Latest / Public Health Review Morning Edition / 1023: AWS Examines AI In Public Health
Transcript
- 0:03This is the award-winning Public Health Review Morning Edition
- 0:06for Friday, October 31st, 2025. I'm Jansen Silvers.
- 0:10Now today's news from the Association of State and
- 0:12Territorial Health Officials. This morning we have a special
- 0:16episode that examines AI in public health.
- 0:19We are joined today by Jim Daniel, the public health leader
- 0:21for state and local government at Amazon Web Services.
- 0:25AWS is a founding member of Asto's Innovation Advisory
- 0:28Council, a new multi sector collaborative of private sector
- 0:31partners helping keep Asto's membership and leadership
- 0:34informed and nimble on emerging topics affecting public health.
- 0:38Jim, I'd love to jump right in. Tell me about how generative AI
- 0:42has assisted with electronic case reports.
- 0:45A couple of years ago, we were approached by many public health
- 0:48departments to help them process the large volume of electronic
- 0:52case reports that were coming in.
- 0:54And, you know, originally our thoughts were, you know, these
- 0:57are standard data feeds with X standard XML tags that we could
- 1:04parse the way we normally parse public health data.
- 1:07We could look for the tags, parse them into a data lake, and
- 1:10look at the data that is in those various fields.
- 1:15What we quickly found out, though, that there was not a lot
- 1:18of standardization from provider organization to provider
- 1:22organization. Basic demographic data might
- 1:25have different fields names with different XML tags.
- 1:28They might be using different vocabulary standards within
- 1:31those fields. So instead of, you know,
- 1:34traditional methods of parsing the data using the XML tags as a
- 1:38way to identify the information critical for public health, we
- 1:42instead turn to Gen. AI and just started using
- 1:45natural language to ask those questions.
- 1:48Is this person pregnant? Is this person a food handler?
- 1:51So that we can take that data from the electronic case
- 1:55reports, transform it with Gen. AI into a format that is
- 1:59actually then used directly in the electronic disease
- 2:03surveillance systems that public health uses.
- 2:05And then tell me how you see AI improving our ability to handle
- 2:08forms. Yeah, I don't think that public
- 2:11health will ever get away from paper forms where, you know,
- 2:14even today we see some electronic lab reports coming
- 2:18in, not electronically, but on paper.
- 2:20And one of the great ways that generative AI can help there is
- 2:25actually looking at those paper based forms and automatically
- 2:29figuring out what are the fields that you're looking at and what
- 2:33are the various options for those fields on.
- 2:37With Jenny I, it's actually able to look at the very complicated
- 2:40form, whether it's a fill in the blank circle, the right answer
- 2:44XS, it can look at all of that and automatically figure out
- 2:49exactly what you're trying to do with that form.
- 2:51You can also have a human in the loop with these where you can
- 2:55get confidence intervals for whether or not the Gen.
- 2:58AI is confident about the answer that it's chosen.
- 3:02And then you can choose, you know, the scores where a human
- 3:06might want to review those that are less than 70% perhaps.
- 3:08And it can bring up the form and show how Gen.
- 3:11AI has proposed that the answer is.
- 3:14And then you can edit that directly the end the user
- 3:16interface before it gets extracted and put over into the
- 3:19system of record. I also know that AI chat bots
- 3:22have been successful. Can you expand?
- 3:25You know, many times public health is overwhelmed with
- 3:29different stakeholders calling in or emailing with questions,
- 3:32whether it's a provider who might have a question about an
- 3:35immunization information system, a person who might have
- 3:39questions about whether or not they're eligible for WIC.
- 3:43You know, these are the kinds of questions that public health
- 3:46departments spent many, many hours on, you know, answering
- 3:50via phone or e-mail. And when it comes to those kind
- 3:53of simple types of questions, we can actually help with an
- 3:58integrated virtual agent or chat bot.
- 4:01They can be placed directly on the website where we can use our
- 4:05tools to consume the information that has all of these answers to
- 4:09the questions and then have a chat bot that can actually help
- 4:13answer those questions. That chat bot can be available
- 4:16on a website. It can be available via phone
- 4:19number where people can call and ask their questions and get an
- 4:23answer verbally. You can ask questions verbally
- 4:26in multiple languages, and even though your source documents are
- 4:30only in English, you can still interface with the people who
- 4:33are asking questions in multiple languages.
- 4:36You can also be very specific with the chat bots about only
- 4:40looking at your documents and not trying to look at the rest
- 4:43of the world's information to find an answer.
- 4:46And then it's just the harder, more complicated questions that
- 4:49end up coming to a live agent, saving public health time and
- 4:53money. And when it comes to telling a
- 4:55story, that's also where AI can be instrumental, right?
- 4:59Yeah. So one of the, I think key pain
- 5:02points for public health departments is the lack of
- 5:05highly skilled data analyst. And public health is always
- 5:10getting questions coming in from leadership from their governor's
- 5:15office where they have to immediately come up with visuals
- 5:19and graphs. And then we can use quick sight
- 5:23for a business user to come in and start asking questions with
- 5:28natural language. So you can start asking
- 5:30questions like can you describe the relationship between obesity
- 5:36and cardiovascular disease? And our GMBI tools were still
- 5:39will start suggesting the types of graphs and charts that you
- 5:42should build and through the the user interface help guide you to
- 5:45creating those. And then once they're created, I
- 5:48can actually create the draft story for you talking about the
- 5:51relationships between those risk factors and diseases and even
- 5:55start suggesting other types of risk factors you might want to
- 5:59look at in your data or you know, what's your interventions
- 6:03or what you might want to do with that information once it's
- 6:06created the story. So it's it's just a a really
- 6:09nice way to kick start that process and make it easier for
- 6:13business analysts to start interfacing with the data and
- 6:16getting the information that they need without a data analyst
- 6:20having to always create those visuals for them.
- 6:22Again, this is not about replacing jobs.
- 6:26Data analysts are still very much needed, especially in this
- 6:30case to help set the data up the right way, but then it makes it
- 6:33easier for everyone to interface with the data.
- 6:36Lastly, what would you say the biggest obstacle to adopting AI
- 6:40is? One of the biggest challenges
- 6:42still with Gen. AI is the policy issues that go
- 6:46around implementing Gen. AI and working with your state
- 6:51policy to actually be able to do that.
- 6:54We do have a lot of examples from the public health
- 6:56departments that have been successful there and we're
- 6:58working on getting a set of shareable tools to help others
- 7:02get through the policy aspects. The technology aspects are not
- 7:06always as difficult to implement, but when they are, if
- 7:09your internal IT department doesn't have the resources to
- 7:13deploy gender to the AI resources on AWS, we have
- 7:17several partners that we can work with depending on the use
- 7:20case, that have expertise in doing that and can help with
- 7:23that. Thank you so much for your time,
- 7:25Jim. You can find more information on
- 7:28AWS, including an on demand webinar with other industry
- 7:31partners on some of their successes in the show notes.
- 7:34That'll do it for today. We're back on Monday morning
- 7:36with more ASTO news and information.
- 7:38I'm Jensen Silvers, you're listening to the award-winning
- 7:41public health review Morning Edition.
- 7:43Have a great weekend.