Mother's care-seeking behavior for neonatal danger signs from qualified providers in rural Bangladesh: A generalized structural equation modeling and mediation analysis

BACKGROUND: Neonatal deaths contribute to nearly half (47%) of under-five mortality globally and 67% in Bangladesh. Despite high neonatal mortality, care-seeking from qualified providers for newborn danger signs remains low. Identification of direct and indirect factors and their pathways affecting care-seeking will help to design a well-targeted intervention. This study assessed the direct, indirect, and total effect of the predictive factors on neonatal care-seeking in Bangladesh. MATERIALS AND METHODS: This was a cross-sectional baseline household survey conducted in 14 districts of Bangladesh in 2019 with 17,251 recently delivered women (RDW) with a live birth outcome in the preceding 15 months. We used a two-stage stratified cluster sampling process to select the samples from 14 districts. We investigated the inter-relationship of maternal background characteristics, maternal health utilizations, child/neonate factors, health service delivery-related factors and newborn danger sign knowledge with newborn care-seeking practices and estimated the direct, indirect, and total effects using Generalized Structural Equation Modeling (GSEM) and mediation analysis. p-value = 0.05 was considered statistically significant. The result of the mediation analysis was reported in Log Odds (LOD). The positive LOD (LOD > 0) implies a positive association. RESULTS: Half of the mothers (50.8%) reported a neonatal illness and among them, only 36.5% mothers of sick neonates sought care from qualified providers. Our mediation analysis showed that maternal health utilization factors, i.e., 4 + antenatal care visits (ANC) from a qualified provider (LOD: 0.63, 95% CI: 0.49, 0.78), facility delivery (LOD: 0.74, 95% CI: 0.30, 1.17) and postnatal care (PNC) from a qualified provider (LOD: 0.50, 95% CI: 0.21, 0.78) showed the highest total effect over other factors domains, and therefore, were the most important modifiable predictors for qualified neonatal care-seeking. Other important factors that directly and/or indirectly increased the chance of newborn care-seeking from qualified providers were household wealth (LOD: 0.86, 95% CI: 0.70, 1.02), maternal education (LOD: 0.48, 95% CI: 0.32, 0.63), distance to nearest health facility (LOD: 0.20, 95% CI: 0.10, 0.30), community health worker's (CHWs) home visits during ANC (LOD: 0.24, 95% CI: 0.13, 0.36), neonatal danger sign counseling after delivery (LOD: 0.20, 95% CI: 0.06, 0.34) and women's knowledge of neonatal danger signs (LOD: 0.37, 95% CI: 0.09, 0.64). CONCLUSION: The inter-relationship and highest summative effect of ANC, facility delivery, and PNC on newborn care-seeking suggested the maternal care continuum altogether from ANC to facility delivery and PNC to improve care-seeking for the sick newborn. Additionally, referral training for unqualified providers, targeted intervention for poorer households, increasing CHWs home visits and neonatal danger sign counseling at the facility and community should also be considered.