Another Kind of Bump in the Road

The small but constant hassles of modern healthcare often feel like a second illness: patients juggle forms, portals, prior authorizations, and billing disputes while already exhausted by their actual medical problems, and clinicians themselves now spend large chunks of their day on documentation instead of direct care, which can delay treatment and erode trust in the system. I trust my doctors but I do not trust the healthcare insurance industry. 

Beneath these everyday bureaucratic frictions lies a more serious safety problem: a widely discussed 2016 analysis by Johns Hopkins researchers, estimated that medical errors in U.S. hospitals contribute to roughly 251,000 deaths a year—about 9.5% of all deaths—leading the authors to argue that such errors should be recognized as the nation’s third leading cause of death, after heart disease and cancer. Critics have since suggested the figure is probably too high.  However, the complex systems, communication breakdowns, medication errors, and paperwork of hospital care can directly translate into avoidable harm rather than healing.

Amazon has made a number of forays into the problems of healthcare bureaucracy without much success. But, this time the behemoth is leaning heavily on AI to attack its health offerings by turning a lot of back-office friction into machine work instead of human work.

In Amazon’s One Medical, a new “Health AI” assistant tied directly to members’ records may answer questions about symptoms, explain test results, renew medications, and book visits, so patients don’t have to navigate separate portals, phone trees, and messaging threads for routine requests.

In the administrative side, Amazon’s AWS teams are rolling out multi‑agent AI systems that automatically assemble clinical documentation, verify insurance eligibility, and submit prior‑authorization requests, shrinking processes that used to take days into minutes and continuously tracking status updates without staff chasing faxes and forms.

Amazon Pharmacy applies similar tools to prescription workflows, using machine‑learning models and generative AI to speed up data entry and order validation, generate real‑time insurance price estimates, and power chatbots that surface policy and safety information for staff, which has cut average prescription processing times dramatically and lets pharmacists spend more time on clinical issues instead of clerical ones.

I have been using Amazon Pharmacy since they began years ago. It is very efficient and things come to your doorstep. Just like with Amazon products, when you place an order, you specify the address or change the address.

For more about what Amazon is doing…