Sociodemographic determinants of maternal health indicators in conflict-affected counties of Kenya: secondary analysis of data from the 2022 Kenya demographic and health survey

A secondary analysis of the 2022 Kenya Demographic and Health Survey reveals that maternal health outcomes in conflict-affected regions are significantly hindered by sociodemographic inequities, with adolescents, rural residents, and socioeconomically disadvantaged women facing the greatest barriers to adequate antenatal care, facility delivery, and skilled birth attendance.

The Gist

Imagine Kenya's maternal health system as a massive, complex train network. In a peaceful world, the train runs on time, stops at every station, and everyone can buy a ticket to reach their destination: a safe birth.

But in conflict-affected counties (areas where fighting and violence have recently happened), the tracks are broken, the stations are damaged, and the "ticket prices" have skyrocketed—not just in money, but in safety and courage.

This paper is a report card on who is still able to catch that train and who is being left behind on the platform. The researchers looked at data from 1,060 women in six of Kenya's most troubled counties (like Mandera, Samburu, and Lamu) to see what factors determine if a mother gets the care she needs.

Here is the story of their findings, broken down into simple concepts:

1. The "Ticket" to Safety: Education and Money

Think of education and wealth as the VIP passes that let you through the gates.

• Education is the Master Key: The study found that education was the single most powerful tool a woman had. A woman with a high school education was nearly three times more likely to get the recommended number of doctor visits (4+ visits) and give birth in a hospital compared to a woman with no schooling. In a chaotic environment, education acts like a flashlight; it helps women navigate the darkness, understand the risks, and find the safe path to a clinic.
• Wealth is the Fuel: Money matters a lot. In these conflict zones, getting to a hospital isn't just about the bus fare. It's about paying for a safe vehicle that won't get stopped by armed groups, or paying for a security escort. Women in the "middle" and "richer" groups were 5 to 11 times more likely to deliver in a hospital than the poorest women. For the poorest, the "hidden costs" of conflict (danger, distance, bribes) make the ticket unaffordable.

2. The "Teenage Detour"

Imagine a young teenager trying to navigate a maze while everyone else is walking confidently.

• The Problem: Teenagers (ages 15–19) were the least likely group to get adequate prenatal care.
• Why? They face a "perfect storm" of barriers: they are often too shy to ask for help, they lack money, they are dependent on parents who might not support them, and they face stigma. In a conflict zone, where the system is already shaky, these young mothers get lost in the cracks. They start the journey but often drop out before reaching the finish line.

3. The "Rural Wall"

Imagine living in a village that is separated from the city by a wide, dangerous river.

• The Reality: Women living in rural areas were 70% less likely to give birth in a hospital than women in towns.
• The Metaphor: In peaceful times, the river might be shallow. But in conflict, the river floods, the bridges are destroyed, and the water is full of crocodiles (insecurity). Even if a woman wants to go to the hospital, the journey becomes impossible. The "Three Delays" model explains this: the first delay is deciding to go, the second is reaching the facility, and the third is getting care. In rural conflict zones, the second delay is a massive, insurmountable wall.

4. The "County Lottery"

Not all conflict zones are the same.

• The Variation: The study looked at six counties. In some (like Lamu), nearly everyone got good care. In others (like Mandera or Samburu), the rates were much lower.
• The Lesson: Conflict doesn't hit everyone equally. It's like a storm that hits some houses harder than others depending on how strong the roof is (local leadership, infrastructure, and humanitarian aid). National averages hide these differences; you can't fix a problem if you treat every county exactly the same.

5. The Surprising Twist: Violence at Home

You might expect that women suffering from domestic violence (Intimate Partner Violence) would be the least likely to seek help.

• The Finding: Surprisingly, the data showed that domestic violence did not statistically predict whether a woman got care or not.
• The Explanation: It's not that violence doesn't hurt; it's that the biggest monsters in the room are poverty, lack of education, and the war itself. These structural barriers are so huge that they "drown out" the effect of domestic violence. It's like trying to hear a whisper (domestic violence) while a jet engine (conflict and poverty) is roaring right next to you. The jet engine is what stops the plane from taking off.

The Bottom Line

This paper tells us that in Kenya's war-torn areas, motherhood is becoming a luxury item.

• Who is left behind? Teenagers, the poor, the uneducated, and rural women.
• Who makes it? Those with money, education, and access to towns.

The Solution? We can't just build more hospitals. We need to:

1. Build bridges for the young: Create special, friendly, non-judgmental clinics just for teenagers.
2. Fix the roads: Invest in safe transport and security so rural women can actually reach the clinics.
3. Hand out flashlights: Keep girls in school. Education is the best protection a mother has in a crisis.

The authors conclude that until we address these deep inequalities, the "train" of maternal health will continue to leave the most vulnerable passengers behind on the platform.

Technical Summary

1. Problem Statement

Maternal health outcomes in low- and middle-income countries are often characterized by systemic inequities. These disparities are exacerbated in conflict-affected settings where health infrastructure is damaged, governance is weakened, and mobility is restricted. While national-level data exists, there is a critical gap in disaggregated evidence regarding how specific sociodemographic determinants (age, education, wealth, residence) influence maternal health indicators specifically within Kenya's conflict-affected counties. Understanding these localized gradients is essential for designing equity-driven interventions in fragile environments.

2. Methodology

• Study Design: A descriptive, cross-sectional secondary analysis.
• Data Source: The 2022 Kenya Demographic and Health Survey (KDHS), a nationally representative survey conducted by the Kenya National Bureau of Statistics (KNBS).
• Study Population: 1,060 women aged 15–49 years who had a live birth in the two years preceding the survey and resided in conflict-affected counties.
• Conflict Identification: Six counties were classified as "conflict-affected" based on the Armed Conflict Location & Event Data (ACLED) database. The criteria were the occurrence of organized violence resulting in >25 fatalities within the 24 months prior to the survey. The identified counties were: Lamu, Mandera, Marsabit, Isiolo, Samburu, and Elgeyo Marakwet.
• Outcome Variables (Dependent):
  1. Adequate Antenatal Care (ANC): Attendance of 4 or more visits.
  2. Facility Delivery: Birth occurring in a public, private, or faith-based health facility.
  3. Skilled Birth Attendance (SBA): Delivery assisted by a doctor, nurse, or midwife.
• Predictor Variables (Independent): Maternal age, educational attainment, household wealth index (quintiles), employment status, place of residence (urban/rural), county of residence, and sex of household head.
• Covariates: Intimate Partner Violence (IPV) experiences (physical, emotional, sexual) were included to adjust for potential confounding.
• Statistical Analysis:
  • Weighted descriptive statistics and Chi-square tests for bivariable associations.
  • Multivariable logistic regression to calculate Adjusted Odds Ratios (aOR) with 95% Confidence Intervals (CI).
  • Significance level set at p < 0.05.
  • Software: Stata.

3. Key Contributions

• Contextual Specificity: This is one of the first studies to isolate and analyze maternal health determinants specifically within the six most conflict-affected counties of Kenya, moving beyond national averages that often mask sub-national disparities.
• Integration of Conflict Metrics: The study rigorously links KDHS data with ACLED fatality data to objectively define the "conflict-affected" sample, providing a robust framework for analyzing health in fragile states.
• IPV Analysis in Fragile Contexts: It uniquely investigates the role of Intimate Partner Violence (IPV) in conflict zones, finding that structural barriers (poverty, geography) may overshadow interpersonal violence as a predictor of service utilization in these specific settings.
• Granular Inequity Mapping: The study provides a detailed "inequity gradient," demonstrating how specific sociodemographic factors interact with conflict to create distinct barriers for different sub-populations (e.g., adolescents vs. older women, rural vs. urban).

4. Key Results

General Prevalence:

• ANC 4+: 53.5% (despite 90.2% attending at least one visit).
• Facility Delivery: 68.2%.
• Skilled Birth Attendance (SBA): 72.2%.

Determinants of Adequate ANC (4+ visits):

• Education: The strongest predictor. Women with higher education had 2.74 times higher odds of adequate ANC compared to those with no education (aOR=2.74; 95% CI: 1.19–6.34).
• Age: Adolescents (15–19) were the least likely to achieve adequate ANC. Women aged 20–24 had significantly higher odds (aOR=1.83) than teenagers.
• County: Significant disparities existed; women in Mandera (aOR=0.39) and Isiolo (aOR=0.49) had significantly lower odds compared to Lamu.
• IPV: No statistically significant association was found between IPV and ANC utilization in the adjusted model.

Determinants of Facility Delivery:

• Wealth: A strong gradient was observed. Women in the middle wealth quintile were 5.5 times more likely to deliver in a facility than the poorest (aOR=5.50; 95% CI: 2.14–14.14).
• Residence: Rural residence was a profound barrier, reducing the odds of facility delivery by 68% (aOR=0.32; 95% CI: 0.15–0.67) compared to urban residents.
• Education: Secondary education increased odds of facility delivery (aOR=2.72).
• County: Mandera (aOR=0.24), Marsabit (aOR=0.32), and Samburu (aOR=0.28) showed significantly lower facility delivery rates compared to Lamu.

Determinants of Skilled Birth Attendance (SBA):

• Wealth: The "Richer" quintile had 11 times higher odds of SBA compared to the poorest (aOR=11.04; 95% CI: 2.06–59.25).
• Residence: Rural women had 78% lower odds of SBA (aOR=0.22).
• Education: Primary education significantly increased SBA odds (aOR=2.86).
• IPV: Similar to ANC, IPV was not independently associated with SBA after adjusting for sociodemographics.

5. Significance and Implications

• Structural Inequities Dominate: The study concludes that in conflict-affected Kenya, structural determinants (wealth, education, geography) are the primary drivers of maternal health outcomes, often overshadowing individual factors like IPV. The "Three Delays" framework is amplified by conflict, where insecurity acts as a "distance multiplier."
• Adolescent Vulnerability: Teenagers are systematically excluded from adequate care, necessitating targeted, adolescent-friendly interventions that address stigma and mobility barriers.
• The "Cost" of Conflict: The findings suggest that safe motherhood is becoming an economic commodity. In conflict zones, the "hidden costs" (transportation, security, informal payments) prevent the poorest and rural populations from accessing free or subsidized government services (e.g., Linda Mama program).
• Policy Recommendations:
  • Targeted Sub-National Planning: National averages are insufficient; interventions must be tailored to specific county realities (e.g., Mandera and Samburu require different strategies than Lamu).
  • Adolescent Continuity: Strengthening ANC continuity for teenagers is critical.
  • Rural Access: Investment in transport and mobile health services is required to overcome geographic isolation.
  • Education & Empowerment: Sustained investment in female secondary education and economic empowerment is identified as a fundamental "survival strategy" for maternal health in fragile settings.

Limitations: The study relies on cross-sectional data (no causality), potential underreporting of IPV due to stigma, and ACLED data which measures fatalities but may not capture the "fear" of violence that also suppresses service utilization.