.svg)
.png)
How do hospitals secure unmanaged IoMT devices without disrupting patient care?
Hospitals secure unmanaged IoMT devices by discovering them agentlessly, classifying each one by identity, and enforcing least-privilege policy through the network hardware already in place, so no endpoint agents, downtime, or re-cabling touch clinical workflows. St. Luke’s University Health Network deployed identity-based microsegmentation across 15 hospitals and roughly 85,000 production devices in about 46 days.[2]
The core problem is that most connected medical devices cannot defend themselves. Infusion pumps, imaging systems, and patient monitors ship with fixed operating systems the hospital is not permitted to modify, and much of the installed base predates the FDA’s premarket cybersecurity guidance for medical devices.[8] Endpoint agents are off the table. So are the change windows and re-cabling projects that traditional segmentation demands, because a patient monitor that goes quiet during a maintenance window is a clinical incident, not an IT ticket.
Identity-based microsegmentation breaks the deadlock in four moves. Discover every device by observing the network, with nothing installed. Classify each one by identity: what it is, who uses it, how it behaves. Simulate least-privilege policy against live clinical traffic until nothing legitimate is blocked. Then enforce, group by group, on the network hardware the hospital already owns.
That sequence is why deployments measure in weeks, not years. The St. Luke’s 46-day microsegmentation deployment went from no microsegmentation to network-wide enforcement two weeks ahead of its own 60-day commitment.[2] MultiCare Health System secures tens of thousands of devices across its Washington State hospital and clinic network.[7] Its CISO, speaking in the MultiCare Health System case study, described what unsegmented complexity costs:
“The people, processes, and technology, each of those comes with their own attack surface. So the impact is non-trivial.”
Jason Elrod, Chief Information Security Officer, MultiCare Health System
For the device-by-device playbook, including per-modality policy patterns and the discovery-tool integrations, see the agentless IoMT and medical device segmentation guide.
Zero trust architecture for healthcare systems
Zero trust architecture for healthcare systems applies NIST SP 800-207 to an environment where most endpoints are medical devices that cannot authenticate themselves. The workable sequence: build a device identity inventory, map real clinical traffic flows, simulate least-privilege policy against live traffic, then enforce microsegmentation through existing network hardware, starting with the highest-risk device groups.[4]
NIST SP 800-207 defines zero trust around policy decision points and policy enforcement points that evaluate every connection continuously. Hospitals meet that model with a constraint few other industries face: a large share of endpoints cannot run software, present credentials, or participate in their own defense. Zero trust for healthcare has to be delivered from the network, not from the endpoint.
In practice, health systems phase the work. Phase one builds the identity inventory, because a policy engine is only as good as its knowledge of what is connected. Phase two maps flows between clinical systems, the EHR, PACS, and vendor maintenance paths. Phase three simulates policy and reviews the exceptions with biomedical engineering. Phase four enforces, beginning with unmanaged IoMT devices, then clinical workstations, then users.
Zero trust in a hospital also covers people. Visiting clinicians and per-diem staff move between facilities and shared workstations every day, so policy anchored to a network location breaks the moment care delivery moves. Identity has to anchor policy for users exactly as it does for devices, and the same enforcement layer serves both.