Stitching the fabric back together: child fostering attitudes in the transition to a chronic HIV epidemic

There are an estimated 56 million orphans and vulnerable children across sub-Saharan Africa. Communities typically care for orphan children through informal caring arrangements – either within or outside of kinship networks. Within Kenya, an estimated 250,000 children live on the streets. There is less research related to fostering attitudes of this special population than orphans and vulnerable children generally. Important research over the past decade has illuminated multiple ways in which children are made more vulnerable because of HIV, including parental death and street-migration from HIV-affected households. As HIV transitions from a terminal illness to a chronic, manageable one, research is also required to establish how parents living with HIV can be an asset to children. In this study, we assess whether mothers living with HIV were very willing to foster biologically-related children, and street-involved children, how these fostering attitudes differed from mothers not living with HIV, and whether differences in fostering attitudes by reported HIV status were mediated by social support, family functioning and general self-rated health. Approximately 40% of mothers living with HIV were very willing to provide long-term foster care to biologically-related or street-involved children. This was less than the percentage of mothers not living with HIV, who were very willing to foster biologically-related children (61%) or street-involved children (58%). Significant portions of these differences were explained by social support, family functioning and general self-rated health. Multi-sectoral approaches are suggested by these findings in order to improve the child-fostering capacity of mothers living with HIV. Improving social support, family functioning and general self-rated health among HIV-infected mothers may not only provide protective benefits for the mothers and their children, but also expand the community’s capacity to care for orphan and vulnerable children.     

基于剂量组学预测肺癌患者放射性肺炎发生的研究

目的 探讨剂量组学在预测肺癌根治性放疗患者放射性肺炎发生中的应用潜能。方法 回顾性收集行根治性放疗的314例肺癌患者的临床资料、放疗剂量文件、定位及随访CT图像,根据临床资料及影像学随访资料对放射性肺炎进行分级,提取全肺的剂量组学特征,构建机器学习模型。应用1000次自助抽样法（bootstrap）的最小绝对值收敛和选择算子嵌套逻辑回归（LASSO‐LR）及1000次bootstrap的赤池信息量准则（AIC）向后法筛选与放射性肺炎相关的剂量组学特征,随机按照7∶3划分为训练集及验证集,应用逻辑回归建立预测模型,并应用ROC曲线及校正曲线评价模型的性能。结果 共提取120个剂量组学特征,经LASSO‐LR降维筛选得到12个特征进入“特征池”,再经过AIC向后法筛选,最终筛选出6个剂量组学特征进行模型构建,训练集AUC为0.77（95%CI为0.65~0.87）,独立验证集AUC为0.72（95%CI为0.64~0.81）。结论 利用剂量组学建立的预测模型具有预测放射性肺炎发生的潜力,但仍需继续纳入多中心数据及前瞻性数据进一步挖掘剂量组学的应用潜能。