A Health Systems Innovation to Reduce Global Maternal Deaths: Quantitative Systems Modeling of Service Delivery Redesign for Maternal Health

Maternal mortality ratios (MMRs) remain high in low- and middle-income countries (LMICs) despite increasing facility births. We developed a quantitative systems model to evaluate a health systems innovation, Service Delivery Redesign (SDR), to reduce maternal deaths by shifting all deliveries to well-equipped hospitals in Kakamega, Kenya. The hybrid agent-based and system dynamics model captures interactions between pregnant women and a reconfigured multi-level health system. Six intervention scenarios with varying coverage and health systems input levels were simulated over 72 months. SDR is most effective when rising demand is matched with improved system capacity. Under the aggressive scenario, MMR declined by 64% (from 210.8 to 75.5 per 100,000 live births), and 189,000 DALYs per 100,000 live births were averted – at 131USD per DALY. By dynamically linking maternal and system behaviors, our model provides novel insights to guide design and implementation of health system strengthening initiatives to reduce maternal deaths in LMICs.