content, frequency, and cross-segment analyses. Researchers studying culturally diverse health experiences can use these analyses to compare patient narratives with community narratives while retaining the original source material for review.
Teams can explore Evidano’s qualitative research features to see how document-based analysis can support a structured thematic workflow.
Problem: Translation can erase culturally specific meanings
Evidano supports translation with custom dictionaries, which can help research teams preserve recurring local terms and preferred translations during multilingual analysis.
A custom dictionary can be useful when a term such as “kokobe” or “gwasi” should remain visible as a culturally specific expression rather than being flattened into a general biomedical label. Human researchers should still validate translations and culturally sensitive interpretations.
Problem: Researchers need to compare themes across participant groups
Evidano supports cross-segment analysis, allowing teams to compare themes across groups such as patients, community members, language groups, locations, or interview types.
For a study modeled on the French Guiana research, analysts could compare patient and community accounts of transmission, stigma, treatment, local remedies, and social support. Researchers can also use visualizations such as code hierarchies and co-occurrence networks to inspect how themes connect.
Problem: Sensitive interview data requires strong safeguards
Evidano encrypts data and does not use customer data to train third-party models, a relevant consideration for health research involving rare diseases and potentially identifiable narratives.
The French Guiana study did not publicly share its interview transcripts because participants could potentially be identified despite pseudonymization. Researchers should apply the same principle of data minimization, access control, and ethical review when using AI-assisted analysis.
FAQ: AI Qualitative Analysis for Health Research
What is AI qualitative analysis?
+
AI qualitative analysis uses artificial intelligence to help organize, code, summarize, compare, and retrieve meaning from unstructured research data such as interviews and open-ended responses.
The 2026 French Guiana leprosy study illustrates a use case where AI could assist with thematic organization across patient and community interviews, while researchers remain responsible for interpretation and validation.
How many people participated in the French Guiana leprosy study?
+
The study included 20 patients and 37 community members in research conducted between April and September 2025.
The participants contributed to 37 interviews, including individual and group formats, with language mediation used extensively in both patient and community research.
What did the study find about cultural beliefs about leprosy?
+
The study found that spiritual explanations, taboo transgression, familial causes, environmental exposures, and local remedies shaped many participants’ understandings of leprosy.
The researchers reported that bacterial causation was rarely mentioned, while “kokobe” and “gwasi” were widely used in Maroon community contexts to describe deformities and skin patches.
Can AI qualitative analysis replace cultural mediators or health researchers?
+
AI qualitative analysis cannot replace cultural mediators or health researchers because meaning depends on language, history, context, and participant relationships.
The study relied on a multidisciplinary team that included dermatologists, a medical anthropologist, and a Maroon cultural health mediator. That team structure helped the researchers refine codes and interpretations across cultural and clinical perspectives.
How should researchers protect sensitive qualitative health data?
+
Researchers should protect sensitive qualitative health data through informed consent, pseudonymization, restricted access, encryption, and careful decisions about whether transcripts can be shared.
The 2026 study withheld its interview transcripts from public release because rare disease narratives and contextual details could make participants identifiable even after pseudonymization.
Conclusion and Next Steps for AI Qualitative Analysis
The 2026 French Guiana study shows that qualitative health research must connect language, cultural explanations, treatment practices, stigma, and social support. AI qualitative analysis can help researchers process multilingual interviews, compare participant segments, and trace themes across a complex dataset, but human and community expertise remain essential.
Researchers should use AI to make evidence easier to organize and interrogate, not to replace reflexivity, ethical safeguards, or culturally informed interpretation. Teams working with interviews, open-ended surveys, or health documents can Try Evidano for free to begin an AI-assisted qualitative analysis workflow.
Topics
- AI qualitative analysis
- AI thematic analysis
- qualitative health research
- leprosy stigma research
- culturally sensitive research
Keep reading
- Commentary on NewsFrom Stories to Themes: AI Qualitative AnalysisLearn how AI qualitative analysis can synthesize sensitive health research while preserving participant voice, using a 2026 Pakistan leprosy study.
- Commentary on NewsFive Hajj Sun-Safety Themes: AI Qualitative AnalysisLearn how AI qualitative analysis clarifies Hajj sun-protection research, from 29 interviews to five themes, and turns multilingual findings into action.
- Commentary on NewsWhat ELLY Shows About AI Qualitative AnalysisLearn how AI qualitative analysis can reveal intervention pathways, community needs, and participant outcomes from the ELLY health study. Try Evidano.