关键词: family planning machine learning reproductive health sexual health social determinants of health spatial analysis

来  源:   DOI:10.1093/haschl/qxae048   PDF(Pubmed)

Abstract:
Equitable access to sexual and reproductive health (SRH) care is key to reducing inequities in SRH outcomes. Publicly funded family-planning services are an important source of SRH care for people with social risk factors that impede their access. This study aimed to create a new index (Local Social Inequity in SRH [LSI-SRH]) to measure community-level risk of adverse SRH outcomes based on social determinants of health (SDoH). We evaluated the validity of the LSI-SRH scores in predicting adverse SRH outcomes and the need for publicly funded services. The data were drawn from more than 200 publicly available SDoH and SRH measures, including availability and potential need for publicly supported family planning from the Guttmacher Institute. The sample included 72 999 Census tracts (99.9%) in the 50 states and the District of Columbia. We used random forest regression to predict the LSI-SRH scores; 42 indicators were retained in the final model. The LSI-SRH model explained 81% of variance in the composite SRH outcome, outperforming 3 general SDoH indices. LSI-SRH scores could be a useful for measuring community-level SRH risk and guiding site placement and resource allocation.
摘要:
公平获得性健康和生殖健康(SRH)护理是减少SRH结果不平等的关键。公共资助的计划生育服务是对有社会风险因素阻碍其获得的人的生殖健康护理的重要来源。这项研究旨在创建一个新的指数(SRH[LSI-SRH]中的本地社会不平等),以基于健康的社会决定因素(SDoH)来衡量社区水平的不良SRH结果风险。我们评估了LSI-SRH评分在预测不良SRH结果和公共资助服务需求方面的有效性。数据来自200多个公开的SDoH和SRH措施,包括Guttmacher研究所对公共支持的计划生育的可用性和潜在需求。样本包括50个州和哥伦比亚特区的72999个人口普查区(99.9%)。我们使用随机森林回归来预测LSI-SRH得分;最终模型中保留了42个指标。LSI-SRH模型解释了复合SRH结果中81%的方差,表现优于3个一般的SDoH指数。LSI-SRH评分可用于衡量社区级SRH风险并指导站点放置和资源分配。
公众号