Bill Gates on the health AI horizon

PROGRAMMING NOTE: We’ll be off next week for the holidays but back to our normal schedule on Tuesday, Jan. 2.

Bill Gates — Microsoft co-founder turned global health philanthropist — sees big advances on the horizon for artificial intelligence in health care.

The technology is likely to “supercharge the innovation pipeline,” he wrote in a reflection on 2023 and a preview of the coming year, adding that “innovation is the key to progress.”

Gates pointed to several promising approaches for using the technology to improve health outcomes, including:

— Keeping antibiotic-resistant bugs at bay by creating an AI system that can help clinicians prescribe drugs smartly

— Aiding in finding pregnancy risks in ultrasounds, with AI eventually acting as a co-pilot for clinicians

— Evaluating an individual’s risk of HIV through an AI chatbot

— Helping create medical records to offer improved data for better clinical decision-making

Still, much of the progress Gates foresees lies outside the realm of AI.

He sees vaccine innovation that helped cut childhood deaths significantly in recent decades as one case study in innovating for maximum impact.

Looking ahead, he sees scalable solutions to malnutrition in children as a major target to improve health outcomes worldwide. And the same is true for climate change, he said.

This is where we explore the ideas and innovators shaping health care.

Your mother probably told you to brush twice daily and evidence keeps coming to confirm her wisdom: Researchers at Brigham and Women’s Hospital in Boston collated data on hospital patients who keep their choppers clean and found they stayed healthier while hospitalized.

Share any thoughts, news, tips and feedback with Carmen Paun at [email protected], Daniel Payne at [email protected], Ruth Reader at [email protected] or Erin Schumaker at [email protected].

Send tips securely through SecureDrop, Signal, Telegram or WhatsApp.

Today on our Pulse Check podcast, host Kelly Hooper talks with POLITICO health reporter Chelsea Cirruzzo, who explains how the Department of Health and Human Services has become one of the top federal agencies using artificial intelligence.

Prior authorization — when insurers require providers to get permission before offering a treatment — “is a universally hated process across the board,” according to Matthew Cunningham, vice president of AuthAI.

Doctors don’t like it when insurers question the treatments they prescribe.

For insurers, it’s a paperwork hassle that can lead to time-consuming disputes.

AuthAI, owned by longstanding health IT company Availity, promises to make it more efficient using artificial intelligence.

Lawmakers and the Biden administration are also looking at fixes, but AI could beat them to it.

Cunningham said AI tools have the potential to return the prior authorization process to its original intent: keeping costs down while preserving access to high-quality care.

“We’re trying to put them on the same side of the table,” he said of payers and providers. “The goal is better patient care.”

How so? Because both providers and payers invest so heavily in completing and reviewing prior authorization paperwork, Cunningham argues that AI tools could save both sides a significant amount of time and money.

By creating a tool that offers providers faster answers and payers more clinical information, they might be able to agree on how to use the tech in a way that works better for everyone — patients included.

Decisions “should be done in seconds, not in days” in most cases, he said.

The Department of Health and Human Services is using AI more than most other agencies — fourth only to NASA and the Commerce and Energy departments, POLITICO’s Chelsea Cirruzzo reports.

That’s according to a recent report from the Government Accountability Office, which helps Congress oversee the executive branch.

Here are four ways HHS uses AI:

1. Food and Drug Administration: The FDA is implementing an AI platform trained on data from 1,500 clinical trials to write clinical study reports using Phase I and II study data. According to HHS’ AI use cases inventory, the AI can “mimic the subject matter experts,” including clinicians and statisticians, to decipher the study design and interpret the results.
2. Administration for Strategic Preparedness and Response: The agency’s EmPOWER program uses AI in a downloadable voice-controlled app for first responders that can tell them how many Medicare beneficiaries with electricity-dependent medical devices are in the area.
3. National Institutes of Health: The NIH uses a tool that predicts the priority level and area of research for grant applications and then ranks submissions using the AI’s analysis, allowing highly ranked applications to be reviewed first.
4. Centers for Disease Control and Prevention: The CDC uses AI and machine learning to improve surveillance testing. Its tools detect tuberculosis in chest X-rays and identify cooling tower locations from aerial imagery to prevent Legionella outbreaks. The agency is also looking into an open-source AI model to improve transcriptions and transcribe videos used in studies for the National Center for Health Statistics.