A UC Berkeley School of Public Health research team has been awarded a $3.6 million, five-year grant by the National Institutes of Health to develop new approaches for simulating and optimizing surveillance networks that detect infectious diseases.
The project will enlist big data to tackle major challenges facing the monitoring of global infectious diseases, such as tracking the progress of disease elimination campaigns, detecting co-infections and maximizing detection of rare diseases in high-risk populations. The researchers will focus on high-priority global infectious diseases, including tuberculosis, malaria and schistosomiasis, and the team will work in partnership with practitioners at the U.S. and Chinese Centers for Disease Control and Prevention.
“Targeted and efficient surveillance systems are critical to detecting outbreaks, tracking emerging infections and supporting infectious disease control efforts, particularly in low- and middle-income countries where estimating the distribution of disease is a major challenge,” said project leader Justin Remais, an associate professor of environmental health sciences at the School of Public Health. “We need to take advantage of new, vast health datasets to identify surveillance strategies that are effective under changing epidemiological and environmental conditions.”
The research is funded by the National Institute of Allergy and Infectious Diseases, under the NIH’s Spatial Uncertainty funding opportunity. Collaborators on the project include statisticians and epidemiologists at the Beijing Institute for Microbiology and Epidemiology, Emory University and the University of Florida.
Infectious disease surveillance systems provide vital data that serve as the foundation of evidence-based programs to improve public health. Globally, surveillance systems vary widely in their design, including how, where and how frequently they survey populations for infections, and the specific diagnostic approaches used. Designing modern systems that provide reliable and timely estimates of disease occurrence, particularly among high-risk groups, is crucial to reducing the burden of global infectious diseases.
The new NIH-funded project, titled spatio-temporal data integration methods for infectious disease surveillance, aims to develop statistical techniques for integrating complex data from multiple surveillance systems, provide critical insights into how surveillance systems function, and lead to key advances in surveillance informatics. The research team will develop algorithms that predict how surveillance systems perform under different configurations, and can estimate the optimal allocation of surveillance resources under various constraints.
“We will feed the insights gained from simulation studies back into the redesign of real-world surveillance systems, helping our partners design systems that are more effective at detecting infections at the outset of an epidemic, for instance, or as disease elimination is approached,” Remais said.
Collaborating on the project is Alan Hubbard, a professor of biostatistics and epidemiology at the School of Public Health. Hubbard is one of the principal faculty in Berkeley’s Center for Targeted Machine Learning and Causal Inference, which develops, implements and disseminates signature methods for exploiting vast, new health datasets.
“Our center provides the perfect environment to develop new platforms for analyzing the project’s big surveillance data, allowing us to assess the performance of specific surveillance architectures in real time,” Hubbard said. “Berkeley has just the right expertise for this research, including our tremendous strength in epidemiological methods, data science, infectious disease and global health.”
By Brett Israel