The Rajshahi City pilot (data collected July–Sept 2024; published 2 July 2026) reports broadly positive user perceptions of an Electronic Immunization Registry (EIR) but flags operational issues like intermittent internet and device limits. This post shows researchers and program teams how to run a rigorous qualitative analysis of EIR perceptions (turning 305 caregiver interviews and 16 HCP responses (n=321) into evidence-based recommendations) and how to operationalize that workflow in Evidano (Evidano). You will get a short reproducible workflow, key checks for bias and transferability, and concrete ways to combine open-ended feedback with registry metrics (for example, 87% facility registration; ≈88% timely registration within 45 days) to inform scale-up decisions.
Key Takeaways
Evidano is an AI-powered qualitative data analysis platform that ingests transcripts, survey sheets, and EIR CSVs, supports PII redaction and auto-coding, and produces stakeholder-ready reports.
The Rajshahi City PLOS One pilot (data Jul–Sep 2024; published 2 July 2026) found broadly positive caregiver and health worker perceptions, while providers reported operational barriers such as slow connectivity and device issues.
- Sample: n = 321 total (305 caregivers, 16 health care providers), cross-sectional perceptions using 5-point Likert plus open-ended questions.
- Operational metrics reported in the pilot include 87% facility-based registration and approximately 88% timely registration within 45 days.
- Caregiver satisfaction ranged roughly 59–78% satisfied and 16–22% highly satisfied; HCPs reported 75–88% highly satisfied, interpret HCP figures cautiously because n=16.
- This analysis is descriptive, so combine thematic counts with registry metrics and targeted follow-ups before making causal program decisions.
Fast take: what the paper shows (quick link)
This PLOS One pilot assessed feasibility and acceptability of an EIR in Rajshahi City, caregivers and health workers reported mostly satisfied or highly satisfied responses, and common provider challenges were slow connectivity and device issues.
Read the source: PLOS One.
- Study period: July–September 2024; Published: July 2, 2026.
- Sample: n = 321 (305 caregivers, 16 HCPs).
- Design: cross-sectional, Likert plus open-ended questions; descriptive analysis only.
Findings snapshot
| Metric | Value | Notes / Implication |
|---|---|---|
| Participants (total) | 321 | 305 caregivers; 16 HCPs |
| Data collection | Jul–Sep 2024 | Cross-sectional perceptions, not longitudinal |
| Published | 2 July 2026 | PLOS One (open access) |
| Facility-based registration | 87% | Most registrations occurred at EPI centers |
| Timely registration | ≈88% within 45 days | Reported in discussion as consistent with other LMIC pilots |
| Caregiver satisfaction | Satisfied 59–78%; Highly satisfied 16–22% | Reflects service perceptions (for example, SMS reminders) rather than direct registry use |
| HCP satisfaction | Highly satisfied 75–88% | Small n (16): interpret cautiously |
What the study did (plain language)
The study used a pretested semi-structured questionnaire with 5-point Likert items and open-ended prompts to measure feasibility and acceptability among caregivers and health care providers.
Quantitative items were summarized descriptively in SPSS, and open-ended responses were reviewed and narratively summarized but not shared publicly because of potential identifiers.
- Focus: perceptions of EIR-supported services (reminders, certificates, access to information), not an effectiveness trial.
- Limitations: purposive sampling at EPI sessions, cross-sectional design, small HCP sample (n=16), self-report bias.
Implications: qualitative analysis of EIR perceptions for researchers and program teams
Overview
This section summarizes implications for qualitative researchers, program teams, and policy leads based on the Rajshahi pilot findings.
Treat caregiver responses as service-experience data and triangulate with registry logs to reduce bias and inform operational priorities.
For qualitative researchers
Qualitative researchers should treat caregiver responses as service-experience data and code for perceived benefits versus system visibility.
Triangulate themes with registry logs, for example timeliness and registration location, to reduce social desirability bias.
For program/implementation teams
Program teams should prioritize operational fixes flagged in open-ended feedback, namely connectivity and device reliability.
Estimate impact by cross-referencing theme frequencies with missed appointments and dropout counts in the EIR and use thematic counts to target training and offline-capable solutions before rural scale-up.
For policy/monitoring leads
Policy and monitoring leads should be cautious interpreting high satisfaction as proof of impact because the study is descriptive.
Commission longitudinal audits or randomized rollout to measure causal effects and use qualitative themes as decision triggers, for example a threshold mention rate to budget offline sync functionality.
Do more, faster with Evidano
Overview
Evidano ingests mixed inputs and automates thematic analysis to accelerate qualitative workflows and produce stakeholder-ready outputs.
Use the features below to operationalize the Rajshahi-style workflow on your own EIR data.
Ingest mixed inputs
Evidano ingests interview transcripts, survey spreadsheets, and EIR CSV exports and links records via child ID or household identifiers.
Upload interview transcripts, survey spreadsheets, and EIR CSV exports and ensure unique keys such as child ID or household are present.
Rapid thematic + frequency analysis
Evidano auto-generates themes from open-text responses and provides exportable theme counts and representative quotes.
Auto-generate themes from open-text responses, export theme counts and representative quotes, and use cross-segment analysis to compare caregivers versus HCPs or facility versus outreach registration locations.
Secure by design
Evidano provides PII redaction, custom dictionaries, encrypted storage, and a guarantee that your data is not used to train third-party models.
Transcription with PII redaction, custom dictionaries for local names and vaccines, encrypted storage, and Evidano guarantees that your data is never used to train third-party models, which is important for sensitive health data.
Close the loop
Evidano automatically generates visualizations and slide-ready summaries to help stakeholders prioritize operational fixes.
Automatically generate visualizations (for example, word clouds, co-occurrence networks, hierarchical code maps) and export slide-ready summaries to prioritize fixes such as offline sync or device procurement.
Collect follow-ups
Evidano can run low-cost follow-up qualitative interviews using AI avatar interviewers to validate themes without adding field staff.
Use AI avatar interviewers to run follow-up qualitative interviews in local languages or to validate themes at scale.
Checklist: reproduce this qualitative analysis in 7 steps
This checklist lists seven reproducible steps to convert open-ended feedback and EIR exports into actionable recommendations.
Step 1: Collate datasets, transcripts, open-text survey fields, and EIR CSV exports; ensure unique keys (child ID or household).
Step 2: Redact PII and map local terms via a custom dictionary before upload.
Step 3: Auto-code with Evidano to generate initial themes; review and refine the codebook (human-in-the-loop).
Step 4: Run cross-segment frequency analysis (caregiver vs HCP; facility vs outreach) to quantify theme prevalence.
Step 5: Link themes to registry metrics (timeliness, missed dose counts) to prioritize operational fixes.
Step 6: Produce stakeholder-ready visuals (top themes, quote bank, co-occurrence network) and a one-page decision memo.
Step 7: Design a two-week follow-up using AI avatar interviews for quota validation or targeted probes.
FAQ: qualitative analysis of EIR perceptions
How do I avoid overestimating satisfaction when participants are recruited at clinics?
Compare clinic-recruited responses with outreach or phone follow-ups to detect recruitment-channel bias.
Weight themes by recruitment channel and check for social desirability language, for example repeated praise without specifics.
Can open-ended data be linked to registry metrics without violating privacy?
Yes, you can link open-ended data to registry metrics if identifiers are pseudonymized and strict access controls are used.
Pseudonymize identifiers, apply strict access controls, and use encrypted linkage workflows; Evidano supports PII redaction and encrypted linkage to preserve privacy while enabling cross-analysis.
What thematic triggers should prompt immediate operational fixes?
Use predefined thresholds in thematic counts as operational triggers, for example >25% mentions of connectivity problems or >15% mentions of confusing SMS reminders.
Operational triggers in the Rajshahi-style workflow could include implementing offline sync for connectivity mentions above a threshold and revising SMS content when confusion is frequently reported.
Conclusion, next steps
The Rajshahi pilot provides descriptive perceptions but does not establish causal impact, so combine thematic counts with registry metrics and targeted follow-ups to inform decisions.
- Ready to reproduce this workflow on your EIR data? Explore a secure, research-focused pipeline at Evidano or Try Evidano for free and spin up a pilot to import transcripts, code open text, and produce stakeholder-ready reports in days.
- Ethics note: qualitative findings in health contexts are research-focused and non-diagnostic, obtain ethical approvals and anonymize data before analysis.