A groundbreaking analytical tool developed by researchers at Columbia University Mailman School of Public Health promises to enhance hospitals’ ability to curb the spread of antibiotic-resistant infections. This innovative method, detailed in the journal Nature Communications, surpasses traditional techniques such as contact tracing by inferring the presence of asymptomatic carriers of drug-resistant pathogens within hospital settings.
The urgency of addressing antimicrobial resistance (AMR) cannot be overstated. In 2019, AMR infections were associated with a staggering 5 million deaths worldwide, highlighting the critical need for improved detection and prevention strategies.
Revolutionizing Infection Control
The inference framework crafted by Columbia researchers marks a pioneering step in infection control. By integrating multiple data sources—including patient mobility data, clinical culture tests, electronic health records, and whole-genome sequence data—the model predicts the spread of AMR infections with unprecedented accuracy. The study focused on carbapenem-resistant Klebsiella pneumoniae (CRKP), a bacterium notorious for its high mortality rate, utilizing five years of data from a New York City hospital.
CRKP colonization levels in healthcare facilities can vary significantly, reaching up to 22 percent in some locations. However, routine screening for CRKP remains uncommon, often relying on symptomatic patients or those suspected of contact with symptomatic individuals, thereby missing asymptomatic carriers.
“Many antimicrobial-resistant organisms colonize people without causing disease for long periods, during which these agents can spread unnoticed to other patients, healthcare workers, and even the general community,” stated Sen Pei, PhD, assistant professor of environmental health sciences at Columbia Mailman School. “Our inference framework better accounts for these hidden carriers.”
Enhanced Detection and Prevention
The researchers demonstrated that their inference framework could estimate CRKP infection probabilities more accurately than traditional methods, even with limited data. By combining the four data sources, the model significantly improved carrier identification. Data simulations revealed that isolating carriers based on the model’s predictions was more effective in preventing infection spread than conventional approaches.
Using the inference model, isolating 1 percent of patients weekly reduced positive cases by 16 percent and colonization by 15 percent. In contrast, traditional contact tracing methods achieved only a 10 percent reduction in positive cases and an 8 percent reduction in colonization.
Building on Previous Research
This study builds on earlier research published in the Proceedings of the National Academy of Sciences (PNAS), which introduced a method for predicting methicillin-resistant Staphylococcus aureus (MRSA) colonization. The current study advances this work by incorporating patient-level electronic health records and whole-genome sequence data, allowing for more precise identification of silent spreaders.
Despite these advancements, eradicating AMR pathogens in hospitals remains challenging due to their widespread community presence, limited surveillance, and high false-negative rates in clinical culture tests. Future research aims to explore AMR spread using ultra-dense sequencing, offering hope for further improvements.
Collaborative Efforts and Future Directions
This study represents a collaborative effort among computational researchers and physician scientists from various institutions. Contributors include Dwayne Seeram and Anne-Catrin Uhlemann from Columbia University Irving Medical Center, Seth Blumberg from the University of California, San Francisco, Bo Shopsin from New York University Grossman School of Medicine, and Jeffrey Shaman from Columbia Mailman School and Columbia Climate School.
The research was supported by funding from the U.S. Centers for Disease Control and Prevention (CDC), the National Institutes of Health (NIH), and the NYU Langone Health Antimicrobial-Resistant Pathogens Program. These cooperative agreements underscore the critical importance of addressing AMR as a public health priority.
As the healthcare community continues to grapple with the challenges posed by AMR, the development of this inference framework offers a promising tool in the fight against drug-resistant infections. By enhancing detection and prevention strategies, hospitals can better protect patients and curb the spread of these dangerous pathogens.
