III: Medium: Collaborative Research: Detecting and Controlling Network-based Spread of Hospital Acquired Infections

University Of Virginia Main Campus, Charlottesville VA

Investigators

Abstract

Hospital Acquired Infections (HAIs) are becoming a major challenge in health systems worldwide. Detection and control of HAIs are challenging and resource intensive, because of the high costs of patient treatment and disinfection of hospital facilities, making them fundamental public health problems. Despite its huge importance for hospitals, and the interest from both clinical and epidemiological researchers, these problems remain poorly understood. This project seeks to develop a novel network-based approach to improve hospital infection control using models and data science. This proposal brings together a highly multi-disciplinary team of researchers, and will lead to fundamental contributions in different areas of computer science (data mining, machine learning, graph mining, social networks, and optimization), network science (mathematical models and dynamical systems) and computational epidemiology (infectious diseases, and hospital epidemiology). The planned work has immediate implications for public health e.g. it can lead to new design policies and guidance for hospital infection control. Research findings will be incorporated into graduate level classes, tutorials, contests and workshops to bring computational biologists and data miners together. There are several challenges in studying HAI outbreaks primarily because the dynamics of HAI spread are much more complex than other diseases, such as influenza, due to many more factors and pathways involved. To overcome these issues, the project team will use a new class of two-mode cascade models, which have very different dynamics than the standard models, and have not been studied in data mining. The will investigate the following topics: (1) Surveillance, early detection of HAI outbreaks, (2) Designing interventions to control the spread of HAIs, and (3) Modeling and estimating exposure risk for HAIs. A unified set of problems will be considered, including modeling, detection, control and inference of missing infections. These are challenging stochastic optimization problems on networks, and the project team will invent rigorous and scalable methods using tools from data mining, machine learning and combinatorial optimization. Their research will use a unique fine-grained, large-scale data set of operations from a public hospital, supplemented with data from other hospitals. The results will be validated with the help of domain experts including epidemiologists and clinicians involved in hospital infection control. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.