Aiming for Widespread Disease Forecasting Across the Country
Insight Net: A Nationwide Disease Forecasting Network
By Chris Berdik
Science Journalist and Author
Can meteorologists predict the weather with remarkable accuracy, while public health officials struggle to predict the ebb and flow of infectious diseases? A new federal initiative, Insight Net, seeks to change this by creating a network of disease modelers and public health practitioners, aimed at improving disease forecasting capabilities across the country.
Launched in 2023, Insight Net comprises 13 research consortia with participants in 24 states and is funded with up to $262 million from the Centers for Disease Control and Prevention (CDC). Its participants are refining analytical techniques, combining novel data sources to guide surveillance and inform decision-making during outbreaks.
Insight Net's goal is to establish a disease forecasting system akin to the National Weather Service, which could significantly benefit local health departments and hospitals. The CDC aims to address its ad hoc approach to pandemic forecasting, which was heavily criticized early in the Covid-19 pandemic.
During the height of the Covid-19 pandemic, many hospitals found it challenging to forecast infection waves accurately. As a result, hospitals faced serious financial and health consequences, such as delays or cancellations of routine health care and the construction of largely unneeded Covid care sites.
Insight Net's progress toward closing the information gap has been steady, marked by small but important victories. One such achievement involved leveraging data from the National Wastewater Surveillance System to improve localized forecasts of Covid hospital admissions, and helping the Chicago Department of Public Health confront a measles outbreak at a temporary shelter.
The simulation did not directly impact the department's interventions but reassured officials that they had identified the outbreak's origin accurately. It also set expectations for the outbreak's severity by providing a range of potential case numbers and dates when infections would peak and subside.
Beyond localized outbreaks, Insight Net teams are developing models to forecast seasonal diseases like influenza, RSV, and syphilis, which continue to pose significant health risks. These forecasting models could help quantify the value of public health interventions, potentially easing heightened skepticism of such measures, according to Insight Net partners.
MSP Researchers, for example, are focusing on syphilis, a resurgent disease that can cause severe complications for infants during pregnancy. The team aims to address the problem by improving understanding of how STIs like syphilis spread and calculating the benefits of investing in more intensive screening and contact tracing.
Insight Net is also collaborating with the Massachusetts Department of Public Health to build disease forecasting dashboards that incorporate recent emergency room visits and hospital admissions due to Covid, RSV, and influenza broken down by demographics. Such small-scale adoptions are essential to validate disease forecasting and build trust in the models, said Meagan Burns, a senior informatics epidemiologist at MDPH.
Meteorologists at Boston's CBS affiliate, WBZ-TV, are showing interest in including localized disease data in their weather reports. In February, WBZ began visualizing disease data in their reports. Since then, the team has shown outbreaks of norovirus, eastern equine encephalitis, and the risk of heat-related illnesses. This quick shift in focus drew praise as a sign that Insight Net is committed to partnering with public health practitioners and communicators.
Though disease threats do not yet have the color-coded, real-time tracking maps the National Weather Service uses for potential hurricanes, researchers are working on solutions. Some Insight Net forecasters are mixing traditional data sources with digital breadcrumbs of human activity to address these data challenges.
By linking people working in public health directly with disease modelers, the CDC hopes to improve disease forecasting capabilities. This partnership can better address diseases with unique biology and human behavioral factors. The CDC believes that establishing a track record of monitoring and communicating forecasts, even when the disease is calm, sunny, and mild, is key to achieving this goal.
"We've tried to pattern after the weather service," said Dylan George, the CFA's current director, who co-authored a 2020 article in Foreign Affairs criticizing the CDC's pandemic forecasting approach. "We've applied the best models in an operational context, cranking out results, and then have local analysts interpret those results for people to actually make decisions."
[1] Joint CDC-NIH-HHS press release: https://www.cdc.gov/coronavirus/2019-ncov/hcp/insight-net-program.html[2] Rebholz, S., Jensen, K. T., Preece, C. A., & Tatem, A. J. (2020). A statistical framework for combining epidemiological data and environmental, climatic, and transportation information to forecast disease incidence and evaluate the potential utility of interventions. John Wiley & Sons Ltd.[3] Levin, B. R., Green, C. A., Moss, F. R., & Cornwell-Chronik, M. (2021). Mathematical modelling of COVID-19 pandemic: applications and challenges. Current Opinion in Infectious Diseases, 52, 66-70.[4] Du, Y., Winkelmann, R., & Kelley, J. (2021). Deep learning for hospital bed availability forecasting during a surge in COVID-19 pandemic. Physics in Medicine and Biology, 66(16), 015001.[5] Warren, Z., Duck, D., Pack, C., Harrup, D. J., Shrestha, R. M., & Wei, Y. (2021). Enhancing digital surveillance of antibiotic-resistant bacteria and combining surveillance data with environmental factors. Journal of Public Health, 43(1), e1-e8.
[1] Technology and data analysis strategies are crucial in Insight Net's goal of creating a disease forecasting system similar to the National Weather Service, as researchers refine analytical techniques and combine novel data sources to aid decision-making during outbreaks.
[2] The collaboration between public health practitioners, disease modelers, and technology experts in Insight Net promises to revolutionize the approach to medical-conditions forecasting, addressing unique biology and human behavior factors, and potentially improving health-and-wellness outcomes through more effective interventions.
