Sociodemographic and Environmental Determinants of Insecticide Treated Net Utilization among Reproductive Age Women in Liberia: A Multilevel Analysis

Abstract

In Liberia, malaria remains hyperendemic with a year-round transmission despite a major scale-up of ITN distribution. The 2022 Liberia Malaria Indicator Survey indicated high household ownership but persistently low utilization among women of reproductive age. Understanding why women do not use available ITNs is essential for closing this ownership–use gap and reducing preventable malaria morbidity and mortality among mothers and children. The study used nationally representative data from the 2022 Liberia Malaria Indicator Survey and environmental and urbanization indicators. The dataset for the study contains 3,593 women aged 15–49 years from households with at least one ITN across all 15 counties and Greater Monrovia out of the 4,513 women who participated in the 2022 MIS. A multilevel logistic regression is used to analyze hierarchical clustering effects and estimate the individual- and county-level determinants of ITN use and random intercepts for each county as follows: a null model, an individual-level model, a contextual/environmental model, and a full model combining all predictors. Model performance is measured with intraclass correlation coefficients and marginal/conditional R². ITN use was significantly clustered by county (null-model ICC = 0.05), indicating considerable geographic heterogeneity. At the individual level, correct malaria knowledge (OR ≈ 1.98, 95% CI: 1.45–2.69) and pregnancy status (OR = 1.36, 95% CI: 1.03–1.79) were the most robust positive predictors of ITN use, and higher wealth, larger household size, and subordinate intra-household (daughter/other relative) status were associated with lower probability of use. Contextual and environmental predictors suggested that, compared to people living in counties of low elevation (<100 m), residence at mid-range elevations (≈101–400 m) was a strong negative factor for ITN use (ORs ≈ 0.51–0.69), independent of individual characteristics. These core effects were preserved in the full model (marginal R² = 0.135; conditional R² = 0.151), which reduced but did not eliminate county-level variance (ICC = 0.02), showing that unmeasured structural-related factors that influence ITN uptake, such as the quality of the program's implementation, robustness of the supply chain, mass ITN distribution, and local norms, continue to influence ITN implementation. These findings suggest that the use of ITNs among Liberia's reproductive-age women shows a multilevel relationship between individual knowledge about risk and maternal health status; intra-household power and resources allocated; and the ecological setting (especially elevation) and county-level health system differences. These findings suggest that simply increasing ITN coverage is not enough; targeted behavior change communication, pregnancy-focused interventions, intra-household equity strategies, and program strengthening for mid-elevation and underperforming areas in the counties are required to address the ownership–use gap and promote malaria control in Liberia.