Individual and Socioeconomic Determinants of Antenatal Care Access for Disadvantaged Pregnant Women: A Systematic Review and Empirical Analysis

Overview

This study investigated the extent to which individual-level socioeconomic and demographic characteristics explain antenatal care access among disadvantaged pregnant women. While socioeconomic disadvantage has been consistently linked to reduced maternal healthcare utilization, the explanatory capacity of individual-level variables remains unclear. The research addressed two complementary objectives: examining associations between socioeconomic status, demographic factors, and antenatal care completion using secondary survey data, and synthesizing global evidence on socioeconomic disadvantage and maternal healthcare access through systematic review and meta-analysis. The analysis integrated individual-level data from 2,019 women with evidence from six observational studies to assess both direct associations and the discriminatory performance of individual-level predictors. The study was motivated by recognition that existing evidence has primarily established associations between disadvantage and outcomes, with limited attention to the extent individual characteristics account for observed disparities in care utilization across heterogeneous health system contexts.

Methods and approach

The study employed a dual methodological approach combining secondary data analysis with systematic evidence synthesis. Individual-level data were obtained from a global maternal health survey administered through a smartphone-based platform and released by the Institute for Health Metrics and Evaluation, comprising 2,019 respondents from 51 countries surveyed between May and June 2021. Eligible participants were women who were currently pregnant or had given birth within the preceding six months. Multivariable logistic regression was applied to examine associations between socioeconomic status, demographic characteristics, and antenatal care completion after adjustment. Exploratory machine learning approaches, specifically K-Nearest Neighbors and XGBoost algorithms, were implemented to assess the discriminative capacity of individual-level variables for predicting antenatal care completion. The systematic review component followed PRISMA 2020 guidelines and included observational studies involving socioeconomically disadvantaged pregnant women reporting outcomes related to antenatal care utilization or maternal healthcare access. Studies were required to report adjusted effect estimates for inclusion in quantitative synthesis. A random-effects meta-analysis was conducted to synthesize adjusted odds ratios from eligible studies. Reference management was performed using Mendeley Reference Manager, and all statistical analyses were executed in R statistical software version 4.3.1.

Results

In the secondary data analysis of the Institute for Health Metrics and Evaluation dataset, socioeconomic status was not significantly associated with antenatal care completion after multivariable adjustment, with an adjusted odds ratio of 0.97 and 95% confidence interval of 0.78 to 1.21. Antenatal care completion rates were nearly identical across socioeconomic groups, at 80.0% for high/middle socioeconomic status and 79.6% for low socioeconomic status, with no statistically significant bivariate association detected. Individual-level predictive models demonstrated limited discriminatory performance, with area under the curve values ranging from 0.49 to 0.50 across both logistic regression and machine learning approaches, indicating minimal capacity to distinguish between women who did and did not complete antenatal care based on individual-level socioeconomic and demographic variables alone. In contrast, the random-effects meta-analysis synthesizing evidence from six observational studies showed that socioeconomic disadvantage was significantly associated with inadequate or delayed prenatal care, with a pooled adjusted odds ratio of 1.96 and 95% confidence interval of 1.26 to 3.07. Substantial heterogeneity was observed across included studies in the meta-analysis. The synthesis incorporated conceptually distinct indicators including migrant status and neighborhood risk as proxies of broader social vulnerability, acknowledging their heterogeneous nature.

Implications

The findings demonstrate that individual-level socioeconomic and demographic variables alone provide limited explanatory value for maternal healthcare utilization patterns, suggesting that broader structural and health system-level factors may play a more substantial role in determining antenatal care access. The discordance between the null association observed in individual-level data after adjustment and the significant pooled effect from the meta-analysis highlights important methodological considerations regarding the measurement and conceptualization of socioeconomic disadvantage in maternal health research. The near-chance discriminatory performance of predictive models incorporating individual-level characteristics underscores the inadequacy of individual-focused explanatory frameworks for understanding disparities in antenatal care completion. These results point toward the potential importance of contextual determinants, including health system organization, social protection mechanisms, transportation infrastructure, and community-level support systems, which may mediate or moderate the relationship between individual socioeconomic characteristics and care utilization. The substantial heterogeneity observed in the meta-analysis further reflects variations across settings in how socioeconomic disadvantage translates into barriers to care access. The study highlights limitations in generalizing individual-level determinants across diverse health system contexts and suggests that interventions targeting antenatal care access among disadvantaged populations may need to address structural barriers and system-level factors rather than focusing exclusively on individual characteristics. Future research should examine multilevel determinants and contextual factors that shape the pathway from socioeconomic disadvantage to maternal healthcare utilization.