Using geographically weighted regression analysis to assess predictors of home birth hot spots in Ethiopia

• Published in: PloS one 2023-06, Vol.18 (6), p.e0286704-e0286704
• Main Authors: Hailegebreal, Samuel, Haile, Firehiwot, Haile, Yosef, Simegn, Atsedu Endale, Enyew, Ermias Bekele
• Format: Article
• Language: English
• Subjects: Analysis; Attended births; Biology and Life Sciences; Birth; Childbirth; Childbirth & labor; Childbirth at home; Distribution; Earth Sciences; Education; Educational Status; Ethiopia - epidemiology; Evaluation; Female; Geographical variations; Health care; Health care industry; Health facilities; Health services; Home births; Home Childbirth; Humans; Infant mortality; Least squares method; Maternal mortality; Medicine and Health Sciences; Nutrition research; People and Places; Pregnancy; Pregnant women; Prenatal Care; Public health; Regression analysis; Religion; Sepsis; Skilled labor; Social Sciences; Software; Spatial Analysis; Spatial Regression; Statistical analysis; Statistics; Surveys; Variables; Womens health
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0286704

Annually, 30 million women in Africa become pregnant, with the majority of deliveries taking place at home without the assistance of skilled healthcare personnel. In Ethiopia the proportion of home birth is high with regional disparity. Also limited evidence on spatial regression and deriving predictors. Therefore, this study aimed to assess the predictors of home birth hot spots using geographically weighted regression in Ethiopia. This study used secondary data from the 2019 Ethiopian Mini Demographic and Health Survey. First, Moran's I and Getis-OrdGi* statistics were used to examine the geographic variation in home births. Further, spatial regression was analyzed using ordinary least squares regression and geographically weighted regression to predict hotspot area of home delivery. According to this result, Somalia, Afar, and the SNNPR region were shown to be high risk locations for home births. Women from rural residence, women having no-education, poorest wealth index, Muslim religion follower, and women with no-ANC visit were predictors of home delivery hotspot locations. The spatial regression revealed women from rural resident, women having no-education, women being in the household with a poorest wealth index, women with Muslim religion follower, and women having no-ANC visit were predictors of home delivery hotspot regions. Therefore, governmental and other stakeholders should remain the effort to decrease home childbirth through access to healthcare services especially for rural resident, strengthen the women for antenatal care visits.