Evidano is an AI-powered qualitative data analysis platform that accelerates reproducible thematic and cross-segment analysis. Problem statement: Digital health pilots often report high-level outcomes but leave researchers and program teams guessing which operational frictions matter most. Context: In July 2026 PLOS published a user-perception study of an Electronic Immunization Registry (EIR) in Rajshahi City Corporation, Bangladesh, using data collected July–September 2024 (n=321). The paper reports broadly positive acceptability but flags connectivity and device issues that could block scale. This post shows how to turn the study's open-ended responses and Likert items into reproducible, actionable findings using AI-enabled qualitative research, including thematic and cross-segment analyses, quote extraction, and offline-capable coding workflows. See the original study at PLOS One. Ready to map perceptions to a rollout plan? Try a secure, research-focused pipeline on Evidano.
Key Takeaways
Core answer: The PLOS assessment (published 2 July 2026) of an EIR pilot in Rajshahi (data collected July–September 2024, n=321) shows strong acceptability among caregivers and health-care providers, but recurring operational barriers such as intermittent internet and device limitations that risk blocking scale.
Brief summary: The study combined Likert items and short open-text responses, producing measurable satisfaction signals (caregivers and HCPs) plus qualitative reports of infrastructure and workflow pain points that program teams should prioritize.
- Primary evidence source: PLOS One, published 2 July 2026, data collection July–September 2024.
- Caregiver satisfaction reported: satisfied 59–78%, highly satisfied 16–22%; HCP highly satisfied 75–88% (provider sample n=16).
- Main operational barriers reported: intermittent internet and device limitations, flagged in open-ended responses.
- Practical path: AI-enabled qualitative methods (thematic extraction, cross-segment analysis, prioritized fixes) convert open-ended feedback into prioritized, testable recommendations.
Findings snapshot (quick numbers)
| Metric | Value | Source / Note |
|---|---|---|
| Published | 2 July 2026 | PLOS One |
| Data collection | July–September 2024 | Study methods section |
| Sample (total) | 321 | 305 caregivers; 16 HCPs |
| Facility-based registration | 87% | Mostly at EPI centers |
| Registration timeliness | ≈88% within 45 days | Reported in discussion |
| Caregiver satisfaction | Satisfied: 59–78%; Highly satisfied: 16–22% | Likert items (descriptive) |
| HCP satisfaction | Highly satisfied: 75–88% | Small provider sample (n=16) |
| Main operational issues | Intermittent internet; device limitations | Open-ended responses summarized |
Fast take: what researchers need to know
Fast take answer: The PLOS study (published 2 July 2026) surveyed 305 caregivers and 16 health-care providers about an EIR pilot introduced in Rajshahi in 2019 and found high satisfaction alongside recurring operational problems such as intermittent internet and device limits.
Study snapshot: Data collection for this assessment occurred July–September 2024, and caregivers reported high satisfaction with EIR-supported services while HCPs reported strong utility for record management but recurring operational issues.
- Primary evidence source: PLOS One (PLOS One, 2 July 2026).
- Why it matters: Perception data identify practical blockers to scaling EIRs (infrastructure, offline workflows, training), exactly the inputs program planners need to prioritize.
- Payoff: AI-enabled qualitative methods convert short open-ended responses into prioritized fixes and measurable recommendations.
What the study actually measured
This section answers what the study measured: The study measured perceptions of feasibility and acceptability among direct users (health-care providers) and service recipients (caregivers) using a cross-sectional, descriptive design.
Design and scope: The study used a cross-sectional, descriptive assessment of perceptions, not an effectiveness trial, combining quantitative Likert items (5-point) with open-ended questions analyzed descriptively.
- Sampling: purposive recruitment of caregivers attending EPI sessions; the HCP sample was small (n=16), which limits generalizability.
- Outcomes measured: satisfaction, perceived usefulness of SMS reminders, preferences for EIR versus paper, and reported operational challenges.
- Limitations to note: self-report bias, lack of a comparison group, single urban setting (RCC), and open-ended responses withheld publicly due to identifiability concerns.
So what for researchers, UX teams, and program leads
For qualitative researchers
Direct answer: Researchers should treat the dataset as a mixed-format corpus suitable for combined thematic and frequency analysis to surface dominant and rare but critical failure modes.
Value: The dataset mixes Likert scales and short open-text answers, ideal for a combined thematic plus frequency analysis to surface dominant pain points and minority but critical failure modes.
Action: Use thematic coding (inductive and deductive) and cross-tabulate themes by respondent type (caregiver versus HCP) and registration channel (facility versus outreach) to reveal divergent needs.
For UX / Product teams
Direct answer: Product teams should treat reported device and connectivity friction as product requirements rather than solely training gaps.
Value: HCPs reported device- and connectivity-related friction despite high satisfaction; these are product requirements (offline mode, lightweight UI, robust sync).
Action: Prioritize offline-capable workflows, smaller payloads, and inline error messaging; validate fixes with rapid usability tests in the same wards used by the study.
For policy & program leads
Direct answer: Program leads should combine investments in digital infrastructure with microplanning using EIR-derived lists to protect service consistency.
Value: Caregiver satisfaction with SMS reminders is a lever for coverage, but infrastructure weak spots threaten consistency.
Action: Combine investment in digital infrastructure with microplanning using EIR-derived lists (target zero-dose clusters), and monitor timeliness metrics (for example, registration within 45 days) as rollout KPIs.
Do more, faster with Evidano, mapped to this study
Problem: Small open-text responses → hard to prioritize
Answer: Evidano accelerates prioritization by turning many short open-text responses into ranked themes with supporting evidence.
Solution (Evidano): Bulk ingest the survey instrument and open-ended fields, run AI-assisted thematic extraction, and return ranked themes with supporting quotes and frequency counts.
Problem: Compare caregiver vs provider views
Answer: Evidano enables cross-segment analysis and visualizes co-occurrence to compare caregivers and providers directly.
Solution (Evidano): Cross-segment analysis by role and registration channel; visualize co-occurrence of themes (for example, 'internet' plus 'sync failures' versus 'SMS' plus 'reminder helpful').
Problem: Multilingual transcripts or local terms
Answer: Evidano preserves domain-specific terminology through custom translation dictionaries and translation pipelines.
Solution (Evidano): Custom translation dictionaries plus a translation pipeline preserve domain terms (vaccine names, local facility labels) and ensure consistent coding.
Problem: Need reproducible reports for stakeholders
Answer: Evidano produces reproducible exports linking findings to source quotes and encrypted storage options for confidentiality.
Solution (Evidano): One-click exports: hierarchical code trees, word clouds, and executive briefs tied to source quotes, encrypted and private (data is not used to train third-party models).
Problem: Follow-up data collection at scale
Answer: Evidano supports autonomous, standardized qualitative follow-ups with consent flows and transcript generation.
Solution (Evidano): AI avatar interviewers for autonomous, standardized qualitative follow-ups (consent flows, PII redaction, and transcript generation).
Two-week checklist: reproduce the study analysis with AI-enabled qualitative methods
This checklist shows the steps to convert the study's approach into an actionable pilot evaluation using AI-enabled qualitative methods.
Step-by-step plan to convert the study's approach into an actionable pilot evaluation:
- 1) Import instruments and CSV survey exports into Evidano, and include metadata (role, date, ward).
- 2) Run automated transcription and cleanup for any audio, using a custom dictionary for Bengali vaccine terms when needed.
- 3) Generate an initial thematic model (auto-suggest codes), then review and finalize the codebook with stakeholders.
- 4) Produce cross-segment frequency tables (caregiver versus HCP; facility versus outreach) and co-occurrence networks.
- 5) Export an executive brief that maps the top three operational fixes to estimated implementation effort and impact.
- Deliverable in 10–14 days: coded dataset, visualization pack, and prioritized recommendations ready for microplanning.
Ethics & a short note for health researchers
Direct answer: The PLOS study appropriately anonymized quantitative data and restricted open-ended text to protect confidentiality, and re-analysis should preserve those constraints.
Guidance: When re-analyzing qualitative health data, preserve consent constraints, apply PII redaction, and treat outputs as research-only (non-diagnostic).
Platform note: Evidano supports PII redaction and encrypted storage to align with these requirements.
Wrapping up & next steps
Bottom line: The July 2, 2026 PLOS assessment of the Rajshahi EIR (n=321) shows strong acceptability but clear operational barriers that can be translated into prioritized, testable fixes.
- Ready to convert open-ended responses into prioritized rollout decisions? Start a secure project and demo the workflows cited above on Evidano.
- If you plan a scale-up, measure and monitor registration timeliness, device sync failures, and SMS delivery success as your core rollout KPIs.
Ethics reminder: Use anonymization and respect participant consent when reusing qualitative data.
Strong CTA: Try Evidano for free to book a walkthrough of an EIR-focused analysis pipeline and see how thematic, frequency, and cross-segment analyses cut synthesis time.
FAQ: Qualitative analysis of EIR perceptions
What did the Rajshahi EIR study measure?
Direct answer: The study measured perceptions of feasibility and acceptability among caregivers and health-care providers using a cross-sectional, descriptive design.
Details: The study combined 5-point Likert items with short open-ended questions to assess satisfaction, perceived usefulness of SMS reminders, preferences for EIR versus paper, and operational challenges.
What were the main findings on caregiver and HCP satisfaction?
Direct answer: Caregivers reported overall satisfaction (59–78%) with 16–22% highly satisfied, and HCPs reported high satisfaction (75–88%).
Context: The caregiver and provider satisfaction figures came from descriptive Likert-item reporting; note the HCP sample was small (n=16).
What operational issues did respondents report?
Direct answer: Respondents reported intermittent internet connectivity and device limitations as the main operational issues.
Implication: These infrastructure and device constraints are practical blockers to consistent EIR use and should be prioritized alongside training.
How can AI-enabled qualitative methods help with short open-text responses?
Direct answer: AI-enabled qualitative methods can extract themes, rank them by frequency and impact, and link each theme back to supporting quotes for prioritization.
Mechanics: Approaches include automated thematic extraction, cross-segment co-occurrence analysis, and reproducible exports for stakeholders.
Are there ethical considerations when reusing the study data?
Direct answer: Yes, re-analysis must preserve consent constraints, anonymize or redact PII, and treat the outputs as research-only.
Best practice: Use encrypted storage, PII redaction, and respect any restrictions on open-ended text, as the PLOS study did.