Blab AI
Allegheny Health Network to Pilot AI-Driven Diabetes Care with UpDoc
Allegheny Health Network (AHN), a 16‑hospital system tied to insurer Highmark, has announced a partnership with Palo Alto‑based UpDoc Inc. to employ artificial intelligence for monitoring and adjusting medication in type 2 diabetes patients. The pilot, set to launch in Western Pennsylvania, aims to deliver continuous care between routine office visits and could reshape chronic‑disease management.
Under the collaboration, UpDoc’s conversational AI platform will be deployed in primary‑care clinics. Physicians will flag patients with type 2 diabetes who will then receive regular telephone or text check‑ins from an AI agent. The agent will periodically ask about symptoms, insulin use, and blood‑glucose readings, and will suggest medication adjustments when appropriate. If a patient’s glucose levels move outside a range defined by the treating clinician, the AI will prompt the patient to contact the doctor. Every interaction is logged in the patient’s electronic medical record for later review.
AI in medicine has largely been limited to administrative tasks such as notetaking and insurance appeals. The AHN pilot marks a shift toward patient‑facing, clinical AI. In a small Stanford Medicine study of 32 participants, the same UpDoc platform achieved effective insulin dosing nearly four times faster than standard care—median 15 days versus more than 56 days—and improved medication adherence to 83% compared with 50% in the control group. The study appeared in the American Medical Association’s JAMA Network Open in 2023.
The potential impact of the pilot is significant. A 2023 JAMA Open Network study, led by UpDoc chief technology officer Ashwin Nayak and colleagues, found that almost one‑quarter of the 33 million U.S. adults with type 2 diabetes have poor glycemic control, often due to suboptimal insulin dosing. Contributing factors include therapeutic inertia, physician time constraints, and competing demands in busy practices. By offering real‑time monitoring and dose‑adjustment recommendations, the AI platform could reduce the need for multiple physician appointments and ease provider burnout.
In the Pittsburgh metropolitan area, roughly 170,000 to 220,000 people live with diabetes, and the direct medical costs of treating the disease can reach $2.6 billion annually. Better glycemic control can lower the risk of long‑term complications—such as kidney disease, eye and nerve damage, and heart disease—thereby cutting overall treatment costs.
AHN’s statement highlighted the expected benefits of the partnership. "Allegheny Health Network is looking forward to introducing this innovative AI platform with a particular emphasis on empowering primary care clinicians," said Amy Crawford‑Faucher, chair of the AHN Primary Care Institute. "The future integration of UpDoc within our electronic health records will further strengthen care coordination, reduce administrative burden and enable our physicians to devote more of their time to high‑value, clinical decision‑making—ultimately driving improved long‑term health outcomes for our patients and improved population health."
UpDoc CEO Sharif Vakili described the AI agent as "adding a cook to the care team" and noted that the technology "is opening the door and allowing the AI agent to join the team." The company has raised $16 million in oversubscribed seed financing from investors including the American Diabetes Association, Cathay Innovation, Eli Lilly & Co., and Mayo Clinic. Vakili said the platform will be expanded to other chronic‑disease management uses in the coming months.
Experts offered cautious optimism. Holly Wiberg, assistant professor of Operations Research and Public Policy at Carnegie Mellon University’s Heinz College, called a patient‑facing interactive tool "an exciting frontier" that can reach more patients at a lower cost. Former University of Pittsburgh and UPMC critical‑care chair Derek C. Angus noted that while AI could "crack the knot of home‑based remote care," risks of "bad recommendations" and other side effects remain.
The pilot is currently underway, with plans to weave UpDoc’s platform into AHN’s electronic health records. The trial’s outcomes will determine whether the AI approach can scale to other chronic conditions and deliver measurable improvements in clinical outcomes and cost savings. Unresolved questions include the long‑term safety of AI‑generated medication adjustments, the accuracy of the recommendations, and how to manage potential regulatory oversight.
As the pilot progresses, stakeholders will monitor its effects on patient adherence, provider workload, and health‑care costs. The results could shape future deployments of AI in chronic‑disease management across the United States.
Under the collaboration, UpDoc’s conversational AI platform will be deployed in primary‑care clinics. Physicians will flag patients with type 2 diabetes who will then receive regular telephone or text check‑ins from an AI agent. The agent will periodically ask about symptoms, insulin use, and blood‑glucose readings, and will suggest medication adjustments when appropriate. If a patient’s glucose levels move outside a range defined by the treating clinician, the AI will prompt the patient to contact the doctor. Every interaction is logged in the patient’s electronic medical record for later review.
AI in medicine has largely been limited to administrative tasks such as notetaking and insurance appeals. The AHN pilot marks a shift toward patient‑facing, clinical AI. In a small Stanford Medicine study of 32 participants, the same UpDoc platform achieved effective insulin dosing nearly four times faster than standard care—median 15 days versus more than 56 days—and improved medication adherence to 83% compared with 50% in the control group. The study appeared in the American Medical Association’s JAMA Network Open in 2023.
The potential impact of the pilot is significant. A 2023 JAMA Open Network study, led by UpDoc chief technology officer Ashwin Nayak and colleagues, found that almost one‑quarter of the 33 million U.S. adults with type 2 diabetes have poor glycemic control, often due to suboptimal insulin dosing. Contributing factors include therapeutic inertia, physician time constraints, and competing demands in busy practices. By offering real‑time monitoring and dose‑adjustment recommendations, the AI platform could reduce the need for multiple physician appointments and ease provider burnout.
In the Pittsburgh metropolitan area, roughly 170,000 to 220,000 people live with diabetes, and the direct medical costs of treating the disease can reach $2.6 billion annually. Better glycemic control can lower the risk of long‑term complications—such as kidney disease, eye and nerve damage, and heart disease—thereby cutting overall treatment costs.
AHN’s statement highlighted the expected benefits of the partnership. "Allegheny Health Network is looking forward to introducing this innovative AI platform with a particular emphasis on empowering primary care clinicians," said Amy Crawford‑Faucher, chair of the AHN Primary Care Institute. "The future integration of UpDoc within our electronic health records will further strengthen care coordination, reduce administrative burden and enable our physicians to devote more of their time to high‑value, clinical decision‑making—ultimately driving improved long‑term health outcomes for our patients and improved population health."
UpDoc CEO Sharif Vakili described the AI agent as "adding a cook to the care team" and noted that the technology "is opening the door and allowing the AI agent to join the team." The company has raised $16 million in oversubscribed seed financing from investors including the American Diabetes Association, Cathay Innovation, Eli Lilly & Co., and Mayo Clinic. Vakili said the platform will be expanded to other chronic‑disease management uses in the coming months.
Experts offered cautious optimism. Holly Wiberg, assistant professor of Operations Research and Public Policy at Carnegie Mellon University’s Heinz College, called a patient‑facing interactive tool "an exciting frontier" that can reach more patients at a lower cost. Former University of Pittsburgh and UPMC critical‑care chair Derek C. Angus noted that while AI could "crack the knot of home‑based remote care," risks of "bad recommendations" and other side effects remain.
The pilot is currently underway, with plans to weave UpDoc’s platform into AHN’s electronic health records. The trial’s outcomes will determine whether the AI approach can scale to other chronic conditions and deliver measurable improvements in clinical outcomes and cost savings. Unresolved questions include the long‑term safety of AI‑generated medication adjustments, the accuracy of the recommendations, and how to manage potential regulatory oversight.
As the pilot progresses, stakeholders will monitor its effects on patient adherence, provider workload, and health‑care costs. The results could shape future deployments of AI in chronic‑disease management across the United States.
