A predictive model for healthcare coverage in Yemen

Area of Science:

• Public Health
• Epidemiology
• Health Informatics

Background:

• Yemen's ongoing conflict severely impacts healthcare systems and information availability.
• Gaps in water, sanitation, and healthcare access contribute to high diarrheal disease mortality in children under five.
• A predictive model is crucial for understanding disease burden and healthcare access in data-scarce environments.

Purpose of the Study:

• To develop a predictive model for diarrheal disease burden in Yemeni children under five.
• To estimate healthcare coverage for treatment services among sick children.
• To inform health authorities and humanitarian partners for better service planning and delivery.

Main Methods:

• A probabilistic Markov model was developed using UNICEF clinical register data.
• The model integrated health system data with environmental and conflict-related factors.
• Estimates of incidence, treatment efficacy, and mortality rates were generated.

Main Results:

• The model accurately recreated observed treatment trends, validating its predictive capability.
• Predicted healthcare coverage for diarrheal treatment in 2019 ranged from 1.73% to over 5% weekly.
• Model predictions offer insights for policy decisions on service delivery timing and modalities.

Conclusions:

• The model reveals seasonal diarrheal disease trends in Yemeni children, influenced by weather, infrastructure, and conflict.
• It highlights that up to 98% of children may lack treatment access during peak rainy seasons.
• Scaling up community outreach and proximity-based services is recommended to improve treatment access.