Mapping the spatial distribution of harmful umbilical cord stump care among neonates in Ethiopia: A spatial with multilevel analysis

Introduction The umbilical cord (UC) serves as the main pathway for bacteria to reach the neonate’s body, potentially causing local and severe infections, sepsis, and even death. Consequently, neonatal mortality remains a significant public health concern, particularly in Ethiopia. The World Health Organization (WHO) recommends that the umbilical cord stump be kept clean and dry, with the exception of applying topical antiseptics. However, various harmful substances are still applied to the umbilical cord of neonates. Data on the geographical distribution and risk factors for harmful umbilical cord stump (UCS) care are scarce. Therefore, this study aims to fill this gap. Methods A secondary data analysis of the Ethiopian Demographic Health Survey (EDHS 2016) was conducted using a weighted sample of 7,168 live births. ArcGIS version 10.7.1 software was utilized to visualize the spatial distribution of harmful umbilical cord stump (UCS) care practices in Ethiopia. Additionally, a Bernoulli probability model-based spatial scan statistic was applied using Kulldorff’s SaTScan version 9.6 software to identify significant clusters of harmful UCS care. A multilevel logistic regression model was used to determine the factors associated with UCS care practices in Ethiopia. Statistical significance was declared at a two-sided P-value of < 0.05. Results Overall, the prevalence of harmful UCS care in Ethiopia was 15.09% (95% CI: 13.9–16.3), with significant spatial heterogeneity across geographical areas. The hotspot areas of harmful US care were observed in the eastern (Somali) and northern (Tigray and Amhara) parts of Ethiopia. In spatial scan analysis, the most likely primary clusters were observed in South Nation Nationalities and Peoples region (SNNPR), secondary clusters in the Somali, tertiary clusters in Tigray, and the next clusters in the Amhara regions, respectively. In the final multilevel model, maternal age (Adjusted odds ratio/AOR 1.07, 95% CI: 1.02–1.12), institutional delivery (AOR 0.64, 95% CI: 0.42–0.97), female neonates (AOR 1.31, 95% CI: 1.04–1.61), rural residence (AOR 2.18, 95% CI: 1.05–4.52), living in Tigray region (AOR 3.79, 95% CI: 1.38–9.38), living in Somali region (AOR, 2.95% CI: 1.02–8.52), and living in Harari region (AOR 3.51, 95% CI: 1.28–9.60) were identified as a significant factors of harmful US care practice in Ethiopia. Conclusion In Ethiopia, the distribution of harmful UCS care practices is non-random and highly clustered in the SNNPR, Somalia, Tigray, and Amhara regions. Both individual and community-level factors were significantly associated with the practice. Special emphasis needs to be provided for neonates from those hot-spot areas and to address the identified predictors of harmful umbilical cord stump care practices.