AI and the Future of Southern Oregon Medicine, Part 1 of 3
There is a version of the Southern Oregon provider shortage story that is told as a local problem. Rural geography. Housing costs. Wildfire risk. The difficulty of recruiting physicians to a region without a major academic medical center. These factors are real, and they matter. But they are not the primary explanation for what is happening here, and treating them as primary leads to responses that are insufficient to the scale of the problem.
The more accurate version of the story is this: Southern Oregon is experiencing, acutely and early, a structural collapse in primary care capacity that is occurring across the United States. The local conditions accelerate and amplify it. They did not cause it. And the national data on where this is heading should concern every health system leader, CCO executive, and state legislator who is responsible for healthcare access in this region.
Isaac Kohane — physician-scientist, chair of the Department of Biomedical Informatics at Harvard Medical School, associate professor of medicine at Brigham and Women’s Hospital, and editor-in-chief of the New England Journal of Medicine AI — has spent four decades at the intersection of medicine and computing. When he describes the state of primary care in America today, he is not speaking in projections. He is describing what he can already see from his own institution.
Mass General Hospital and Brigham and Women’s Hospital — two of the most prestigious academic medical centers in the United States — have officially announced they are no longer accepting new primary care patients. In Kohane’s own subspecialty of pediatric endocrinology, half of training slots nationally are going unfilled. In pediatric developmental disorders, including autism, the same is true. The American Association of Medical Colleges estimates that by 2035, the United States will be short approximately 50,000 primary care physicians.
Read those figures again in the context of Jackson and Josephine counties, where fewer than 60 primary care providers serve every 100,000 residents against a state average of 83, where La Clinica’s waitlist runs six to nine months, and where most Asante family medicine clinics are closed to new patients. Southern Oregon is not falling behind a national average that is holding steady. It is falling behind a national average that is itself in freefall.
This is the structural argument. The shortage is not a gap that will close when the housing market softens or when the next cohort of medical school graduates makes different specialty choices. It is the predictable output of a system in which the financial and professional incentives of medical training have systematically steered physicians away from primary and preventive care for decades. Interventional radiology is oversubscribed. Dermatology has more applicants than slots. Primary care, pediatric endocrinology, behavioral health — these are the specialties that carry the greatest burden of population health, and they are the ones emptying out.
Kohane is direct about what this means: “Go around medical schools and ask, who’s becoming a primary care doctor? Almost nobody.” This is not a complaint about individual choices. It is a description of a system that has, through its reimbursement structure, its prestige hierarchy, and its debt burdens, produced exactly the workforce distribution it incentivized.
Why the Pipeline Argument Fails
The instinctive policy response to a shortage is to fill the pipeline. Recruit more aggressively. Offer better loan repayment. Create rural training programs. Build more capacity in the specialties that are underfilled.
These interventions are worth pursuing. Oregon has taken some meaningful steps — OB 476 creates a pathway for internationally trained physicians to receive provisional state licenses, and the HPSA designations that cover Jackson and Josephine counties unlock federal support for recruitment and retention. None of this should stop.
But the pipeline argument fails as a primary strategy for one simple reason: the timeline is incompatible with the urgency. A physician who begins medical school today will complete residency training in primary care around 2033 at the earliest. The AAMC’s 50,000-physician shortage figure is a 2035 projection. The gap between now and any pipeline solution is a decade or more, and during that decade, the existing primary care workforce continues to age out, burn out, and redirect toward higher-reimbursement specialties.
In Southern Oregon, that timeline is not abstract. It is the behavioral health patient who waits three months for an appointment and doesn’t make it. It is the person with uncontrolled hypertension who has no primary care relationship because there is no panel to join. It is the undiagnosed child whose symptoms don’t fit a clean pattern and whose case requires the kind of differential reasoning that a specialist might catch but a stretched urgent care system will not.
The pipeline will not close this gap in time. Something else has to.
The Misdiagnosis Problem Nobody Is Talking About
Alongside the access shortage is a problem that receives even less attention in regional healthcare conversations: the epidemic of misdiagnosis and undiagnosed patients.
Kohane serves as the principal investigator of the coordinating center of the Undiagnosed Network, a consortium of twelve academic hospitals spanning the University of Washington, Stanford, UCLA, Baylor, Harvard’s hospitals, and the NIH. The network sees several thousand patients annually — people who have been to multiple providers, received multiple tests, and still have no diagnosis. They are in pain. Some have been sick for years.
This population, large as it is, represents a fraction of the true burden. The patients who reach the Undiagnosed Network are the ones with enough resources, persistence, and access to navigate referrals to major academic centers. Behind them is a much larger group of people whose symptoms are being managed, suppressed, or ignored without an accurate underlying diagnosis — people who are being failed not by malice but by the structural limitations of an overstretched system where no single clinician has the time or the angle to see the full picture.
One case illustrates both the failure and the possibility. A child was having trouble walking, trouble chewing, and had developed intractable headaches. His mother brought him to multiple doctors. Multiple imaging studies were ordered. No diagnosis was made. The child remained in pain.
His mother typed the accumulated reports into GPT-4 and asked what the diagnosis might be. The response was tethered cord syndrome. She brought the imaging to a neurosurgeon, who confirmed it.
This is not a story about a mother being smarter than her son’s doctors. It is a story about what happens when the tool is matched to the problem. The doctors she saw were competent. They were also operating in a system that does not give any individual clinician the time, the computational support, or the wide-angle view across an entire case history that accurate diagnosis of complex presentations requires. The AI had all of it at once.
Kohane’s comment on this pattern is precise: most clinicians he knows do not have what he calls the “Google reflex” — the instinct, when facing an unknown, to ask a powerful information retrieval tool what it knows. They use Google in their personal lives. They do not use it in the clinic. They are too busy, too constrained by workflow, and too formed by a training culture that values internal recall over external augmentation.
In a region where the ratio of patients to providers is already stressed, and where complex cases have fewer specialist options to be referred to, the cost of that missing reflex is borne directly by patients.
What This Means for Southern Oregon
The specific contours of Southern Oregon’s shortage make the structural argument even more urgent here than in many other underserved regions.
This is not a community that lacks healthcare institutions. Asante Health System invested $177 million in community benefit spending in fiscal year 2022 — 202% of the floor assigned by the Oregon Health Authority, the highest percentage in the state. AllCare and Jackson Care Connect coordinate care for a significant portion of the region’s Medicaid population. Providence operates facilities here. Southern Oregon University’s health sciences programs and the OHSU nursing school on the Ashland campus represent genuine workforce development infrastructure.
What this community lacks is not commitment. What it lacks is capacity — specifically, the primary and behavioral health capacity to match the volume of need, particularly among populations with complex, chronic, and undiagnosed conditions. And it lacks that capacity for structural reasons that are not going to be resolved by any single recruitment program or facility expansion.
The question that Kohane’s work forces is this: if the pipeline cannot close the gap in time, and if the structural forces driving the shortage are national in origin and decades in the making, what response is actually proportionate to the problem?
His answer — and it is not a speculative one — is AI augmentation of the workforce that exists. Not replacement of physicians. Not a technological fix that sidesteps the human relationships at the center of good care. But a fundamental rethinking of what a nurse practitioner, a physician assistant, or a community health worker can do when they are supported by tools that have been trained on the full body of medical literature, can hold the entire history of a patient’s case in view at once, and can flag patterns that no individual clinician could be expected to catch alone.
That argument is the subject of the next article in this series. But it begins here, with a clear-eyed account of the problem it is responding to.
The Southern Oregon provider shortage is not a local anomaly. It is the local expression of a national structural failure. The response has to be proportionate — not to the scale of what has been tried before, but to the scale of what the situation actually requires.
This is the first article in a three-part series on AI and the future of medicine in Southern Oregon. Part 2 examines what AI augmentation can concretely do for a region without enough doctors. Part 3 addresses the risks, the resistance, and what local institutions need to decide. ReImagine Healthcare is a subsidiary of Flourish Charity, a 501(c)(3) nonprofit. We publish research and analysis on healthcare system design in Southern Oregon. We welcome responses, corrections, and partnership inquiries at reimagine-healthcare.org.