Does antenatal care utilisation vary in urban and rural health facilities in Ghana? A comparative cross-sectional study

Technical Summary: Urban-Rural Variations in Antenatal Care Utilisation in Ghana

Problem Statement
Despite improvements in maternal health access in Ghana, evidence regarding urban-rural disparities in antenatal care (ANC) utilization remains limited. While national data suggests high overall coverage, significant gaps persist between urban and rural populations regarding the timing of first booking and the number of visits. Previous literature often fails to disaggregate determinants by residence, limiting the development of targeted interventions. Furthermore, high utilization rates in some settings have not translated into proportional reductions in maternal mortality, suggesting potential compromises in the quality and continuity of care, particularly in rural areas. This study addresses the knowledge gap by examining whether significant differences exist in ANC utilization between urban and rural health facilities in the Bono Region of Ghana and identifying the socio-demographic determinants associated with these disparities.

Methodology
The study employed a facility-based comparative cross-sectional design involving 722 postpartum women (361 urban, 361 rural) within six weeks of delivery. Participants were selected from five districts in the Bono Region using a multistage stratified sampling technique. Data collection occurred between October 2022 and May 2023 using Computer-Assisted Personal Interviewing (CAPI) with a pre-tested structured questionnaire adapted from maternal health records and census instruments.

Key variables included:

• Outcome Measures: Timing of first ANC booking (defined as early if $\le$ 12 weeks), total number of ANC visits, and adequacy of care (defined as $\ge$ 4 visits per WHO recommendations).
• Predictors: Maternal age, marital status, educational level, occupation, and ethnicity.

Statistical analysis was conducted using Jamovi version 2.4.1. Descriptive statistics were used for demographic profiling. Pearson's chi-square tests and independent t-tests (including Welch's t-test where variance homogeneity was violated) assessed differences between urban and rural groups. To identify independent determinants while controlling for the urban-rural setting, the authors utilized Mantel–Haenszel stratified analysis to compute pooled Odds Ratios (ORs) with 95% Confidence Intervals (CIs). Statistical significance was set at $P < 0.05$.

Key Results

• Demographic Disparities: Significant differences were observed between urban and rural participants regarding age distribution ($P=0.001$), educational level ($P=0.002$), marital status ($P=0.001$), and occupation ($P=0.004$). Specifically, rural participants had a higher proportion of adolescents (11.1% vs. 3.9%) and unmarried women (18% vs. 11.2%).
• ANC Utilization Indicators: Contrary to prevailing assumptions, no statistically significant differences were found between urban and rural facilities regarding:
  • Timing of first booking ($P=0.377$): 67.3% of urban and 70.4% of rural women initiated care after 12 weeks.
  • Adequacy of care ($P=0.349$): 90.6% of urban and 92.5% of rural women received adequate care ($\ge$ 4 visits).
  • Mean number of visits ($P=0.100$): Urban mean was 7.12 ($\pm$ 2.55) and rural mean was 6.83 ($\pm$ 2.55).
• Determinants of Utilization: In the stratified Mantel–Haenszel analysis, marital status was the only factor independently associated with adequate ANC utilization. Married women had significantly higher odds of adequate utilization compared to unmarried women (MH OR = 2.06; 95% CI: 1.21–3.51; $P=0.007$). This association was driven primarily by rural women.
• Non-Significant Factors: Maternal age, educational status, ethnicity, and occupation showed no statistically significant association with ANC utilization in the stratified analysis.

Key Contributions
The paper makes three primary contributions to the literature:

1. Methodological Rigor: It utilizes a comparative cross-sectional design with Mantel-Haenszel stratification to control for setting-specific confounders, providing a replicable model for future research.
2. Identification of Determinants: By controlling for urban-rural confounding, the study identifies marital status as a significant, independent determinant of adequate ANC utilization, highlighting the role of social support structures.
3. Challenging Assumptions: The study challenges the prevailing narrative that urban-rural disparities in ANC utilization are ubiquitous. It demonstrates that in the Bono Region of Ghana, utilization metrics (timing, adequacy, and frequency) do not significantly differ between urban and rural health facilities, suggesting potential success in equitable service distribution in this specific context.

Significance and Conclusion
The study concludes that while ANC utilization is high in both urban and rural facilities in the Bono Region, a critical delay in initiation (beyond the first trimester) persists across both settings. The lack of significant urban-rural difference in utilization metrics suggests that access barriers may have been mitigated in this region, possibly due to equitable service distribution, comparable health literacy, or transport improvements.

However, the independent association between marital status and adequate utilization underscores a specific vulnerability for unmarried pregnant women. The authors posit that unmarried women may face stigma, lack of partner involvement, or economic insecurity that hinders service uptake. Consequently, the study advocates for interventions that prioritize early ANC initiation and specifically address social barriers affecting unmarried women, rather than focusing solely on geographic disparities. The findings support evidence-based policymaking to achieve Sustainable Development Goals (SDG 3 and 5) by ensuring healthy lives and gender equality.