Identifying Barriers to Maternal and Child Healthcare in Catholic Sister-Run Hospitals
Using AI-enabled avatar-led surveys to identify barriers to maternal and child healthcare in Catholic sister-run hospitals across East and Southern Africa.
Background
Justice and Ecology Network Africa (JENA) commissioned a study to better understand barriers to access, timeliness, and quality of maternal and child healthcare services in Catholic sister-run hospitals in Kenya, Uganda, and Zambia. The study formed part of JENA's broader interest in integrating digital health technologies within faith-based health facilities across Africa.
Barriers to maternal and child healthcare
Identifying key barriers affecting women's access to, experience of, and outcomes in maternity and childcare services, drawing on the perspectives of pregnant and postpartum women, healthcare workers, and community health workers.
Three-delays model with AI-enabled multilingual data collection
AI-based categorisation of responses using a large language model was combined with human-led qualitative analysis. The automated layer handled initial transcription, translation, and thematic coding across four languages, while the human review added contextual interpretation and addressed limitations related to repetition, lack of synthesis, and cultural nuance that automated systems alone can miss.
Results by country
Key findings
Financial constraints: The most pervasive barrier — leading to delayed decisions to seek care and incomplete treatment across all three countries.
Geography, transport, and infrastructure: Distance and road conditions prevented timely access to facilities, particularly in rural areas.
Traditional beliefs and cultural practices: Influencing decisions to avoid or delay facility-based care in favour of traditional birth attendants.
Gender norms and household power dynamics: Women frequently reported needing permission from husbands before seeking care, creating critical delays at the decision stage.
Knowledge and awareness gaps: Limited understanding of the importance of early antenatal care and recognition of danger signs during pregnancy.
Decision Impact
The three-delays framework proved effective at structuring the analysis, revealing how barriers compound across the decision, access, and care stages. The avatar-led approach achieved high completion rates and enabled detailed multilingual qualitative data collection in low-resource settings. Findings informed recommendations for transport support, health education, and gender-sensitive interventions.
Have a similar challenge?
We'd be happy to discuss how we might approach your specific context and requirements.