Qualitative analysis of electronic immunization registry

Evidano is an AI-powered qualitative data analysis platform that speeds PII-safe ingestion, de-identification, AI-assisted coding, and reproducible thematic analysis.

• A July–September 2024 cross-sectional EIR pilot in Rajshahi City Corporation (n=321: 305 caregivers, 16 HCPs) reported high satisfaction but providers noted intermittent internet and device issues (PLOS ONE, published 2 July 2026).
• Converting open-ended responses into a coded theme set lets teams quantify barrier vs enabler frequency and produce cross-segment comparisons (caregivers vs HCPs) for program decisions.
• A two-week reproducible workflow moves raw open-ends to stakeholder-ready briefs using PII redaction, translation harmonization, AI-assisted codebooks, and visual exports.

A cross-sectional descriptive assessment conducted July–September 2024 in Rajshahi City Corporation (RCC) asked caregivers (n=305) and healthcare providers (n=16) about feasibility and acceptability of an EIR introduced in 2019. Respondents reported high satisfaction and easier record access, while providers flagged intermittent internet and device issues. Full study: PLOS ONE.

• Why researchers care: the paper includes Likert-scale results plus open-ended responses that were summarized but not publicly shared, ideal raw material for deeper qualitative coding and cross-segment analysis.
• Quick payoff: convert descriptive open-ends into coded themes, quantify theme frequency across caregivers vs HCPs, and produce visual co-occurrence maps for program decisions.

The study design: cross-sectional purposive sampling of caregivers attending EPI sessions and HCPs using the EIR, using a pretested semi-structured questionnaire with 5‑point Likert items and open-ended fields. Analysis used descriptive statistics (SPSS) and a summary review of open responses, with qualitative answers summarized but not released due to identifiability.

• Strength for analysts: mixed numeric plus open-ended data are well suited for rapid thematic work and segment comparisons (caregivers vs providers).
• Limitation to note: purposive sampling and the small HCP n (16) can bias satisfaction estimates; open-ends were summarized descriptively rather than systematically coded.

Week 1 objectives: import and prepare data, run PII redaction, harmonize language with translation dictionaries, and create a 10–15 item seed codebook from a purposive sample.

• Day 1–2: Upload and map columns (respondent type, question id).
• Day 3–4: Auto-redact PII and run translation (if needed).
• Day 5–7: Seed-code 10–20% manually and generate an AI-suggested codebook.

Week 2 objectives: scale and synthesize by applying AI-assisted coding to the full corpus, run thematic frequency and cross-segment analysis, generate a co-occurrence network and export top quotes.

• Day 8–10: Review and refine automated codes; lock the final codebook.
• Day 11–12: Produce visuals (theme frequencies, co-occurrence map) and segment comparisons (caregiver vs HCP).
• Day 13–14: Draft a one-page brief with top findings and six verbatim quotes for stakeholders.

Ethics guidance: follow local IRB approvals and consent procedures for studies involving human subjects. The PLOS ONE paper anonymized data and withheld raw open-ends for identifiability concerns, mirror that caution when sharing quotes or datasets.

Next steps for teams preparing qualitative analysis of EIR pilot data: prioritize reproducible coding, segment comparisons, and secure handling of open-ends. Evidano speeds each step: from PII-safe ingestion and translation to AI-assisted codebooks, thematic frequency, and visual reports, so teams can move from descriptive summaries to actionable program fixes in two weeks.

• Try Evidano for free: upload one survey export and a sample of open-ends to see an auto-generated codebook and segment analysis in hours.
• Read the original PLOS ONE study here: PLOS ONE.