Enhanced Detection System for Healthcare-Associated Transmission of Infection

摘要 尽管最近在减少医疗保健相关感染(HAI)方面取得了进展，但美国疾病控制和预防中心 美国疾病控制和预防中心估计，2011年美国急性护理医院发生了72.2万例甲型HAI， 导致75,000人死亡。目前在医院检测疫情的方法是初级的，很可能是 完全遗漏了一些疫情，并导致对其他疫情的检测出现很大延误。 有两个主要的 医疗保健方面的发展有可能彻底改变医疗保健如何与疾病爆发相关 细菌病原体在医院中被识别和控制。首先， 《负担得起的医疗法案》要求使用 电子病历(EMR)导致了它的 在医疗保健中广泛使用。 第二，成本 细菌全基因组测序(WGS)已大幅下降，这使得它被感染所利用 这些计划越来越可行。在此应用程序中，我们建议建立和评估 加州大学医疗协会传播增强检测系统(EDS-HAT) 匹兹堡医疗中心(UPMC)。EDS-HAT使用WGS和EMR分析相结合的方式 增强的疫情检测功能。我们的具体目标是1a)：确定EDS-HAT识别HAT的效用 这不是通过常规感染预防实践来确定的，1b)：提高效率并减少 通过使用电子病历限制使用WGS的EDS-HAT的费用，2a)：计量减少的HAIs如下 实施EDS-HAT，和2b)：估计预防和保健的感染和死亡人数 通过EDS-HAT避免的成本。对于AIM 1a，在常规感染的同时，将回顾地进行EDS-HAT 预防做法(怀疑爆发时要求进行分子分型)仍在继续，因此允许 两种方法的直接比较。对于目标1b，我们将提高效率和降低成本 EDS-HAT通过使用机器学习和EMR的数据挖掘来选择WGS的分离物。对于Aim 2a，我们 将实时监测实施EDS-HAT前后HAI比率的变化，这将 最后，对于目标2b，我们将进行临床和预算影响分析，以 确定EDS-HAT的总体影响。为了实现这些目标，我们组建了一个团队， 传染病、暴发调查、感染预防、微生物基因组学和基因组学方面的专门知识 流行病学、机器学习和数据挖掘、经济分析和建模、流行病学和 生物统计学。EDS-这可能会大幅降低感染、死亡和医疗成本，并 可作为医院如何检测HAT的模型。

ABSTRACT Despite recent progress in reducing the incidence of healthcare-associated infections (HAIs), the Centers for Disease Control and Prevention estimated that 722,000 HAIs occurred in U.S. acute care hospitals in 2011, resulting in 75,000 deaths. Current methods for detecting outbreaks in hospitals are rudimentary and likely to miss some outbreaks altogether and result in substantial delays in detection of others. There are two major developments in healthcare that have the potential to revolutionize how healthcare associated outbreaks of bacterial pathogens are identified and controlled in hospitals. First, the Affordable Care Act mandates use of the electronic medical record (EMR), which has led to its widespread use in healthcare. Second, the costs of bacterial whole genome sequencing (WGS) have declined substantially, which is making its use by infection programs increasingly feasible. In this application, we propose to establish and evaluate the impact of the Enhanced Detection System for Healthcare Association Transmission (EDS-HAT) at the University of Pittsburgh Medical Center (UPMC). EDS-HAT uses a combination of WGS and analysis of the EMR for enhanced outbreak detection. Our specific aims are to 1a): Determine the utility of EDS-HAT to identify HAT that is not identified through routine infection prevention practice, 1b): Improve the efficiency and reduce the cost of EDS-HAT by using the EMR to restrict the use of WGS, 2a): Measure reductions in HAIs following implementation of EDS-HAT, and 2b): Estimate the number of infections and deaths prevented and healthcare costs averted by EDS-HAT. For Aim 1a, EDS-HAT will be performed retrospectively while routine infection prevention practice (requests for molecular typing when an outbreak is suspected) continues, thus allowing a direct comparison of the two approaches. For Aim 1b, we will improve the efficiency and reduce the cost of EDS-HAT by using machine learning and data mining of the EMR to select isolates for WGS. For Aim 2a, we will monitor changes in HAI rates both before and after implementation of EDS-HAT in real time, which will occur at the beginning of year 3. Finally, for Aim 2b, we will perform clinical and budget impact analyses to determine the overall impact of EDS-HAT. To accomplish these aims, we have assembled a team with expertise in infectious diseases, outbreak investigation, infection prevention, microbial genomics and genomic epidemiology, machine learning and data mining, economic analysis and modeling, epidemiology, and biostatistics. EDS-HAT will likely lead to substantial reductions in infections, deaths, and healthcare costs and can serve as a model for how HAT is detected in hospitals.