Site Logo
All articles
Commentary on News

AI for Qualitative Analysis: Developmental Assessment in Ethiopia

Evidano7 min read

This post explains how AI-enabled qualitative research can translate the PLOS One findings on child developmental assessment barriers in Ethiopia into actionable program and research steps. The PLOS One study (Gebeyehu et al., PLOS One, published August 10, 2026) collected 20 key informant interviews between August 1, 2023 and January 1, 2024 and identified six implementation themes, according to the PLOS One article. Health program designers, qualitative researchers, and digital research teams will find practical mappings from those themes to AI-assisted methods and workflows that speed synthesis and preserve rigour, according to the PLOS One study (Gebeyehu et al., PLOS One, 2026).

Key Takeaways

PLOS One found that implementation of routine child developmental assessment in Ethiopia is constrained by system-level shortages, workforce gaps, community stigma, and parental barriers, according to Gebeyehu et al., PLOS One (published August 10, 2026).

  • The PLOS One study conducted 20 semi-structured key informant interviews between August 1, 2023 and January 1, 2024 with 14 health professionals and 6 parents, according to Gebeyehu et al., PLOS One (2026).
  • As of July 2024 the Gamo Zone population was 2, 391, 628 and children under five numbered about 368, 310 (15.4%), according to Gebeyehu et al., PLOS One (2026).
  • As of December 2025 the study area had 61 health centers, 9 hospitals, and 272 health posts providing pediatric services, a service footprint that the authors say is under-resourced for developmental assessment (Gebeyehu et al., PLOS One, 2026).
  • A prior Ethiopian study cited in the paper reported that in Addis Ababa only 28.6% of providers assessed developmental milestones and 3.7% used a standardized screening tool (study cited in Gebeyehu et al., PLOS One, 2026).

What Happened and how the study measured barriers

The PLOS One study used a descriptive qualitative design to identify barriers and facilitators to developmental assessment implementation by interviewing 20 informants from August 2023 to January 2024, according to Gebeyehu et al., PLOS One (2026).

The PLOS One study purposively sampled one zonal MCH focal person, district officers, pediatricians, nurses, general practitioners, and six parents and then transcribed and coded interviews using ATLAS.ti 7, according to Gebeyehu et al., PLOS One (2026).

The PLOS One study mapped emergent themes to the Consolidated Framework for Implementation Research (CFIR) to organise findings across outer setting, inner setting, individual characteristics, and intervention characteristics domains, according to Gebeyehu et al., PLOS One (2026).

Direct quotations in the PLOS One paper illustrate frontline perspectives, for example a BSc nurse said, "The service being provided currently is not sufficient. The children come to the health facility by coincidence... it has not received attention like other diseases, " (BSc nurse, Gebeyehu et al., PLOS One, 2026).

Findings Snapshot

DateMetricValueImplication (as reported in PLOS One)
Aug 1, 2023–Jan 1, 2024Key informant interviews20 interviews (14 providers, 6 parents)Qualitative depth and saturation achieved in parent group (Gebeyehu et al., PLOS One, 2026)
July 2024Gamo Zone population2, 391, 628 total, 368, 310 under-five (15.4%)Large rural child population implies high service demand but limited accessibility (Gebeyehu et al., PLOS One, 2026)
Dec 2025Facility footprint in study zone61 health centers, 9 hospitals, 272 health postsExisting PHC structure can be leveraged but is under-resourced for developmental assessment (Gebeyehu et al., PLOS One, 2026)
2016 (cited in paper)Provider practice in Addis Ababa28.6% assess milestones; 3.7% use standardized toolLow adoption of standardized screening tools constrains early detection (study cited in Gebeyehu et al., PLOS One, 2026)

Implications for researchers and program implementers

Program designers and researchers should prioritise integrated, low-cost monitoring corners and targeted training because the PLOS One study found that overcrowded clinics and limited provider skills block routine assessment (Gebeyehu et al., PLOS One, 2026).

  • Policy: The authors recommend elevating developmental assessment in national child-health policy and allocating dedicated budgets because limited political commitment was repeatedly reported (Gebeyehu et al., PLOS One, 2026).
  • Service design: The authors recommend colocating brief screening within immunization and growth-monitoring visits because EPI contact points are reliable touchpoints, according to Gebeyehu et al., PLOS One (2026).
  • Community engagement: The authors recommend community education through Health Extension Workers to reduce stigma and increase parental knowledge because cultural beliefs and stigma were key access barriers (Gebeyehu et al., PLOS One, 2026).

How Evidano helps researchers and implementers map barriers to action

What is Evidano and why it fits this use case

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.

Evidano can ingest the PLOS One interview transcripts and reproduce reproducible thematic, frequency, and cross-segment analyses that map directly to CFIR domains as Gebeyehu et al. did manually, enabling faster iteration for implementers.

Problem: Large qualitative corpora and slow synthesis

Solution: Evidano automates transcript ingestion, supports verbatim transcription and translation, and produces thematic coding with exportable codebooks so implementation teams can move from raw text to CFIR-mapped themes in hours rather than weeks.

Feature links: Learn more about these capabilities in the features page.

Problem: Need for accurate transcripts in local languages

Solution: Evidano supports transcription with custom dictionaries and PII redaction and supports translation workflows that preserve local terms for cultural beliefs, which is essential when analysing community stigma as described in PLOS One (Gebeyehu et al., PLOS One, 2026).

Feature links: See speech-to-text for details on transcription accuracy and custom vocabulary.

Problem: Comparing themes across participant types (providers vs parents)

Solution: Evidano provides cross-segment analysis and co-occurrence visualizations so teams can quantify which themes (for example, 'workload' or 'stigma') appear more frequently among providers or parents, matching the CFIR cross-checking approach used in the PLOS One study (Gebeyehu et al., PLOS One, 2026).

Evidano preserves data confidentiality and uses proprietary LLMs that are not shared with third parties, enabling secure policy-sensitive analysis.

FAQ: qualitative analysis developmental assessment Ethiopia

What were the main barriers to implementing developmental assessment in Ethiopia according to the PLOS One study?

Answer: The main barriers were limited political commitment and funding, inadequate facility space and tools, workforce training gaps, the complexity and cost of diagnostic pathways, community stigma, and parental financial and knowledge constraints, according to Gebeyehu et al., PLOS One (2026).

Supporting detail: The PLOS One authors mapped these issues to CFIR domains to show multi-level drivers of low service availability (Gebeyehu et al., PLOS One, 2026).

How was data collected in the PLOS One study and how many participants were interviewed?

Answer: The PLOS One study conducted 20 semi-structured key informant interviews from August 1, 2023 to January 1, 2024, including 14 health professionals and 6 parents, according to Gebeyehu et al., PLOS One (2026).

Supporting detail: Interviews were audio-recorded, transcribed verbatim, translated into English, and coded using ATLAS.ti 7 for thematic analysis (Gebeyehu et al., PLOS One, 2026).

Can AI tools reproduce the CFIR mapping done by the PLOS One authors?

Answer: Yes, AI-assisted qualitative platforms can reproduce CFIR mapping at scale by combining thematic coding, automated code-to-construct mapping, and cross-segment frequency analysis, as a way to mirror the PLOS One methodology (Gebeyehu et al., PLOS One, 2026).

Supporting detail: The PLOS One authors applied CFIR retrospectively to inductive themes; AI workflows can accelerate that step while preserving human validation, according to implementation science best practices cited in the literature.

Is the PLOS One study generalisable beyond Gamo Zone?

Answer: The PLOS One authors state that findings are not statistically generalisable because the study was purposive and local, but they argue transferability is plausible because Ethiopia’s primary health care structure and IMNCI guidelines are nationally standardised (Gebeyehu et al., PLOS One, 2026).

Supporting detail: The authors recommend contextual adaptation when applying findings to other regions due to variations in resources and stakeholder engagement (Gebeyehu et al., PLOS One, 2026).

Conclusion & Next Steps

The PLOS One study shows that transforming developmental assessment in Ethiopia requires simultaneous policy commitment, modest infrastructure changes, workforce training, and community engagement (Gebeyehu et al., PLOS One, 2026).

AI-enabled qualitative workflows accelerate that work by turning interviews into validated CFIR mappings, quantifying theme prevalence across segments, and creating exportable reports for policymakers and funders, according to implementation researchers and platform case studies.

If you are a researcher or program lead who needs to turn qualitative interviews into actionable implementation plans, Try Evidano for free and see how automated transcription, thematic coding, and cross-segment analysis shorten the time from data collection to policy-ready evidence.

Topics

  • qualitative analysis developmental assessment Ethiopia
  • developmental assessment Ethiopia
  • AI qualitative research
  • child development screening Ethiopia

Keep reading

Browse all articles
Company
About
Newsletter

Product updates, research, and tips — straight to your inbox.

© Evidano, All Rights Reserved.