Two papers announced among 10 most influential in healthcare and infection control

The papers provide data-driven solutions to hospital infection and the use of machine learning in healthcare.

Prof. Jenna Wiens Enlarge
Prof. Jenna Wiens

Prof. Jenna Wiens’ group had two papers highlighted in a session on the top 10 most influential papers in healthcare epidemiology and infection control at Infectious Disease Week (IDWeek 2018). The papers were selected for their impact, the number of times they were cited in the preceding two years, and their potential effect on future research and technology.

The first paper, “A Generalizable, Data-Driven Approach to Predict Daily Risk of Clostridium difficile Infection at Two Large Academic Health Centers,” produced a data-driven approach to building models for estimating daily patient risk for contracting C. difficile in a particular hospital. An estimated 293,300 healthcare-associated cases of C. difficile occur annually in the US. Existing research on the spread of this infection has focused on developing risk prediction models that work well across institutions. The research team found that this one-size-fits-all approach ignores important hospital-specific factors.

Instead, they developed a generalizable method for building facility-specific models. This approach was used to build institution-specific models at two large hospitals with different patient populations and produced models that achieved 95% confidence in both cases. These hospital-specific models allow for earlier and more accurate identification of high-risk patients and better targeting of infection prevention strategies.

The second paper, “Machine Learning for Healthcare: On the Verge of a Major Shift in Healthcare Epidemiology,” is a user’s guide on how Machine Learning (ML) can transform healthcare epidemiology, providing examples of successful applications, challenges, and pitfalls.

The increasing availability of electronic health data presents a major opportunity in healthcare for developing improved healthcare techniques. However, for healthcare epidemiologists to best use the data, computational techniques that can handle large, complex datasets are required. The group’s paper presents special considerations for healthcare epidemiologists who want to use and apply ML to these challenges, and provides a basic overview of the field and its accomplishments.

The presentation of these papers was given by Grace M. Lee, MD MPH from Stanford Medicine’s Lucile Packard Children’s Hospital.