Clinical outcomes and mortality risk among inborn and referred newborns admitted to hospitals in Kenya

This study of over 130,000 newborns in Kenya reveals that referred infants face nearly three times higher mortality odds than inborn infants, a disparity driven by greater clinical vulnerability and systemic challenges in the referral process that underscores the urgent need for strengthened pre-referral stabilization and transport systems.

The Gist

The Big Picture: A Tale of Two Newborns

Imagine a hospital as a fortress designed to protect the most vulnerable people in the world: newborn babies. This study looked at two types of babies entering this fortress in Kenya:

1. The "Home Team" (Inborn): Babies who were born inside the fortress walls. They were handed directly to the doctors and nurses waiting for them.
2. The "Travelers" (Referred): Babies who were born elsewhere (in smaller clinics or at home) and had to be transported to the fortress for help because they were too sick or too small to stay where they were.

The researchers wanted to know: Do the Travelers survive at the same rate as the Home Team?

The Shocking Discovery

The answer was a resounding no.

• The Home Team: About 11 out of 100 babies born inside the hospital died.
• The Travelers: About 29 out of 100 babies who arrived from other places died.

The Analogy: Imagine a race where the Home Team starts right at the finish line with a full tank of gas. The Travelers start miles away, driving a car with a flat tire, on a bumpy road, and they arrive late. It's no surprise they struggle more, but the study shows the gap is huge—Travelers were three times more likely to die than the Home Team, even after the researchers accounted for how sick the babies were when they arrived.

Why Are the Travelers in Such Danger?

The study found that the Travelers were already in a much worse spot when they arrived. Think of it like this:

• The "Sicker" Selection: The Travelers weren't just random babies; they were the ones who were already in critical condition. They were more likely to be born extremely early (premature) and weigh almost nothing (extremely low birth weight).
• The "Journey" Problem: By the time these babies got to the big hospital, they had already lost precious time.
  • The Delay: 80% of Home Team babies were admitted within 24 hours of birth. Only 60% of Travelers were.
  • The Rough Ride: The journey to the hospital might have been unsafe. Maybe the ambulance didn't have oxygen, maybe the baby got cold (hypothermia) during the trip, or maybe the baby wasn't stabilized (stabilized like a wobbly table) before leaving the first clinic.
  • The "Triage" Gap: When they finally arrived, they were often already in respiratory distress (trouble breathing) or hypoxic (not getting enough oxygen).

The "Risk Factors" (The Villains)

The researchers identified specific "villains" that made it harder for these babies to survive, regardless of where they were born:

1. Tiny Size: Babies weighing less than 1kg (2.2 lbs) had a massive risk of death. It's like trying to run a marathon with a broken leg; their bodies are just too fragile.
2. Breathing Trouble: If a baby couldn't breathe well on arrival, their odds of survival dropped sharply.
3. The Cold: If a baby arrived cold (hypothermia), it was a sign they hadn't been cared for properly during the journey.
4. Birth Complications: Babies who had trouble during birth (like lack of oxygen) or had birth defects faced higher risks.

What Does This Mean for the Future?

The study suggests that the problem isn't just that the Travelers are sicker; it's that the system of getting them to safety is broken.

The Solution: The "Hub-and-Spoke" Model
Imagine a bicycle.

• The Hub is the big hospital with all the fancy equipment and specialists.
• The Spokes are the smaller clinics and rural hospitals.

Right now, the spokes are disconnected. The study suggests we need to strengthen the spokes so they can keep the babies stable before they even leave.

• Better Training: Teach the nurses at the small clinics how to warm the baby and help them breathe before calling for an ambulance.
• Better Ambulances: Ensure the transport vehicles have oxygen and heating.
• Faster Decisions: Don't wait until the baby is critical to send them; send them earlier.

The Bottom Line

This paper is a wake-up call. In Kenya, if a baby is born in a big hospital, they have a fighting chance. If that same baby is born in a small clinic and has to be moved, their chances drop drastically.

It's not just about building more big hospitals; it's about fixing the bridge between the small clinics and the big ones. If we can make the journey safer and faster, we can save thousands of lives and get closer to the global goal of ending preventable newborn deaths.

Technical Summary

1. Problem Statement

Neonatal mortality remains critically high in sub-Saharan Africa, with the region failing to meet Sustainable Development Goal (SDG) target 3.2 (ending preventable newborn deaths by 2030). A significant disparity exists in outcomes between inborn newborns (born and managed within the same facility) and referred newborns (transferred from other facilities). While referred newborns are often sicker due to selection bias, evidence suggests that systemic failures in the referral pathway—such as inadequate pre-referral stabilization, unsafe transport, and admission delays—may further exacerbate mortality risks. There is a lack of large-scale, routine clinical data quantifying these disparities and identifying specific risk factors within the Kenyan context to inform targeted health system interventions.

2. Methodology

• Study Design: A retrospective observational analysis of routine clinical data.
• Setting: 13 hospitals across 10 counties in Kenya implementing the NEST360 (Newborn Essential Solutions and Technologies 360) program, a health systems strengthening initiative.
• Data Source: The Neonatal Inpatient Dataset (NID), collected via the Clinical Information Network (CIN) using REDCap. Data covers the period from January 2019 to October 2024.
• Population:
  • Total Analyzed: 130,773 newborns.
  • Inclusion Criteria: Newborns admitted to the 13 hospitals, born in a health facility, and aged <28 days at admission.
  • Exclusion Criteria: Home births (n=1,724), missing birth location/age data (n=15,273), and admissions >28 days.
  • Groups: 114,084 (87.2%) Inborn vs. 16,689 (12.8%) Referred.
• Variables:
  • Demographics/Perinatal: Age at admission, sex, gestational age, birthweight, mode of delivery, singleton/multiple birth.
  • Clinical: Respiratory distress, hypoxia (SpO2 <90%), temperature (hypothermia/fever), primary diagnosis.
  • Maternal: Age, HIV status, survival status.
• Statistical Analysis:
  • Descriptive statistics (frequencies, medians, IQRs) to compare characteristics.
  • Univariable Logistic Regression: To assess associations between predictors and mortality, stratified by birth location.
  • Multivariable Logistic Regression: Forward stepwise selection (entry p-value <0.25) to calculate Adjusted Odds Ratios (aOR) for factors independently associated with mortality.
  • Software: Stata version 19.

3. Key Contributions

• Scale and Scope: This is one of the largest studies in Kenya to compare inborn and referred newborns using routine clinical data, covering over 130,000 admissions.
• Granular Risk Stratification: It moves beyond simple mortality rates to dissect specific risk factors (e.g., extremely low birthweight, respiratory distress) within the referred population, highlighting that referred infants are not only sicker but face higher mortality risks even after adjusting for these conditions.
• Systemic Insight: The study provides empirical evidence linking the "referred" status to a nearly three-fold increase in mortality, suggesting that the referral process itself (transport, stabilization, triage) is a critical point of failure in the continuum of care.

4. Key Results

• Mortality Disparity:
  • Referred Newborns: 29.0% mortality rate (4,835 deaths).
  • Inborn Newborns: 11.3% mortality rate (12,900 deaths).
  • Adjusted Risk: In multivariable analysis, being referred was independently associated with 2.54 times higher odds of death (aOR 2.54, 95% CI 2.39–2.71) compared to inborn newborns.
• Clinical Characteristics:
  • Referred newborns were significantly more likely to be extremely preterm (6.1% vs 3.1%) and extremely low birthweight (4.6% vs 2.6%).
  • They presented with higher rates of respiratory distress (26.2% vs 15.0%) and hypoxia (23.2% vs 15.3%).
  • Admission Timing: Only 59.6% of referred newborns were admitted on Day 1 of life, compared to 80.2% of inborn newborns, indicating significant delays in transfer.
• Risk Factors for Mortality (Referred Group):
  • Extremely Low Birthweight (<1,000g): OR 13.57 (95% CI 11.19–16.44).
  • Respiratory Distress: OR 4.07 (95% CI 3.77–4.39).
  • Congenital Anomalies: OR 1.66 (95% CI 1.41–1.95).
  • Intrapartum-related Complications: OR 1.35 (95% CI 1.20–1.52).
• Diagnoses: Prematurity was the most common diagnosis overall (38.4%), but significantly higher in the referred group (48.8%) compared to inborn (36.8%).

5. Significance and Implications

• Health System Performance: The mortality gap highlights critical inefficiencies in the referral ecosystem, including delays in decision-making, lack of pre-referral stabilization, and potential gaps in transport safety.
• Policy Recommendations:
  • Hub-and-Spoke Model: Strengthening networks where lower-level facilities (spokes) are linked to tertiary centers (hubs) for capacity building and oversight.
  • Referral Protocols: Implementation of clear referral criteria, standardized pre-referral stabilization checklists, and improved transport logistics.
  • Data-Driven Quality Improvement: Utilizing routine data (like NID) to monitor referral times and outcomes to drive continuous quality improvement.
• Future Research: The authors call for studies focusing on the "time-to-transfer" metric, the quality of pre-referral care, and linking neonatal data with maternal records to better understand antenatal risk factors.

Conclusion: The study concludes that while referred newborns are inherently sicker, the nearly three-fold increase in mortality suggests that systemic gaps in the referral continuum significantly compound this vulnerability. Addressing these gaps through improved coordination, stabilization, and transport is essential to achieving national and global neonatal survival targets.