Socioeconomic inequality in maternal healthcare utilisation in Ghana: evidence from concentration index decomposition of the 2022 Demographic and Health Survey

Despite high overall coverage of basic antenatal care in Ghana, significant pro-rich socioeconomic inequalities persist in maternal healthcare utilization, driven primarily by rural residence and educational disparities, which the National Health Insurance Scheme alone fails to eliminate.

The Gist

Imagine Ghana's healthcare system as a massive, nationwide bus network designed to get everyone to the hospital safely when they are about to give birth. The government built this network with two major upgrades: a National Health Insurance Scheme (NHIS) to pay the fare, and a Free Maternal Policy to make the ride free for pregnant women.

The goal was simple: get every woman, rich or poor, to ride the bus to a hospital.

However, this new study acts like a traffic camera that took a snapshot of the system in 2022. The camera reveals a surprising truth: even though the bus is free, the passengers aren't evenly distributed. The "rich" are still taking the luxury limousines (private hospitals), while the "poor" are often still walking the long, dusty road to deliver at home.

Here is the breakdown of what the study found, using simple analogies:

1. The "Good News" vs. The "Real News"

• The Good News: Most women are getting on the bus. About 88% of women visited a doctor at least 4 times during pregnancy. This is a huge success.
• The Real News: When you look closer at the quality of the ride, the gap widens. The World Health Organization (WHO) says women should visit the doctor 8 times for a truly safe pregnancy.
  • The Analogy: Imagine a marathon. Everyone can run the first 4 miles (the basic checkups). But when the race gets to the full 8 miles, the rich runners are sprinting ahead, while the poor runners are struggling to keep up. The study found that the inequality in getting those full 8 visits is four times worse than the inequality in just getting 4 visits.

2. Where Do Women Give Birth? (The Destination)

The study looked at where women actually delivered their babies. Think of this as choosing between three destinations:

• The Home (The Dirt Road): 46.7% of the poorest women still delivered at home.
• The Public Hospital (The Community Bus): This is the main stop for the middle class and the poor. It's reliable and free, but often crowded.
• The Private Hospital (The Luxury Limo): 46.1% of the richest women delivered here. They pay extra for comfort, shorter wait times, and better service.

The Takeaway: The system works like a sorting machine. If you are poor, you are likely to walk the dirt road or take the crowded bus. If you are rich, you are likely to take the limo.

3. Why is this happening? (The Decomposition)

The researchers used a special tool called a "Concentration Index Decomposition" to figure out why this gap exists. Think of this as a recipe for inequality. They broke down the "cake" of inequality to see which ingredients made it taste so bad.

Here are the main ingredients they found:

• The "Rural" Ingredient (16.4%): Living in the countryside is a huge barrier. It's like living 50 miles away from the bus stop. Even if the bus is free, the walk to get there is too long and dangerous.
• The "Insurance" Paradox (16.4%): This is the most surprising finding. The researchers found that having health insurance actually contributed to the inequality.
  • The Analogy: Imagine a VIP club that is free to enter, but the bouncer only lets in people who have a specific ID card. It turns out the rich people have the ID cards, but the poor people don't (maybe they can't afford the paperwork or the hidden fees). So, the "free" club ends up being mostly for the rich, making the gap wider.
• The "Region" Ingredient (15.6%): Where you live in Ghana matters. Some areas are just better connected than others.
• The "Education" Hero (-22.5%): This is the only ingredient that reduced inequality.
  • The Analogy: Education is like a superpower. Women with secondary education are more likely to know how to navigate the system, demand better care, and make the trip to the hospital, regardless of how much money they have. It acts as a great equalizer.

4. What Should We Do? (The Fix)

The study suggests that just giving out "free tickets" (free healthcare) isn't enough. We need to fix the whole system:

1. Fix the Roads: Build better clinics in rural areas so the "bus stop" is closer to the poor.
2. Fix the VIP Club: Make it easier for the poorest women to get their insurance cards. Remove the hidden paperwork barriers.
3. Upgrade the Public Bus: Make the public hospitals so good (clean, respectful staff, no waiting) that even the rich people want to ride them. This stops the rich from leaving for the limos.
4. Invest in Girls' Schools: Keep girls in school. Education is the most powerful tool to ensure a mother gets the care she needs.

The Bottom Line

Ghana has built a fantastic healthcare system, but it's like a garden where the rich flowers are getting all the water and sunlight, while the poor flowers are still struggling in the shade. The government has removed the financial cost, but they haven't removed the distance, the paperwork, or the quality gaps. To truly help everyone, they need to water the whole garden, not just the top layer.

Technical Summary

1. Problem Statement

Despite Ghana's implementation of the National Health Insurance Scheme (NHIS) in 2003 and the Free Maternal Healthcare Policy (FMHCP) in 2008 to remove financial barriers and promote Universal Health Coverage (UHC), socioeconomic disparities in maternal healthcare utilization likely persist. Previous literature indicates pro-rich gradients in skilled birth attendance and antenatal care (ANC), but there is a lack of recent evidence using the most current nationally representative data (2022 GDHS) to quantify these inequalities and decompose their drivers. Specifically, it remains unclear whether financial policies have successfully neutralized the influence of wealth, education, and geography on maternal health outcomes, or if structural barriers continue to drive inequality.

2. Methodology

Data Source and Sample:

• Data: The study utilized the 2022 Ghana Demographic and Health Survey (GDHS), a nationally representative cross-sectional survey.
• Sample: Restricted to women aged 15–49 who reported a live birth in the five years preceding the survey ($n = 5,134$; weighted population $\approx$ 4.66 million).
• Missing Data: Handled using Multiple Imputation by Chained Equations (MICE) with 20 datasets to preserve statistical power and reduce bias.

Variables:

• Outcomes:
  1. Adequate ANC: Defined as $\ge$4 visits (standard) and $\ge$8 visits (WHO recommended threshold, used in sensitivity analysis).
  2. Place of Delivery: Categorized as Home, Public Facility, or Private Facility.
• Explanatory Variables: Grouped into socioeconomic position (wealth quintile, education, residence), demographics (age, birth order, marital status), access barriers (distance, money, permission), and health system factors (NHIS membership, media exposure, region).

Statistical Analysis:

• Survey Design: All analyses accounted for complex survey design (weights, strata, primary sampling units) using Stata 18.0.
• Multinomial Logistic Regression: Used to identify factors associated with place of delivery (reference: home delivery), reporting Relative Risk Ratios (RRR).
• Concentration Index (CI): Calculated to measure socioeconomic inequality in adequate ANC. The Wagstaff (2005) normalization correction was applied to adjust for the binary nature of the outcome.
• Decomposition Analysis: The Wagstaff et al. (2003) method was employed to decompose the CI. This quantified the contribution of each determinant (e.g., wealth, education, NHIS) to the total inequality, distinguishing between "pro-rich" contributors (increasing inequality) and "equalizing" contributors (reducing inequality).
• Sensitivity Analyses: Included re-estimation using the $\ge$8 ANC threshold, complete-case analysis, and substitution of NHIS card possession for membership.

3. Key Results

Coverage and Distribution:

• ANC Coverage: 88.6% of women achieved $\ge$4 ANC visits. However, the Concentration Index (CI) for this metric was 0.0391 ($p < 0.001$), indicating statistically significant pro-rich inequality.
• WHO Threshold Impact: When applying the stricter WHO standard ($\ge$8 visits), the CI increased more than fourfold to 0.1728, revealing that adherence to the full recommended schedule is heavily concentrated among the wealthy.
• Place of Delivery: A sharp socioeconomic gradient exists.
  • Home Delivery: Most prevalent among the poorest (46.7%) and declines to 4.1% among the richest.
  • Private Facility Delivery: Dominates among the richest (46.1%) but is rare among the poorest (9.3%).
  • Public Facility Delivery: Serves as a safety net, remaining relatively stable across wealth quintiles (15.9%–21.7%), though predicted probability drops for the richest.

Regression Findings (Place of Delivery):

• Wealth: The richest women were >1,000 times more likely to deliver in a private facility compared to the poorest (RRR = 1,013.3).
• NHIS Membership: Associated with a 76% reduction in the odds of home delivery vs. public facility (RRR = 0.24), yet it did not eliminate the wealth gradient.
• Barriers: Distance and financial problems were significant predictors of home delivery.

Decomposition of Inequality (ANC):

• Explained Variance: The included determinants explained 71.3% of the total inequality.
• Major Pro-Rich Drivers (Increasing Inequality):
  • Rural Residence: 16.4% contribution.
  • NHIS Membership: 16.4% contribution. Crucially, while NHIS increases utilization, it contributes to inequality because enrollment is disproportionately concentrated among wealthier women.
  • Geographical Region: 15.6% contribution.
• Major Equalizing Drivers (Reducing Inequality):
  • Secondary Education: The strongest equalizing factor, contributing -22.5% to inequality reduction.
  • Primary Education: -6.4%.

4. Key Contributions

1. Policy-Relevant Evidence: This is the first study to use 2022 GDHS data to decompose maternal health inequalities in Ghana, providing an up-to-date assessment of the NHIS and FMHCP impacts.
2. Methodological Rigor: The study applies Wagstaff decomposition with appropriate normalization for binary outcomes and utilizes multiple imputation to handle missing data, ensuring robust estimates.
3. Nuanced Insight on NHIS: It challenges the assumption that health insurance automatically promotes equity. The study demonstrates that while NHIS improves average utilization, its stratified enrollment (higher uptake among the wealthy) actually amplifies pro-rich inequality in ANC.
4. Threshold Sensitivity: The study highlights that monitoring ANC via the minimum $\ge$4 visit threshold masks severe inequality, which becomes pronounced when the WHO $\ge$8 contact standard is applied.

5. Significance and Implications

• Structural Barriers Persist: Financial policies alone are insufficient. Geographic inaccessibility (rural residence) and regional disparities remain dominant drivers of inequality.
• Quality-Driven Sorting: Wealthier women are "sorting" into private facilities, likely due to perceived quality differences in public facilities (e.g., wait times, drug availability). This suggests a need for public sector quality improvement to retain cross-socioeconomic trust.
• Education as a Lever: Secondary education is a powerful equalizer. Policies promoting girls' secondary education could yield significant co-benefits for maternal health equity.
• Policy Recommendations:
  • Targeted NHIS Enrollment: Shift from passive to active enrollment strategies (community outreach, simplified registration) to reach the poorest quintiles.
  • Infrastructure Investment: Upgrade rural health infrastructure (CHPS compounds, maternity waiting homes) to address distance barriers.
  • Monitoring Reform: Adopt the WHO $\ge$8 ANC contact standard for monitoring to accurately capture disparities in care quality.
  • Public Sector Quality: Invest in provider training and resource availability to prevent the "quality-driven exit" of wealthier populations from the public system.

In conclusion, while Ghana has achieved high basic coverage, significant pro-rich inequalities persist. Achieving true Universal Health Coverage requires addressing non-financial barriers, improving public facility quality, and ensuring equitable access to insurance and education.