GeoHAI: A novel geographic tool for Hospital Acquired Infection visualization and assessment

项目总结 医院获得性感染很常见，影响了3.2%的急性护理住院患者。最近的报道说 总体HAI率有所改善，主要是由于手术部位(SSI)和导管的改善 尿路感染(CAUTI)。传播感染，如艰难梭菌(CDI)，有 没有显示出随着时间的推移同样的下降。这可能是因为预防CDI需要全面的 解决环境和患者层面风险因素的医院范围内的方法。地理信息 系统(GIS)和空间分析技术已成为公共卫生信息学的重要工具 因为它们可以集成大量的数据源，并探索数据中的关联和模式 使用传统的生物统计学方法可见。地理信息系统和空间分析的应用范围很广，但有 在医院里很大程度上被忽视了。本研究的目的是开发一种HAI评估工具， 它结合了医院的地理数据和来自电子健康记录的患者级别数据 系统，这对于医院感染预防人员更好地识别HAI集群和评估 潜在的风险。我们汇集了一支由临床、操作和学术研究人员组成的多学科团队， 在地理信息系统和空间分析、患者安全、公共卫生信息学、可用性评估和混合- 方法评价。我们的目标是：1)创建一个动态的、空间参考的、可供地理信息系统使用的数据库，以及一组 HAI爆发和风险检测的统计算法，旨在便于在其他地方实施 医院；2)创建既可用又有用的地理HAI可视化和评估工具(GeoHAI 支持感染预防人员发现新出现的HAI聚集区和医院的高危地区； 3)进行混合方法分析，以确定GeoHAI的实施如何影响行为， 过程和HAI结果。该工具将使用时空贝叶斯模型来识别国家 医疗安全网络(NHSN)定义的医院发病CDI和多药耐药生物(MDRO)和 根据医院和患者的风险因素预测潜在的高风险区域。对该工具的综合研究 可用性和有用性将与工具开发相结合。工具开发将侧重于可重复性 在没有当地地理方法专业知识的情况下，使其他医院系统能够开展类似的工作。是我们独一无二的 方法是一种评估战略，重点是减少医院获得性感染，但也寻求 了解该工具及其衍生的信息如何影响医院中的患者安全实践。 我们希望这一工具的实施将从根本上改变感染的工作流程和反应速度 预防人员，大大提高了他们预防HAI的能力，而不是在HAI发生后做出反应。

PROJECT SUMMARY Hospital Acquired Infections are common, affecting 3.2% of acute care hospital admissions. Recent reports have shown an improvement in overall HAI rates, primarily driven by improvements in surgical site (SSI) and catheter associated urinary tract infections (CAUTI). Transmissible infections, such as Clostridium difficile (CDI), have not shown the same decrease over time. This may be because prevention of CDI requires a comprehensive hospital-wide approach addressing environmental and patient-level risk factors. Geographic Information Systems (GIS) and spatial analysis techniques have become an important tool in public health informatics because they can integrate a vast number of data sources and explore associations and patterns in the data not visible using traditional biostatistical methods. Applications of GIS and spatial analysis are wide ranging but have largely been ignored in the hospital setting. The objective of this research is to develop a HAI assessment tool, which incorporates geographic data on the hospital and patient-level data from the electronic health record system, that is useful for hospital infection preventionists in better identifying clusters of HAI and assessing potential risk. We bring together a multidisciplinary team of clinical, operational, and academic investigators with expertise in GIS and spatial analysis, patient safety, public health informatics, usability assessment, and mixed- methods evaluation. Our aims are to: 1) create a dynamic, spatially referenced, GIS-ready database, and a set of statistical algorithms for HAI outbreak and risk detection, which are designed for easy implementation at other hospitals; 2) create a Geographic HAI visualization and assessment tool (GeoHAI) that is both usable and useful in supporting infection preventionists in detecting emerging clusters of HAI and high risk areas of the hospital; and 3) Conduct a mixed methods analysis to determine how implementation of GeoHAI influences behaviors, processes, and HAI outcomes. The tool will use spatio-temporal Bayesian models to identify clusters of National Healthcare Safety Network (NHSN)-defined hospital onset CDI and multidrug resistant organisms (MDRO) and predict potential high risk areas given hospital and patient risk factors. Comprehensive studies of the tool's usability and usefulness will be integrated with tool development. Tool development will focus on reproducibility to enable similar work at other hospital systems without local expertise in geographic methods. Unique to our approach is an evaluation strategy that focuses on the reduction of hospital acquired infection, but also seeks to understand how the tool and the information derived from the tool impacts patient safety practices in the hospital. We expect the implementation of this tool to radically change the workflow and speed of response of infection preventionists, greatly improving their ability to prevent HAI instead of reacting after they have occurred.