Evidano is an AI-powered qualitative data analysis platform that accelerates transcription, thematic coding and cross-segment analytics for program teams handling sensitive field data. A new quasi-experimental mixed-methods protocol published 15 July 2026 examines an IYCF intervention in the tribal Ganjad PHC of Palghar District, Maharashtra. The study enrolls 460 mother–infant (6–12 months) dyads across 44 Anganwadi Centres, runs a 24-month program (6-month prep, 12-month intervention, 6-month post), and is registered CTRI/2024/06/068427 (06 June 2024). For researchers working on Infant and Young Child Feeding (primary keyword: qualitative analysis of IYCF practices), the protocol is a rich source of IDIs, process logs, diet cards and weekly indicators, ideal for thematic and cross-segment analysis. Read the original protocol at PLOS ONE. This post shows how AI-enabled qualitative research workflows (transcription, thematic coding, segment comparison, and visualization) accelerate synthesis and help convert field data into program decisions using Evidano.
Key Takeaways
This post shows how teams can convert Palghar's mixed-methods IYCF data into rapid, defensible themes and program decisions using AI-enabled qualitative workflows.
- The Palghar protocol (published 15 July 2026) combines IDIs with structured pre/post surveys and process monitoring across 77 AWCs, enrolling 460 mother–infant dyads at baseline and 460 at endline.
- The study area is Ganjad PHC in Palghar District, Maharashtra, and the trial is registered CTRI/2024/06/068427 (06 June 2024).
- Evidano streamlines the pipeline: ingest audio and logs, auto-transcribe with local dictionaries, apply PII redaction, auto-code at scale, and produce cross-segment visualizations for stakeholders.
Findings snapshot
| Metric | Value | Source / Note |
|---|---|---|
| Published | 15 July 2026 | PLOS ONE |
| Design | Quasi-experimental mixed-methods (pre/post + qualitative IDIs) | Protocol details |
| Sample | 460 mother–infant dyads (6–12 months) per survey | 44 AWCs / cluster sampling |
| Study area | Ganjad PHC, Palghar District, Maharashtra (tribal communities) | Context: geographic & cultural barriers |
| Duration | 24 months (6m prep, 12m intervention, 6m post) | Timeline in Fig 2 |
| Registration | CTRI/2024/06/068427 (06 June 2024) | Clinical trial registry |
Fast take: Why this study matters for qualitative teams
This section explains why the Palghar protocol is valuable for qualitative researchers and program teams.
The Palghar protocol combines in-depth interviews (IDIs) with structured pre/post surveys and process monitoring across 77 AWCs to evaluate a multi-component IYCF intervention.
- Key program components include SBCC training for frontline workers, pictorial IEC, hot cooked complementary feeds at Anganwadi Centres, and micronutrient supplementation.
- Qualitative inputs include IDIs with ASHAs, AWWs and supervisors collected during Feb–Sep 2024 situational analysis.
- Qualitative analysis explains why and how uptake varies, for example due to cultural beliefs, logistics and counselling quality.
What the protocol collected (methods in plain English)
This section summarizes the data streams the protocol generated for mixed-methods analysis.
The study pairs structured quantitative surveys (eight WHO IYCF core indicators, anthropometry) with qualitative IDIs of frontline workers and stakeholders.
Process monitoring logs (weekly), diet cards, IEC materials, WhatsApp group threads and audio recordings from cooking demonstrations and role plays create a multi-modal corpus.
- Qualitative data are IDIs, role-plays, field notes and supervisor feedback, transcribed and analyzed thematically (Nvivo v13 in protocol).
- Quantitative data are baseline and endline surveys (n=460 each), with process indicators tracked weekly (for example, number receiving ≥2 complementary feeds at AWC).
- Intervention elements that generate text and audio artifacts include SBCC sessions, recipe demos, community competitions and mother diet cards.
So what for qualitative researchers and program teams?
What to code for
This subsection lists the thematic areas the Palghar protocol makes important to code for.
Counselling quality, including AWW and ASHA language, use of pictorial aids and recall of messages, should be coded.
Barriers and enablers such as food access, cultural taboos, hot meal acceptability and caregiver time should be coded.
Process fidelity, such as frequency of demonstrations, diet card completeness and WhatsApp engagement, should be coded.
Segments to compare
This subsection identifies the participant and provider segments worth comparing in analysis.
Frontline workers (AWW versus ASHA versus supervisor) should be compared on knowledge, confidence and counselling style.
Caregiver subgroups should be compared across different tribal communities, literacy levels and households using hot meals versus THR.
Time windows should be compared, for example baseline IDIs versus reflections during and after the intervention to capture behaviour change narratives.
Key indicators for mixed analysis
This subsection recommends indicators that link qualitative themes to quantitative outcomes.
Map thematic prevalence of barriers to changes in MDD, MMF and MAD outcomes.
Use process logs to explain heterogeneity in anthropometric results across AWCs.
Prioritize quotes that explain anomalous quantitative signals, for example low take-up despite supply.
Do more, faster with Evidano (mapped to this protocol)
From messy audio and WhatsApp threads to clean transcripts
Evidano converts messy audio and WhatsApp threads into clean transcripts using auto-transcription and custom dictionaries.
Auto-transcribe IDI and field audio with a custom dictionary for local tribal terms and food names, and apply PII redaction to ensure ethics and anonymity while preparing public datasets.
Consistent coding and rapid theme discovery
Evidano enables consistent coding and rapid discovery of themes by proposing and applying codebooks at scale.
Import or build a codebook from the protocol (IYCF indicators, SBCC elements); Evidano auto-suggests codes, applies them at scale and produces hierarchical codes to subcodes for reflexive thematic analysis.
Cross-segment and frequency analysis
Evidano supports cross-segment comparisons and frequency analytics to quantify where barriers appear most often.
Run cross-segment comparisons (AWW versus mothers, different padas and tribes) and frequency analytics to quantify how often barriers appear and where to target refresher training.
Visuals that convince stakeholders
Evidano generates visual outputs designed for stakeholder briefs and program meetings.
Generate co-occurrence networks, code hierarchies and word clouds for reports and stakeholder meetings, appropriate for ICDS and Poshan program briefs.
Secure, research-grade environment
Evidano provides an encrypted, research-grade environment and does not use customer data to train third-party models.
Data is encrypted end-to-end in Evidano and not used to train third-party models, a critical safeguard when handling data from vulnerable populations.
7-step workflow: reproduce Palghar’s qualitative synthesis in Evidano
This 7-step workflow outlines how to move from raw collection to an evidence brief using Evidano.
- 1) Ingest: upload audio (IDIs, demos), photos of diet cards and survey spreadsheets into Evidano.
- 2) Transcribe & translate: run auto-transcription with a custom dictionary for local food terms and tribal names; apply PII redaction.
- 3) Import codebook: load the protocol’s thematic codes (IYCF core items, SBCC constructs) or let Evidano propose starter codes.
- 4) Auto-code + review: apply AI-assisted coding, then human-review high-uncertainty segments.
- 5) Cross-segment analysis: compare themes by AWC, worker type and caregiver subgroup; export frequency tables.
- 6) Visualize: create co-occurrence networks and hierarchical code trees for presentations to district officials.
- 7) Report & iterate: produce a 1-page executive brief and clickable quote bank for policy conversations; schedule follow-up AI avatar interviews for unresolved questions.
Wrapping up: next moves
This conclusion recommends next steps for teams with IYCF mixed-methods data.
The Palghar protocol is a strong example of programmatic mixed-methods data that benefits from AI-assisted qualitative workflows.
For teams running IDIs, process logs and diet-card backed interventions, AI tools that handle audio, local language, secure PII redaction and cross-segment thematic analytics shorten the path from fieldnotes to policy.
- Read the full protocol on PLOS ONE.
- Ready to pilot this workflow on your IYCF corpus? Explore Evidano or Try Evidano for free.
FAQ: qualitative analysis of IYCF practices
What did the Palghar protocol study?
The Palghar protocol evaluated a multi-component IYCF intervention using a quasi-experimental mixed-methods design.
The protocol paired structured quantitative surveys (including eight WHO IYCF core indicators and anthropometry) with qualitative IDIs of frontline workers, process monitoring logs and diet cards to assess intervention uptake and program fidelity.
How many participants and AWCs were included?
The study enrolled 460 mother–infant dyads at baseline and 460 at endline across 44 Anganwadi Centres.
The protocol describes cluster sampling across 44 AWCs and collected additional qualitative data from ASHAs, AWWs and supervisors during a situational analysis in Feb–Sep 2024.
How can teams use Evidano with this protocol's data?
Teams can use Evidano to transcribe, redact PII, auto-code and run cross-segment analytics on the protocol's mixed data streams.
Evidano supports ingesting audio, photos of diet cards and survey spreadsheets, auto-transcribing with custom local dictionaries, applying PII redaction and producing visualizations and exportable frequency tables.
What safeguards are in place for vulnerable populations?
Evidano provides encryption and does not use customer data to train third-party models, protecting vulnerable populations.
PII redaction and encrypted storage are highlighted as critical safeguards when preparing public datasets from tribal community data.