As federal agencies advance their IT modernization goals, many are exploring the potential of using artificial intelligence tools that can supplement human employees. Federal agencies are currently applying AI to a variety of missions, and public health is no exception. The U.S. Centers for Disease Control and Prevention’s new Office of Public Health Data Surveillance Technology (DST) is exploring how to apply AI to public health data and leverage generated AI to enhance efforts.
“There was actually a series of 15 pilots conducted at various centers and offices within the agency,” said DST Director Jennifer Layden. Federal Monthly Insights – Operationalizing AI. “These were used to evaluate what kind of infrastructure we needed, what kind of features we wanted to use, what security factors we needed to consider. This ranged from basic work to more operational work such as website redesigns, evaluations, and commenting back on protocols.”
CDC launched DST last year to align its data strategy. This includes improving data exchange with other federal, state, and local agencies and nongovernmental partners. Improving the way data informs public health efforts. How to better visualize and disseminate data to the public. AI is rapidly becoming part of these efforts.
AI use cases
For example, automated processes can quickly flag potential health threats, facilitating faster notification and communication. But it could also improve internal workflows and make CDC staff more efficient at their jobs. Generative AI can also quickly create fact sheets about emerging public health threats to educate both those at risk and the medical professionals who need to treat them.
For example, Leyden described one test use case in which AI inspected public cooling facilities to identify areas at risk for the spread of Legionella bacteria spread through contaminated water.
Ensuring data privacy and reducing bias
While there are a variety of potential use cases, DST is focused on guardrailing the use of these tools, Layden said.
“What we’re trying to do in the process is provide guidance for programs and scientists to develop basic strategies for using such tools to help people do it safely and reliably. ,” she said. Federal Drive with Tom Temin. “[Two:] We recognize that we don’t want to create risks for anonymization or sharing information that shouldn’t be shared. And third, how to consider ethics and bias. ”
Data privacy and ethics and bias considerations are especially important when working with public health data. One of the major concerns about AI tools is that they could be used by malicious parties to compromise the data privacy of patients and the public by manipulating them to reveal personally identifiable information. That’s true. So you need to consider anonymization and deciding what data is appropriate to share. However, that data should be as fair and diverse as possible to avoid introducing bias and potentially creating new underserved populations or worsening the situation of existing populations. Must be something.
Choose the right team and AI tools
That’s why Leyden said DST encourages the use of multidisciplinary teams when working with public health data. She is an expert on this disease and other public health threats, those who understand the affected or at-risk populations, and those who understand the data tools and methodologies for performing advanced analytics. We proposed a team that included.
“It really takes a multidisciplinary team coming together to understand what the question is that we’re trying to answer,” Leyden said. “What should we consider when understanding the data we are using? So what is the best tool to answer that question? So instead of just using a tool because it’s new; Is it the best tool to answer the question at hand?”
Another consideration around these tools is the fact that they evolve. After all, AI tools have been around for a while, but generative AI only gained traction in the last year or so. As that evolution occurs, experts must continually reevaluate them for a variety of reasons. Are they still the best tools for the job? Has the nature of identifiable information changed in any way?
Sharing information and tools
The appropriate community also needs access to that information. Best practices and lessons learned can prevent other teams from making similar mistakes and save you time evaluating your own tools. Leyden said stakeholders will need to continually build, test and validate the framework to continue their work.
Because threats evolve, capabilities must continue to evolve. Therefore, public health experts need to be aligned.
“One of the challenges in public health in general, which is not unique or new, is to bring in more advanced analytical capabilities, the expertise of the workforce,” Leyden said. “We also have the bandwidth, the ability to understand the full range of tools and how they can be used… recognizing that building these capabilities in-house will take time, and partnering with academic and private partners. So, how we can partner with experts in academia or in the private sector is another way for us to build capacity, understanding and expertise.”
One way CDC accomplishes this is through the use of shared tools. For example, Layden said more than half of state jurisdictions use incident investigation tools operated and maintained by the CDC. Similarly, there are shared monitoring systems to track emergency room data. Additionally, we have a shared governance model to support the development and sharing of even more tools.
One benefit of such sharing tools is that they facilitate the sharing of data more broadly and in the same format, reducing the amount of work required by data scientists to reconcile data before starting analysis. That’s it.
“In my opinion, the more we share public health, the more we need to continue to grow by building enterprise-wide tools that can be used and leveraged appropriately. At the same time, we can share best practices.”
Copyright © 2024 Federal News Network. All rights reserved. This website is not directed to users within the European Economic Area.
