When the System Cannot Stop, Patients Carry the Risk

A hospital cannot pause for an incident the way an office can. When clinical systems cannot stop safely, the risk lands on patients.

Goran Svetlecic’s wartime dilemma is sharpest in a hospital, because both horns of it hurt patients. Cut the suspect segment and you may disrupt care: delay a dose, blind a care team, force a unit onto paper mid-shift. And cutting tells a capable adversary its channel is burned, which can push it toward destructive action or into dormancy behind a door nobody found. Wait and watch instead, and the adversary keeps walking the graph toward the rest of the fleet. There is no clean answer in wartime. Which is exactly why the work has to happen in peacetime.

The fleet is a graph

A hospital’s device fleet is thousands of endpoints deep: pumps, monitors, imaging, laboratory analyzers, building systems, and the workstations that manage them. Many share credentials. Many sit on flat networks where one compromised device can reach the rest. Few have an identity of their own, which means the network cannot tell a legitimate pump from an intruder using a pump’s address.

Read that as a graph and the problem clarifies. Every shared credential and every flat segment is an edge. The clinical systems that cannot go down are the crown jewels. A compromised device does not need a million paths. It needs one edge into the flat network, and the fan of routes to the rest of the fleet opens from there.

Peacetime: segment the graph

Segmentation is peacetime choke-point cutting. Segment by clinical function, so a compromised device in one unit cannot traverse the fleet. Give each device an identity, a network segment, and least authority: it may reach exactly the systems its function requires, nothing more. Each segment boundary is a cut edge, made while cuts are cheap, tested, documented, and reversible in a maintenance window.

Downtime is not an IT event in a hospital. It is a patient-safety event. Segmentation is what lets a team isolate the threat without isolating the patient. It converts the wartime dilemma from an impossible choice into a survivable one: the cut is survivable precisely because the graph was already divided before anyone was on it.

HIPAA is enforceable now

There is a dangerous misunderstanding circulating in healthcare: that the HIPAA Security Rule overhaul, pushed to a July 2027 target, means enforcement is on hold. It does not. The current rule is fully enforceable now, and the overhaul is delayed, not cancelled. Organizations treating the delay as a holiday from enforcement are building a finding for their next audit, or their next incident review.

The same logic applies to device security. Waiting for a final rule to segment a fleet is waiting for permission to do what the current rule already requires: protect electronic protected health information with access controls, audit controls, integrity controls, and transmission security that match the risk.

A window that is open now

The FDA’s draft guidance on medical device cybersecurity has comments open until November 24, 2026, with a workshop following on December 2-3, 2026. That is a live window to shape expectations while they are still draft: what evidence manufacturers should provide, what hospitals should demand in procurement, and how postmarket vulnerabilities get handled. Once guidance finalizes, the conversation moves from shaping to complying. The time to have a position is while the docket is open. It is also a peacetime window in the other sense: the time to segment is before the incident, not during it.

What to do this quarter

Inventory the fleet, including the devices nobody owns on paper. Map it as a graph: which devices can reach which clinical systems, through which shared credentials and flat segments. Segment by clinical function, starting with the units where isolation is hardest. Give each device an identity and least authority. Then rehearse the stop: isolate a segment in a drill, without stopping care, and measure how long it takes and what breaks. A stop that has never been rehearsed will not happen cleanly during a real incident.

When the system cannot stop, patients carry the risk. Cut the graph in peacetime, so the team never has to choose between security and care in the middle of an incident again.

Source: Goran Svetlecic, “Attack Path Management in Peacetime and Wartime,” LinkedIn, October 27, 2026: https://www.linkedin.com/pulse/attack-path-management-peacetime-wartime-goran-svetlecic-cissp-eqeif/

Recent Posts

Tag Cloud

Stop Threats Before They Execute

Your free Kernel-Level Defense Buyer’s Guide is ready to download.

By providing my email address, I consent to receive emails and text messages—including newsletters and marketing communications—from creators of Warden Secure, Cyber Strategy Institute, our flagship zero-trust platform for ransomware prevention, and agree to the Terms and Privacy Policy. You may unsubscribe at any time.