Rural health systems caught between Medicaid cuts and AI promises

Rural providers in the United States have struggled with money for years. Now they face cuts to Medicaid that, by the estimates cited by STAT, come close to a trillion dollars over the next decade. The administration points to artificial intelligence as a way to save this part of the system. The "best way to help" rural communities, as Centers for Medicare and Medicaid Services head Mehmet Oz put it, is supposed to be "AI-based avatars". Officials have also promised concierge care for every American and AI nurses "as good as any doctor". The argument goes that technology, backed by a 50-billion-dollar rural health transformation fund, can remake care in the countryside.

Practice versus announcements

Lori Dwyer is president and CEO of Penobscot Community Health Care, a system that covers parts of Maine. She says plainly: "I don't think AI is going to lower the cost of care anytime soon. I don't think it's going to save us". Her system already uses documentation assistants. The benefits are real: less administrative burden, less burnout on teams, shorter documentation time. But, as Dwyer stresses, "it doesn't lower costs".

STAT spoke with dozens of rural providers, system heads and medical AI experts. The same picture keeps coming back. AI is expensive to adopt, especially for systems already in a difficult financial position, and its savings potential has not been proven so far. At the same time, some heads are speeding up deployments precisely because of the cuts. Trampas Hutches of MaineHealth says the system makes AI "a member of the care team" and that "the only way to meet demand is technology and AI", though he admits cost is a problem.

One survey cited in the report found that more than 70 percent of hospitals use predictive AI. But the question of what exactly they are buying comes up. James McHugh of Berkeley Research Group does not mince words: "There's a lot of junk on the market". In his view, many AI deployments in health care have not gone well so far, and adoption has consumed significant resources. The expectation that this will lower labor costs "usually doesn't pan out".

Infrastructure and inequalities

Technology does not have to be the newest to run into barriers. Telemedicine, often using solutions from a decade ago, still has trouble taking hold in remote areas, because of poor internet or unregulated billing rules. Some facilities do not have data in a form that AI systems can process: forms sent by fax or filled in by hand are hard to enter into digital systems. "If you work in a big system, you have a much better chance of access to this technology. If you run a private practice, a much smaller one", says James Jarvis of the Maine Medical Association.

Some vendors do report growth in the countryside as well. Aidoc, a company working on radiology AI, recorded a "clear increase" in more remote regions, reports Jesse Ehrenfeld, the company's chief medical officer. He adds that many of these places are poorly equipped: they have no AI governance or the capacity for deployments at scale. It is this mechanism, benefits for the large and barriers for the small, that makes technology, according to STAT's interviewees, likely to widen inequalities in health care rather than reduce them.