PDF(Pubmed)
Abstract：
BACKGROUND: There has been extensive research conducted on open defecation in Ethiopia, but a notable gap persists in comprehensively understanding the spatial variation and predictors at the household level. This study utilizes data from the 2021 Performance Monitoring for Action Ethiopia (PMA-ET) to address this gap by identifying hotspots and predictors of open defecation. Employing geographically weighted regression analysis, it goes beyond traditional models to account for spatial heterogeneity, offering a nuanced understanding of geographical variations in open defecation prevalence and its determinants. This research pinpoints hotspot areas and significant predictors, aiding policymakers and practitioners in tailoring interventions effectively. It not only fills the knowledge gap in Ethiopia but also informs global sanitation initiatives.
METHODS: The study comprised a total weighted sample of 24,747 household participants. ArcGIS version 10.7 and SaT Scan version 9.6 were used to handle mapping, hotspots, ordinary least squares, Bernoulli model analysis, and Spatial regression. Bernoulli-based model was used to analyze the purely spatial cluster detection of open defecation at the household level in Ethiopia. Ordinary Least Square (OLS) analysis and geographically weighted regression analysis were employed to assess the association between an open defecation and explanatory variables.
RESULTS: The spatial distribution of open defecation at the household level exhibited clustering (global Moran\'s I index value of 4.540385, coupled with a p-value of less than 0.001), with significant hotspots identified in Amhara, Afar, Harari, and parts of Dire Dawa. Spatial analysis using Kuldorff\'s Scan identified six clusters, with four showing statistical significance (P-value < 0.05) in Amhara, Afar, Harari, Tigray, and southwest Ethiopia. In the geographically weighted regression model, being male [coefficient = 0.87, P-value < 0.05] and having no media exposure (not watching TV or listening to the radio) [coefficient = 0.47, P-value < 0.05] emerged as statistically significant predictors of household-level open defecation in Ethiopia.
CONCLUSIONS: The study revealed that open defecation at the household level in Ethiopia varies across the regions, with significant hotspots identified in Amhara, Afar, Harari, and parts of Dire Dawa. Geographically weighted regression analysis highlights male participants lacking media exposure as substantial predictors of open defecation. Targeted interventions in Ethiopia should improve media exposure among males in hotspot regions, tailored sanitation programs, and region-specific awareness campaigns. Collaboration with local communities is crucial.
METHODS: The study comprised a total weighted sample of 24,747 household participants. ArcGIS version 10.7 and SaT Scan version 9.6 were used to handle mapping, hotspots, ordinary least squares, Bernoulli model analysis, and Spatial regression. Bernoulli-based model was used to analyze the purely spatial cluster detection of open defecation at the household level in Ethiopia. Ordinary Least Square (OLS) analysis and geographically weighted regression analysis were employed to assess the association between an open defecation and explanatory variables.
RESULTS: The spatial distribution of open defecation at the household level exhibited clustering (global Moran\'s I index value of 4.540385, coupled with a p-value of less than 0.001), with significant hotspots identified in Amhara, Afar, Harari, and parts of Dire Dawa. Spatial analysis using Kuldorff\'s Scan identified six clusters, with four showing statistical significance (P-value < 0.05) in Amhara, Afar, Harari, Tigray, and southwest Ethiopia. In the geographically weighted regression model, being male [coefficient = 0.87, P-value < 0.05] and having no media exposure (not watching TV or listening to the radio) [coefficient = 0.47, P-value < 0.05] emerged as statistically significant predictors of household-level open defecation in Ethiopia.
CONCLUSIONS: The study revealed that open defecation at the household level in Ethiopia varies across the regions, with significant hotspots identified in Amhara, Afar, Harari, and parts of Dire Dawa. Geographically weighted regression analysis highlights male participants lacking media exposure as substantial predictors of open defecation. Targeted interventions in Ethiopia should improve media exposure among males in hotspot regions, tailored sanitation programs, and region-specific awareness campaigns. Collaboration with local communities is crucial.
摘要：
背景：对埃塞俄比亚的开放式排便进行了广泛的研究，但是在全面理解家庭水平的空间变化和预测因素方面仍然存在明显的差距。本研究利用2021年埃塞俄比亚行动绩效监测（PMA-ET）的数据，通过确定开放性排便的热点和预测因素来解决这一差距。采用地理加权回归分析，它超越了传统模型来解释空间异质性，对开放性排便患病率及其决定因素的地理差异提供了细致入微的理解。这项研究指出了热点领域和重要的预测因素，帮助政策制定者和从业者有效地调整干预措施。它不仅填补了埃塞俄比亚的知识空白，而且还为全球卫生倡议提供了信息。
方法：该研究包括24,747名家庭参与者的总加权样本。ArcGIS版本10.7和SaTScan版本9.6用于处理制图，热点,普通最小二乘，伯努利模型分析，和空间回归。基于伯努利的模型用于分析埃塞俄比亚家庭水平的露天排便的纯空间聚类检测。采用普通最小二乘（OLS）分析和地理加权回归分析来评估开放性排便与解释变量之间的关联。
结果：在家庭水平上，露天排便的空间分布表现出聚集性（全球MoranI指数值为4.540385，p值小于0.001），在阿姆哈拉发现了重要的热点，Afar,Harari,以及DireDawa的部分地区.使用Kuldorff扫描的空间分析确定了六个簇，在阿姆哈拉，有四个显示出统计学意义（P值<0.05），Afar,Harari,提格雷,埃塞俄比亚西南部。在地理加权回归模型中，作为男性[系数=0.87，P值<0.05]并且没有媒体接触（不看电视或听广播）[系数=0.47，P值<0.05]成为家庭水平的有统计学意义的预测因素在埃塞俄比亚排便。
结论：该研究表明，埃塞俄比亚家庭的露天排便情况因地区而异，在阿姆哈拉发现了重要的热点，Afar,Harari,以及DireDawa的部分地区.地理加权回归分析强调了缺乏媒体暴露的男性参与者是开放性排便的重要预测因素。埃塞俄比亚有针对性的干预措施应改善热点地区男性的媒体曝光率，量身定制的卫生计划，和针对特定地区的宣传运动。与当地社区的合作至关重要。
方法：该研究包括24,747名家庭参与者的总加权样本。ArcGIS版本10.7和SaTScan版本9.6用于处理制图，热点,普通最小二乘，伯努利模型分析，和空间回归。基于伯努利的模型用于分析埃塞俄比亚家庭水平的露天排便的纯空间聚类检测。采用普通最小二乘（OLS）分析和地理加权回归分析来评估开放性排便与解释变量之间的关联。
结果：在家庭水平上，露天排便的空间分布表现出聚集性（全球MoranI指数值为4.540385，p值小于0.001），在阿姆哈拉发现了重要的热点，Afar,Harari,以及DireDawa的部分地区.使用Kuldorff扫描的空间分析确定了六个簇，在阿姆哈拉，有四个显示出统计学意义（P值<0.05），Afar,Harari,提格雷,埃塞俄比亚西南部。在地理加权回归模型中，作为男性[系数=0.87，P值<0.05]并且没有媒体接触（不看电视或听广播）[系数=0.47，P值<0.05]成为家庭水平的有统计学意义的预测因素在埃塞俄比亚排便。
结论：该研究表明，埃塞俄比亚家庭的露天排便情况因地区而异，在阿姆哈拉发现了重要的热点，Afar,Harari,以及DireDawa的部分地区.地理加权回归分析强调了缺乏媒体暴露的男性参与者是开放性排便的重要预测因素。埃塞俄比亚有针对性的干预措施应改善热点地区男性的媒体曝光率，量身定制的卫生计划，和针对特定地区的宣传运动。与当地社区的合作至关重要。
