AI and rural health care: A paradigm shift in America’s heartland
The use of artificial intelligence is happening in “unlikely places.” So says a recent working paper published by the National Bureau of Economic Research. While much of the nation is debating the future of AI, health care providers in rural America are pioneering new uses of it in their practices. As the leader of the U.S.’s largest rural health care system, I predict the impact of AI on rural health care will be consequential.
After a decade of decline before the pandemic, a recent study by the U.S. Department of Agriculture indicates that the population in rural areas is rising a bit. Rural counties across the country — defined as those with cities of up to 50,000 people — grew one-quarter of a percent from 2020 to 2022.
That tiny population surge, however, isn’t likely to mend the greatest demographic challenge for rural health care: recruiting and retaining enough clinicians to work in the medical specialties that are in short supply across rural America.
Three out of five federally designated medical provider-shortage areas are in rural regions. Rural communities have only 30 specialists per 100,000 people, compared to 263 specialists per 100,000 in urban regions. Meanwhile, 25% fewer rural physicians will be practicing by 2030 due to an aging workforce with looming retirements.
A paradigm shift
Nearly all of America’s doctors are experiencing burnout as the pace of their practice has relentlessly sped up in recent years. But many are encouraged by the potential for AI to help them improve efficiencies in ways that allow them to refocus on their patients, rather than trying to keep up with electronic health records.
A paradigm shift is happening in rural America as rural health providers come to embrace the idea that what we do won’t change, but how we do it must.
Clinicians working in Nebraska corn country at Bryan Health are now using AI-enabled software that takes notes on their phones with the press of a button so they can look their patients in the eyes instead of clicking away at a keyboard. The technology securely records a provider’s conversation with a patient during a visit using “ambient listening,” which is then transcribed in the electronic medical record.
Bryan Health CEO Russ Gronewold told me that one physician in Grand Island called this “career extension technology.” Another in Lincoln said he’ll look back at this moment as “one of the most pivotal moments” in his medical profession.
In the next few weeks, Bryan Health will also launch a new generative AI tool in its electronic medical record that’s designed to reduce the significant amount of time physicians spend responding to large volumes of patient messages. The new tool will “pre-populate responses,” but physicians will have the ability to tailor and edit each message before sending.
While recent studies have shown that the use of large language models may not actually save clinicians time, saving time might not be the only measurement that matters. As Dr. Michael Pfeffer, the chief information officer and associate dean at Stanford Health Care and Stanford School of Medicine recently shared, his team found that generated draft responses for patient messages “reduced cognitive burden” by giving physicians a place to start.
In the dairy country of Wisconsin and the upper peninsula of Michigan, Marshfield Clinic Health System is taking a slightly different approach when it comes to combating physician burnout related to patient messages. Marshfield will soon deploy AI technology within its electronic medical record to “reduce noise” for physicians by sorting and routing messages to the appropriate member of the care team.
According to Marshfield’s chief information and digital officer, Jeri Koester, nearly 60% of the system’s patient-initiated messages are related to prescription refills, scheduling, or completing forms — tasks that can be managed by a nurse or medical assistant. This new tool will “remove clutter” from physicians’ inboxes and allow them to instead focus on clinical and urgent messages that require their expertise and immediate attention.
When the U.S. Preventive Services Task Force recommended colorectal cancer screenings begin at age 45, clinical teams at Sanford Health gathered to determine how to manage screening for 100,000 newly eligible individuals in the rural Dakotas, where there was a limited supply of gastroenterologists. They developed their own AI model that includes additional risk factors that may put people at heightened risk for colon cancer. The model is designed to help physicians understand the risk of the patient in front of them without having to scroll through medical records, saving doctors time they can spend with the patient instead.
AI’s next chapter
AI can and will do much more than streamline administrative tasks. These technologies will soon serve as another tool in clinicians’ black bags — wherever their practices are located — metropolis, suburb, or farming town. AI-enabled clinical decision-support tools will help to identify serious health threats, improve diagnoses and calibrate the precision of medical treatments.
A new effort, led by the White House, is focused on developing a voluntary framework of health care AI commitments. The initiative does not shy away from the trust barriers that must be addressed, both for consumers and health care practitioners, including ensuring data that an AI model is trained on is representative of the population it will serve as a guard against bias.
So far, 38 payers and health systems have come together in this collaboration to determine how to harness AI models safely, securely and transparently. Bringing diverse voices to the table is a critical component of this work.
The challenges are not the same for urban and rural health care systems.
Rural America has some of the highest rates of late-stage breast cancer diagnoses in the country. A recent report from the Centers for Disease Control and Prevention found that rural Americans are more likely to die early from preventable causes like cancer. The potential to reverse this trend through new AI technologies that forecast the risk of disease will be a game-changer for rural clinicians and the patients they care for.
Preterm birth is another rural issue. In rural northern Minnesota, one of the poorest and most geographically isolated regions in the state, OB-GYNs at Sanford Bemidji Medical Center launched a pilot using FDA-cleared AI-enabled non-stress test belts to monitor fetal heart rate and the presence of contractions among patients who may be at higher risk of pre-term delivery, allowing them to intervene earlier to ensure the best possible maternal health outcomes.
My children are enamored with a movie called “The Croods,” which tells the story of a family of prehistoric cave-dwellers about as far from high-tech living as one could imagine. In the original 2013 film, the stubborn, cautious father (“Fear keeps us alive!”) refuses to let anyone leave the cave except for brief forays at daybreak to gather food, admonishing his children to “never not be afraid.” But they defy that edict and eventually make it to the other side of the mountain to see a peaceful paradise there.
It makes sense to be cautious about AI in health care, no matter where one lives or practices medicine. But some health care providers, including those in the most remote and rural locations in our nation, have already crept over the mountain and have seen a new world of promise.
Bill Gassen is president and CEO of Sanford Health, the largest rural health system in the United States, with headquarters in Sioux Falls, South Dakota.
link