Geographic Analysis of Emergency Newborn Referral Patterns in Southern Ghana and Implications for Neonatal Survival

Abstract Introduction: From 1990 to 2020, neonatal mortality declined by 52% globally and 45% in Ghana. Yet this fell short of the national goal. Critical contributors to this shortfall include limited and unreliable access to Emergency Obstetric and Newborn Care (EmONC). Effective and efficient health care referral networks are an integral but under-studied component of EmONC. Objectives: This study aimed to assess the longitudinal and geographic patterns in newborn referral rates in Making Every Baby Count Initiative (MEBCI), a project that aimed to improve the quality of facility-based care for newborns in 4 administrative regions through clinical training, facility readiness, and advocacy. Methods: 155 health facilities from the 4 targeted administrative regions were included in the sample. A multilevel Poisson growth curve model was used to examine changes in newborn referral rates over time, and correlation analysis was conducted to assess the association between travel time and referral rates. A geographic information system was used to estimate and visualize the average travel time with a 4-wheeled automobile. Results: The overall rate of newborns referred out of MEBCI-participating facilities increased over time. Mean travel times from referring facilities to their respective regional hospitals varied across regions (lowest: 78mins, highest: 120mins). Overall, across the 4 MEBCI-participating regions, there was a negative correlation between referral rates and travel time (r = -0.37, p < .001). Conclusion: The MEBCI-supported facilities experienced statistically significant overall increases in newborn referral rates over time. Facilities furthest from the regional hospitals experienced the lowest referral rates, suggesting that long distances and travel times may have influenced providers’ decisions to refer or patients’ acceptance of referral.