Johns Hopkins Prevention Epicenter: Transdisciplinary Research Approaches to Prevent Healthcare Associated Infections and Antibiotic Resistance (TRAP HAI & AR)

项目概要/摘要 该提案的目的是将基础、流行病学和技术发现转化为新的 预防医疗保健相关感染和抗生素耐药性以及改善抗生素使用的策略 所有医疗机构都使用诊断测试。这项工作包括六个核心项目和三个 可选的合作项目。核心项目包括：（1）利用人因工程方法 改善各种长期护理机构的环境清洁； (2) 开发和测试临床医生- 使用来自以下机构的数据为成人和儿童提供知情的电子抗生素相关不良事件伤害评分 电子健康记录（EHR），旨在改变医院的抗生素处方行为； (3) 制定一种措施来检测连续居住的患者的生物体传播事件 同一医院病房使用电子病历的数据，并试行将其用作改善房间清洁的指标； (4) 表征广谱产β-内酰胺酶肠杆菌的分子流行病学， 预测哪些患者被定植，并确定感染控制失误的作用和 传播过程中接触抗生素； (5) 描述新生儿鼻腔微生物群的特征 金黄色葡萄球菌菌血症，了解如何预防这种致命的感染； (6) 开发一个 使用 EHR 数据的算法快速识别接触过病毒的医护人员和患者 发现患有传染性疾病的患者，目的是快速降低继发性感染的风险 曝光。可选的合作项目包括：(1) 制定血培养基准和 血培养阳性率和血培养不适当的共识定义以及使用这些 结果来实施和评估基于证据的算法，以改善多国的血培养实践 成人医学患者机构队列； (2) 描述微生物学和地理的风险特征 在医院外（例如，在 家庭、长期护理机构、透析）并试点预防工具包，以降低感染风险 出院时需要中心导管的患者，以及 (3) 确定产后年龄的影响 新生儿重症监护病房 (NICU) 中出生体重与菌血症发生率之间的关联，以便 制定一种实用方法来报告 NICU 发病菌血症率，并根据出生体重和出生体重进行调整 产后年龄。该提案展示了整合医疗保健流行病学和 抗生素管理与其他学科——人因工程、数据科学、机器学习、 微生物学、数学建模、微生物组科学和实施科学——目标是 确定预防医疗保健相关感染和抗生素耐药性的新方法，这可以 告知 CDC 指导并应用于多种环境，以减少对患者和医疗保健的伤害 工人。

PROJECT SUMMARY/ABSTRACT The purpose of this proposal is to translate basic, epidemiologic, and technologic discoveries into new strategies to prevent healthcare-associated infections and antibiotic resistance and to improve how antibiotics and diagnostic tests are used across all healthcare settings. This work includes six Core Projects and three Optional Collaborative Projects. The Core Projects include: (1) using a human factors engineering approach to improve environmental cleaning in diverse long-term care settings; (2) developing and testing a clinician- informed, electronic antibiotic-associated adverse event harm score for adults and children using data from the electronic health record (EHR) with the goal of changing antibiotic prescribing behaviors in the hospital; (3) developing a measure to detect episodes of organisms transmission from patients sequentially occupying the same hospital room using data from the EHR and piloting its use as an indicator for improved room cleaning; (4) characterizing the molecular epidemiology of extended spectrum β-lactamase-producing Enterobacterales, predicting what patients are colonized with them, and determining the role of infection control lapses and antibiotic exposure in their transmission; (5) characterizing the nasal microbiota of neonates that develop Staphylococcus aureus bacteremia to understand how to prevent this deadly infection; and (6) developing an algorithm using EHR data to identify rapidly healthcare workers and patients who have been exposed to an index patient with a transmissible infectious disease with the goal of quickly mitigating the risk of secondary exposures. The Optional Collaborative Projects include: (1) developing a benchmark for blood culture and blood culture positivity rates and a consensus definition of blood culture inappropriateness and using these results to implement and evaluate an evidence-based algorithm to improve blood culture practices in a multi- institution cohort of adult medicine patients; (2) characterizing the risk for and microbiology and geographical patterns of central line-associated bloodstream infections that are acquired outside of the hospital (e.g., at home, in long-term care facilities, at dialysis) and pilot a prevention toolkit to reduce the risk of infection for patients requiring a central catheter upon hospital discharge, and (3) determining how postnatal age impacts the association between birthweight and bacteremia rates in the neonatal intensive care unit (NICU) in order to develop a practical approach for reporting NICU-onset bacteremia rates that adjusts for birthweight and postnatal age. This proposal demonstrates capacity to integrate expertise in healthcare epidemiology and antibiotic stewardship with other disciplines—human factors engineering, data science, machine learning, microbiology, mathematical modeling, microbiome science, and implementation science—with the goal of identifying novel approaches to prevent healthcare associated infections and antibiotic resistance and that can inform CDC guidance and be applied across multiple settings to reduce harm to patients and healthcare workers.