A new platform is being developed by several major universities with the aim of combating misinformation about healthcare and public health policies.

It's spearheaded by the University of Pittsburgh, the University of Illinois Urbana-Champaign, UC Davis' Health Cloud Innovation Center and Amazon Web Services. 

The platform, called Project Heal, will use machine learning, generative artificial intelligence and predictive analytics to help public health officials shift from reactivity to proactivity in their efforts to address health misinformation.

WHY IT MATTERS

Misinformation has long been a public health challenge, but the pandemic put it in stark relief. The U.S. Department of Health and Human Services, for instance, estimates that COVID-19 vaccine misinformation cost $50 million to $300 million per day during 2021.

That figure was based on a portion of the cost of voluntarily declining a COVID-19 vaccination – including hospitalizations, a valuation of lives lost and long-term morbidity – an estimated $1 billion. 

Moreover, as such misinformation proliferates, the effort to combat it causes burnout among public health officials, clinicians and caregivers.

With its new open source toolkit, Project Heal's creators aim to inform public health officials precisely when their clarifying communications are most needed to improve outcomes and empower more individuals to make better-informed decisions about their health.

The platform classifies and detects emerging misinformation before it can go viral in communities. 

As Project Heal collaborators explain in an AWS blog post, trained machine learning models classify the likelihood of a statement having misleading content and allow for categorization based on the statement entities and context. 

They then evaluate misleading statements to help score the severity of the threat to human health.

To account for unique cultural, historical and linguistic nuances affecting how various demographics respond to misinformed rumors that hinder health equity and corrective counter-messaging, the system also uses retrieval-augmented generation to optimize large language models' outputs to generate more personalized messaging for targeted communities.

Once developed and deployed on AWS cloud, the platform will allow public health officials to more efficiently manage workloads by shifting community education tasks from reactivity to proactivity. 

THE LARGER TREND

Testing the prototype, public health experts were enthusiastic that the system would be a source of support – particularly where there is an intentional delineation between verified and unverified data sources, the collaborators said.

Groups with low health literacy tend to be more susceptible to misinformation, said Denise Scannell, department manager of health behavior and social sciences at MITRE.

"One of the things that doesn't exist today – but we're looking at developing – is early warnings, so we can work with public health folks to inoculate against mis- and disinformation before it amplifies within the local community," she said in 2022. "That's critical."

Through a partnership with Florida International University, MITRE helped to identify COVID-19 vaccine misinformation and disinformation and aided active community interventions to counter these messages in the Haitian community, she said.

"We increased vaccination rates from close to zero when we started there to somewhere in the thousands."

ON THE RECORD

"It is clear that health misinformation remains a major threat to patient wellness in the U.S. and beyond," Project Heal collaborators said.

Andrea Fox is senior editor of Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.