The AI arms race over your medical bill

The ideas and innovators shaping health care

Jan 05, 2024 View in browser   By Daniel Payne, Ruth Reader, Carmen Paun and Erin Schumaker WEEKEND READ You're stilling getting paper bills, but AI may be processing them. | AP An artificial intelligence arms race is afoot between medical providers and insurers seeking advantage in dividing the $4 trillion in health care expenses Americans accrue each year. “Everyone’s trying to maximize revenue while toeing the line on, effectively, fraud,” Nick Stepro told Daniel. He’s the chief product and technology officer at Arcadia, a company that works with health care organizations on both sides of the divide looking to build the technology. It’s true, he believes, that advanced AI will bring a host of positive impacts to the health system — but perhaps not before it further inflames the feuds over bills between your health plan and your medical provider. What’s up? For providers, the dream is an AI tool that can quickly and aggressively code procedures and file claims. Insurers — and the government agencies that pay for health care — want comparable technology to scrub those bills. The provider side: Leaders at several health systems told Daniel they want to reap savings by coding their bills more accurately, thus reducing their liability. Others highlight the benefits of reducing the labor needed to complete administrative tasks. And nearly all providers are keenly interested in making their staff happier by reducing burnout. Payers respond: Insurers say they’re making major investments in AI to cut their own administrative costs and weed out fraud in the bills they receive. “It feels like the perfect storm coming of the technology really becoming a more significant asset to the company if deployed correctly,” said Craig Richardville, chief information officer of Salt Lake City-based Intermountain Health, which does business as both a provider and payer. Government’s stake: Agencies already use AI to combat fraud, according to the National Health Care Anti-Fraud Association, an alliance of government agencies and private insurers. A spokesperson for the Centers for Medicare and Medicaid Services said that “CMS continually assesses opportunities to safely and responsibly leverage new, innovative strategies and technologies, including AI, to more effectively accomplish its mission.” At the same time, the agency has to answer to patients skeptical of strict anti-fraud measures that can result in denied care. Daniel has more on the race here. WELCOME TO FUTURE PULSE St. John, U.S. Virgin Islands | Shawn Zeller/POLITICO This is where we explore the ideas and innovators shaping health care. Researchers at the University of Maryland and the National Institutes of Health identified the enzyme that makes urine yellow and its significance goes beyond just satisfying their curiosity, they argue: The discovery can help with further research into jaundice and even lead to a better understanding of the gut microbiome’s role in human health. Share any thoughts, news, tips and feedback with Carmen Paun at cpaun@politico.com, Daniel Payne at dpayne@politico.com, Ruth Reader at rreader@politico.com or Erin Schumaker at eschumaker@politico.com. Send tips securely through SecureDrop, Signal, Telegram or WhatsApp. TECH MAZE Kids in Utah will soon need mom and dad's permission to open a social media account. | AFP via Getty Images In less than two months, Utah will become the first state to curtail social media use for all residents under 18, requiring them to get parental permission to open accounts, allow parents or guardians access and restrict nighttime usage. The idea is to combat mental health harms tied to scrolling for hours on the sites. And several states will soon follow. A new study from researchers at Boston Children’s Hospital and Harvard’s T.H. Chan School of Public Health shows why some tech giants are lobbying to stop Utah’s law from becoming a trend. “We hope to catch the attention of lawmakers who are looking for these numbers to provide context for why government regulation is needed,” said researcher Amanda Raffoul, the lead author on the study. How so? The study found that in 2022, YouTube, TikTok, Snapchat, Facebook and Instagram collectively generated $11 billion in revenue from ads targeted at users under 18, mostly from those between 13 and 17. Meanwhile, researchers say Snapchat derived 41 percent of its revenue from ads aimed at users under 18 — the most of any platform. YouTube, which has the largest number of young users, raked in the most revenue from the under-12 set: an estimated $959.1 million. Instagram, which isn’t nearly as popular with kids, generated more than $800 million in ad revenue aimed at those under 12, according to researcher estimates, implying that Instagram makes more money off of fewer users. On Capitol Hill: Bipartisan Senate legislation from Republicans Katie Britt of Alabama and Tom Cotton of Arkansas and Democrats Brian Schatz of Hawaii and Chris Murphy of Connecticut would bar children under 13 from using social media. Meanwhile, Sens. Ed Markey (D-Mass.) and Bill Cassidy (R-La.) are pushing for an update to the Children and Teens’ Online Privacy Protection Act that would bar targeting ads to kids. Even so: None of the platforms studied break out ad revenue derived from children, so the estimates rely on market research and survey data. SAFETY CHECK How many calories are really in that double cheeseburger? | Getty Images Counting calories is a perennial challenge for people trying to lose weight — so much so that Congress included a provision in Obamacare requiring chain restaurants to include counts on their menus. So a bot that could accurately estimate a meal’s tally could be useful. But that AI use case hasn’t arrived yet, according to a research letter in JAMA Network Open by nutrition specialists from Taiwan, Malaysia and Indonesia. How so? The researchers asked a current and earlier version of ChatGPT, the bot from the San Francisco firm OpenAI, to estimate the calories in eight meals. Neither could provide an estimate within plus or minus 10 percent of the true tally more than half the time. They also failed to accurately assess fat, carbohydrate and protein counts most of the time. Even so: The researchers hope ChatGPT could help people understand the nutritional makeup of the food they eat if the program trains on better data.   Follow us on Twitter Carmen Paun @carmenpaun Daniel Payne @_daniel_payne Ruth Reader @RuthReader Erin Schumaker @erinlschumaker   Follow us

To change your alert settings, please log in at https://www.politico.com/_login?base=https%3A%2F%2Fwww.politico.com/settings This email was sent to rouf@idiot.cloudns.cc by: POLITICO, LLC 1000 Wilson Blvd. Arlington, VA, 22209, USA Unsubscribe | Privacy Policy | Terms of Service