The 2022–2024 cholera outbreak in Malawi (57, 639 reported cases and 1, 727 deaths) exposed how climate shocks, fragile health services, and social vulnerabilities converge into prolonged epidemics. This post shows researchers and policy teams how to convert interview transcripts (n=24 first-responder interviews, Aug–Sep 2024) and field notes into system-level evidence using AI-enabled qualitative analysis. We reference the original PLOS study (see PLOS Neglected Tropical Diseases) and map a reproducible workflow you can run in Evidano to produce thematic, frequency, and cross-segment outputs for decision-ready reports.
Key Takeaways
This post shows how qualitative interviews from Malawi’s 2022–2024 cholera epidemic reveal converging climate, health-system, and social pathways that prolonged transmission, and how to reproduce that system-mapping with AI tools.
Evidano is an AI-powered qualitative data analysis platform that encrypts data, supports versioned analyses, and does not use your data to train third-party models.
- Qualitative evidence from 24 frontline interviews (data collected Aug–Sep 2024) supports the conclusion that pathway convergence, not a single cause, amplified the outbreak.
- Reported national burden in Malawi was 57, 639 cases and 1, 727 deaths during 2022–2024 (study figures).
- Use AI-enabled workflows to produce thematic frequencies, co-occurrence networks, and cross-segment comparisons that make qualitative findings decision-ready.
Fast take, why this matters for qualitative researchers
This section explains why the Malawi findings and methods matter to qualitative researchers and decision-makers.
A field-based qualitative study published July 10, 2026, used 24 in-depth interviews with first responders in Neno and Chikwawa to trace three converging pathways (climate, health-system fragility, and social/economic determinants) that turned a seasonal cholera spike into a two-year epidemic, see PLOS Neglected Tropical Diseases.
- Why readers should care: structured qualitative evidence like this is ideal for informing cross-sector preparedness, WASH investments, and vaccine or stockpile placement.
- Payoff: you will learn a compact AI-enabled workflow to code, quantify, and visualize pathway convergence (themes, co-occurrence, segment comparisons) so findings drive operational decisions.
Findings snapshot
| Metric | Value | Source / Note |
|---|---|---|
| Reported cases | 57, 639 | Malawi Ministry of Health (study figure) |
| Reported deaths | 1, 727 | Malawi Ministry of Health (study figure) |
| Study method | 24 in-depth interviews; field visits; thematic analysis | Authors: Livne et al.; data collected Aug–Sep 2024 |
| Geography | Neno & Chikwawa Districts (Southern Malawi) | Rural, cyclone-affected districts |
| Key climate events | Cyclone Ana (Jan 2022), Cyclone Gombe (Mar 2022), Cyclone Freddy (2023) | Authors’ timeline |
| Primary pathways identified | Climate shocks, health-system vulnerabilities, social/economic determinants | Convergence amplified transmission |
What happened (plain English)
This section summarizes the field findings about how the outbreak evolved and why transmission persisted.
Field teams interviewed 24 frontline health workers and managers and combined transcripts with site observations to map how successive cyclones destroyed WASH and transport infrastructure, displaced about 90, 000 households in Chikwawa, and interrupted supplies and staffing.
The study reports that a new, more-contagious Vibrio cholerae strain circulated in 2022–2023, damage to latrines and boreholes forced communities to use contaminated surface water, and IDP camp conditions and cross-border movement with Mozambique sustained transmission.
- Temporal note: data collection occurred August–September 2024; the article was published July 10, 2026.
- Analytic approach: qualitative descriptive study with iterative codebook development and thematic analysis (Dedoose used by authors).
- Main finding: the outbreak’s severity resulted from pathway convergence rather than any single factor.
So what for researchers, UX teams, and policy analysts
For qualitative researchers
This subsection explains what qualitative researchers should prioritize when studying outbreaks like Malawi’s cholera epidemic.
Qualitative researchers should prioritize system-mapping: code for cross-cutting pathways (climate, infrastructure, social) and extract co-occurrence patterns rather than isolated themes.
Qualitative researchers should quantify relevance: report theme frequency, intensity, and segment-level differences (for example, district or role) to move from narrative to decision-ready evidence.
For UX / design teams
This subsection explains how UX and design teams can use frontline qualitative findings to inform product and service prototypes.
UX and design teams should translate frontline quotes into user journeys: map access pain points (roads, supply chains, service availability) to prioritize prototypes, for example portable water-treatment kits or decentralized stockpiles.
UX and design teams should use co-occurrence visuals to show stakeholders how infrastructure failure links to behavior changes that increase risk.
For policy & health analysts
This subsection explains actionable steps policy and health analysts can take based on thematic timelines and convergence evidence.
Policy and health analysts should use thematic timelines to align funding windows with recovery cycles, because the study shows there was no room for recovery between repeated cyclones.
Policy and health analysts should require multi-sector indicators (WASH coverage, road access, IDP camp density) in outbreak early-warning dashboards.
How to run a qualitative analysis of cholera outbreak with AI
Problem: messy, multilingual field data
This subsection states the data challenges that AI-enabled qualitative workflows need to solve.
Transcripts, field notes, and short survey CSVs arrive in different formats, sometimes with local terms and sensitive details.
Evidano setup (what to ingest)
This subsection lists the recommended inputs for an AI-enabled qualitative workflow.
Upload interview audio, transcripts, observation notes, and survey spreadsheets into Evidano.
Use custom dictionaries to normalize local terms (for example local names for boreholes or camp types) before coding.
Automated prep (transcription & translation)
This subsection summarizes automated preparation steps for raw audio and transcripts.
Auto-transcribe interview audio with PII redaction and a custom dictionary to preserve local terms; translate non-English excerpts while retaining original-language anchors for validation.
AI-assisted coding & themes
This subsection explains the AI-assisted coding workflow and outputs.
Import your draft codebook or let Evidano suggest codes from a sample; run AI-assisted coding across transcripts, then refine iteratively with coder reconciliation.
Produce thematic frequencies and intensity scores to show how extensively and intensely themes appear.
Cross-segment & co-occurrence analysis
This subsection explains how to compare groups and identify pathway convergence in the data.
Generate cross-segment comparisons, for example Neno versus Chikwawa or clinicians versus environmental health officers, and co-occurrence networks to surface pathway convergence evidence similar to the Malawi study.
Deliverables for decision-makers
This subsection lists the outputs that make qualitative findings actionable for stakeholders.
Export visualizations, for example word clouds, co-occurrence graphs, hierarchical code to subcode trees, and clickable quote spreadsheets for reports and policy memos.
Security & reproducibility
This subsection states platform security and reproducibility features relevant to sensitive health data.
Evidano encrypts data, supports versioned analyses, and does not use your data to train third-party models, which is critical for sensitive health transcripts.
Checklist: 7 steps to go from raw interviews to operational insight
This checklist gives a seven-step sequence to convert raw qualitative inputs into decision-ready outputs.
Step 1: Gather inputs, audio, transcripts, field notes, and any survey CSVs; tag by district, role, and date.
Step 2: Upload to Evidano and run transcription with a custom dictionary and PII redaction.
Step 3: Build or import a codebook based on your research question (include climate, WASH, displacement, supply-chain codes).
Step 4: Run AI-assisted coding; review a 10–20% sample for reliability and refine codes.
Step 5: Produce thematic frequency tables, co-occurrence matrices, and cross-segment reports.
Step 6: Create stakeholder-ready outputs: a one-page findings snapshot, an evidence map, and a quotes appendix linked to codes.
Step 7: Iterate after stakeholder review; add targeted follow-up interviews using AI avatar interviewers if you need new strands of evidence.
FAQ: Qualitative analysis of cholera outbreak
What is 'qualitative analysis of cholera outbreak' and when should I use it?
Answer: Qualitative analysis of a cholera outbreak is the systematic coding and synthesis of interviews, observations, and contextual documents to surface drivers, coping mechanisms, and system failures, and you should use it when you need actionable context that complements case counts and lab data.
The study defines this approach as suitable for revealing pathway convergence and operational barriers that are not visible in epidemiological surveillance alone.
How do I compare segments reliably, for example district-level differences?
Answer: To compare segments reliably, segment all inputs at ingestion, ensure balanced sampling where possible, and use frequency plus intensity metrics alongside representative quotes to avoid over-interpreting rare mentions.
The post recommends tagging inputs by district and role and reporting both frequency and intensity when presenting differences between Neno and Chikwawa or between respondent roles.
What inputs should I ingest for this workflow?
Answer: Ingest interview audio, transcripts, observation notes, and survey spreadsheets, and tag them by district, role, and date.
The checklist and Evidano setup sections list these inputs and recommend custom dictionaries for local terms.
What ethics and data safeguards are required for outbreak interviews?
Answer: Ethics and data safeguards require informed consent, PII redaction, local ethics approvals, and secure storage of transcripts.
The post notes that Evidano supports PII redaction and secure, encrypted storage and that qualitative outputs are research-focused and non-diagnostic.
Wrapping up: next steps and CTA
This section recommends next steps for teams working on outbreak response, disaster recovery, or WASH programming.
If your team is working on outbreak response, disaster recovery, or WASH programming, convert transcripts into operational evidence faster by combining thematic, co-occurrence, and cross-segment analyses and reproduce the Malawi study’s system-mapping approach to show convergence of climate, health-system, and social pathways in your context.
Try a pilot: upload a small sample (5–10 interviews) to Evidano to generate a thematic snapshot and co-occurrence network you can share with decision-makers within days.
For the full PLOS study, see PLOS Neglected Tropical Diseases.