Effective coverage for maternal health: operationalising effective coverage cascades for antenatal care and nutrition interventions for pregnant women in seven low- and middle-income countries.

Background: Efforts to improve maternal health have focused on measuring health and nutrition service coverage. Despite improvements in service coverage, maternal mortality rates remain high. This suggests that coverage indicators alone do not fully capture the quality of care and may overestimate the health benefits of a service. Effective coverage (EC) cascades have been proposed as an approach to capture service quality within population-based coverage measures, but the proposed maternal health EC cascades have not been operationalised. This study aims to operationalise the effective coverage cascades for antenatal care (ANC) and maternal nutrition services using existing data from low- and middle-income countries (LMICs).

Methods: We used household surveys and health facility assessments from seven LMICs to estimate EC cascades for ANC and maternal nutrition services provided during ANC visits. We developed theoretical coverage cascades, defined health facility readiness and provision/experience of care scores and linked the facility-based scores to household survey data based on geographic domain and facility type. We then estimated the coverage cascade steps for each service by country.

Results: Service contact coverage for at least one ANC visit (ANC1) was high, ranging from 80% in Bangladesh to 99% in Sierra Leone. However, there was a substantial drop in coverage from service contact to readiness-adjusted coverage, and a further drop to quality-adjusted coverage for all countries. For ANC1, from service contact to quality-adjusted coverage, there was an average net decline of 52 percentage points. For ANC1 maternal nutrition services, there was an average net decline of 48 percentage points from service contact to quality-adjusted coverage. This pattern persisted across cascades. Further exploration revealed that gaps in service readiness including lack of provider training, and gaps in provision/experience of care such as limited nutrition counselling were core contributors to the drops in coverage observed.

Conclusions: The cascade approach provided useful summary measures that identified major barriers to EC. However, detailed measures underlying the steps of the cascade are likely needed to support evidence-based decision-making with more actionable information. This analysis highlights the importance of understanding bottlenecks in achieving health outcomes and the inter-connectedness of service access and service quality to improve health in LMICs.