FDA’s 2018 device playbook made cyber incidents a hospital staffing job

The October 2018 medical device cybersecurity playbook asked hospitals to run device incidents through incident command. That meant staffing four jobs that fell between departments: the inventory, delegated authority to pull devices, regional mutual aid and supplier duties written at purchase.

Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.

Two-minute summaryThe dek and the first sentence of each section, quoted from the article

The October 2018 medical device cybersecurity playbook asked hospitals to run device incidents through incident command. That meant staffing four jobs that fell between departments: the inventory, delegated authority to pull devices, regional mutual aid and supplier duties written at purchase.

  1. The week On Monday, October 1, 2018, the Food and Drug Administration issued a statement from its Commissioner on the agency’s efforts to strengthen its medical device cybersecurity program.
  2. The inventory: knowledge someone has to keep The playbook called knowing what systems are connected to the network a foundational preparedness principle.
  3. The command roles: authority delegated in advance The playbook built on the hospital incident command system, which defines a hospital incident management team to lead the response to all-hazards incidents.
  4. The region: capacity beyond the hospital’s walls The playbook’s most distinctive idea was in its title.
  5. The manufacturer: duties written at purchase The playbook began its preparedness section with procurement.
  6. Training the people who notice first The playbook gave training its own section, and its reasoning places the front line of detection with clinical staff.
  7. What changes for each department Seen as staffing, the playbook asked something specific of each department that touches devices.
  8. Objections: optional guidance, small hospitals, rare events The first objection is that the playbook was not guidance and created no obligation.
  9. Staffing the plan: who does what on an ordinary Monday The playbook was written for the incident.
  10. With hindsight The FDA’s device cybersecurity work continued along the premarket track that the October 2018 statement previewed.

The week

On Monday, October 1, 2018, the Food and Drug Administration issued a statement from its Commissioner on the agency’s efforts to strengthen its medical device cybersecurity program. The statement announced that the MITRE Corporation, with support from the FDA, had released a Medical Device Cybersecurity Regional Incident Preparedness and Response Playbook. It described the playbook as a tool to help health care delivery organizations, such as hospitals, respond better to medical device cybersecurity incidents, and said readiness activities in it included developing a medical device inventory and conducting training exercises.

The playbook itself, labeled Version 1.0 and dated October 2018, said it was prepared by MITRE under contract with the FDA. It stated that its views did not constitute agency guidance, policy or legally enforceable requirements, and that following it did not constitute compliance with the Federal Food, Drug, and Cosmetic Act or any other law. Its background section explained the gap it was meant to fill. After the WannaCry and Petya/NotPetya attacks, the FDA had heard from stakeholders that health care delivery organizations did not know with whom to communicate, what actions they might take and what resources were available. MITRE, with FDA support, had consulted hospitals of different sizes, state health departments, device manufacturers and government agencies to build it.

The playbook’s primary audience was hospital staff involved in medical device cybersecurity preparedness and response: clinicians, healthcare technology management professionals, and IT, emergency response, risk management and facilities staff. Its scope was device cybersecurity issues that affect device function, with a focus on threats with the potential for large-scale, multi-patient impact and patient safety concerns. It said it was not meant for day-to-day patch management, and that it presented target capabilities, which not every hospital would be able to execute fully because of operational constraints. Its structure followed the incident response life cycle in NIST Special Publication 800-61 revision 2: preparation, detection and analysis, containment, eradication and recovery, and post-incident activity.

The same statement announced two memoranda of understanding to create information sharing analysis organizations for medical device cybersecurity, and said the FDA was discussing a memorandum of agreement with the Department of Homeland Security on coordination and information sharing about device vulnerabilities and threats. It said the FDA planned, in the coming weeks, to publish a significant update to its 2014 premarket cybersecurity guidance, which would highlight the use of a “cybersecurity bill of materials,” a list of commercial or off-the-shelf software and hardware components of a device that could be susceptible to vulnerabilities. It also said the FDA had developed its own internal playbook for responding to device cybersecurity threats, vulnerabilities and incidents.

The statement said the FDA was not aware of any reports of an unauthorized user exploiting a cybersecurity vulnerability in a medical device in use by a patient, but that the risk of such an attack persisted. It listed exercises that had informed the agency’s work, including a DHS-led functional exercise in 2013, a DHS-led national level exercise in 2016 and a MITRE-convened tabletop exercise for the FDA in 2017. It also said the FDA had proposed, in its fiscal year 2019 budget, a Center of Excellence for Digital Health that would support a cybersecurity unit.

The 2018 playbook is easy to file as a technical document about devices. Its own content points somewhere else. Much of it describes people: who keeps the inventory, who sits in the hospital’s incident command when devices are affected, who calls the manufacturer, who the hospital can borrow devices from, who has the authority to take a device out of service at night. The playbook treated a device cybersecurity incident as an emergency like any other, run through the hospital’s existing incident command system, with new roles added.

Our argument is that the playbook’s real demand on hospitals was organizational. It asked them to staff four jobs that sat between clinical engineering, IT security and emergency management, and that, as we read the playbook, no single department was positioned to own. The four are:

  • The inventory: a current record of every connected device, with an owner, a location on the network and a maintenance status.
  • The command roles: named people with device and security expertise inside the hospital incident management team, and decisions delegated in advance.
  • The region: mutual aid with other hospitals and public partners, including loaner devices and patient diversion, arranged before an incident.
  • The manufacturer: duties written into the purchase, so that the supplier takes part in exercises and supplies what responders need.

The four are mutually exclusive in what they govern. The inventory is about knowledge. The command roles are about authority. The region is about capacity beyond the hospital’s walls. The manufacturer is about obligations that only a supplier can meet. Together they cover what a hospital needs to respond to a device incident: knowing what it has, deciding what to do, getting help and getting the supplier to act. After the four parts, we take the main objections and then set out how a hospital can staff the plan.

The playbook was not a regulation, and it said so. Hospitals had no legal duty to follow it. But hospitals that participate in Medicare already had a legal duty that the playbook explicitly cited: the emergency preparedness condition of participation, which requires an all-hazards emergency plan, a risk assessment, policies, a communication plan and training and exercises. The playbook framed device cybersecurity as part of that all-hazards plan. That framing is what made it an operations document.

The timing also mattered. The same statement that released the playbook announced information sharing organizations for manufacturers, a planned agreement with DHS and an upcoming update to premarket guidance. Those were measures aimed at manufacturers and at government coordination. The playbook was the one item in the package aimed at the people who operate devices every day. It asked hospitals to do their part of what the statement called shared responsibility, and it described that part almost entirely in terms of roles, relationships and agreements.

The inventory: knowledge someone has to keep

The playbook called knowing what systems are connected to the network a foundational preparedness principle. It asked hospitals to maintain a centrally managed, baseline set of information about each medical device, including legacy devices and devices on research or other nonstandard networks, and to update it regularly, ideally in real time or when there are changes.

The list of fields it proposed reads like a job description. Device name and description. Physical location. Logical location: IP address, switch port, wireless access point. Device owner and manager. Maintenance parameters, including whether the manufacturer still supports it and who maintains it under what service level agreement. Operational status, including operating system and patch status. Embedded components, through a software bill of materials, with versions and patch status. Interactions with and dependencies on other devices and IT resources. Log files that capture operating and diagnostic information.

Those fields come from different places in a hospital. Physical location and clinical owner belong to the department that uses the device. Maintenance parameters and service contracts belong to clinical engineering or healthcare technology management. Network location and patch status belong to IT. Software components come from the manufacturer. No single team holds all of them, which is why the inventory is a staffing question before it is a tooling question.

The playbook was candid that this was a goal. A footnote said: “This is considered a goal capability; many HDOs currently do not have the capability to catalog all their medical devices to this degree.” It listed the inability to conduct a complete medical device asset inventory as a potential cybersecurity risk in its own right, for inclusion in the hospital’s hazard vulnerability analysis.

Why the inventory decides the first hour

The playbook explained the purpose: an inventory that is current when an incident arises lets devices be quickly located and patched, pulled offline or replaced. Each of those actions depends on a field. Locating a vulnerable infusion pump model requires the model and the physical location. Pulling it offline requires the network location. Replacing it requires knowing whether spares exist and whether a neighbor has the same model. Deciding whether a patch is safe requires the manufacturer’s contact and the maintenance arrangement.

When a manufacturer or a federal agency announces a vulnerability in a device model, the first question at a hospital is whether it has that model, how many and where. A hospital with a current inventory can answer from its records. One without it may have to send staff to walk the floors. That difference is the most concrete measure of whether the inventory job is staffed.

Who should own it

The playbook did not assign the inventory to a department, and we would not either without knowing a hospital’s structure. We would insist on three things. One person is accountable for the inventory as a whole. Each field has a named source team responsible for keeping it current. And there is a reconciliation step, at a set interval, that compares what the inventory says with what the network sees, so that devices plugged in without being recorded are found before an incident rather than during one.

From inventory to hazard analysis

The playbook linked the inventory to a second preparedness task: the hazard vulnerability analysis, which hospitals use to identify gaps in emergency planning. It suggested including cybersecurity incidents and their potential impact on devices in that broader analysis, reviewing anticipated threats and existing mitigations to identify residual risks to accept, avoid or transfer.

Its list of potential risks reads as a list of staffing gaps: the inability to conduct a complete device inventory, the inability to collect and correlate audit logs across the enterprise, limited sensor coverage to detect adversary activity, a procurement process that does not address cybersecurity, and a lack of staff able to detect and respond to an incident. Its list of mitigations included tiering functions and assets by importance, reviewing and prioritizing remote connections that might need to be cut during a response, placing devices, especially legacy devices, on dedicated protected network segments, improving procurement, training users, and intrusion detection or security event management.

The playbook said the results of that assessment could identify the need for more mitigation, such as hiring skilled incident responders or training designated staff, and could inform the device cybersecurity portions of the hospital’s emergency operations plan. That is the bridge between the inventory and the people: the analysis shows which gaps need staff, and the emergency plan says who those staff are.

The command roles: authority delegated in advance

The playbook built on the hospital incident command system, which defines a hospital incident management team to lead the response to all-hazards incidents. Where an incident includes medical device cybersecurity concerns, it said, the activated team should include medical-technical specialists with cybersecurity and medical device expertise.

It then described additional roles. An information security officer, designated by the incident commander, would lead the cybersecurity portion of the team and act as liaison between the commander and cybersecurity staff. The chief medical information officer would be involved in IT decisions with a potential impact on patient care, such as taking part of the network offline or shutting off devices. Specialized technical experts in areas such as healthcare technology management, intrusion detection, malware analysis and forensics would augment the specialists, and the playbook noted that not all hospitals would have staff with those skills. A medical device cybersecurity liaison, ideally a healthcare technology management professional familiar with the affected devices, would coordinate with regional partners and the manufacturer. Other departments, including legal, risk management, finance, human resources and public affairs, would support the team.

The most important sentence in this section concerned authority. During the preparedness phase, the playbook said, a senior leadership champion such as the chief information officer may officially sanction, for example through policy, the cybersecurity decisions and actions the team takes during an incident, such as curtailing device usage. It explained why: during a cyberattack, there is not always time to make calls through a chain of command, so the hospital should clarify in advance who has what authority.

The decision that cannot wait

Curtailing device use is the decision that makes device incidents different from other IT incidents. Taking an email server offline inconveniences staff. Taking a fleet of infusion pumps or patient monitors out of service changes how nurses deliver care on every unit that uses them. The clinical risk of keeping a possibly compromised device in use has to be weighed against the clinical risk of working without it. That judgment needs a clinician and a security specialist in the same conversation, with someone authorized to decide.

If that authority is not settled beforehand, the decision waits for whoever can be reached. The playbook’s answer was to write it into policy, with a senior champion’s sanction, so that the incident team can act at three in the morning. We would add one thing: write down which decisions the team can make alone and which must go to the incident commander or a clinical executive, and rehearse the difference in an exercise.

Contact lists that work during an outage

The playbook asked hospitals to create a chart identifying all medical device cybersecurity roles, the people filling them and two methods of contact for each person. It also asked them to establish points of contact at each manufacturer, starting with whether the manufacturer had a product security page with incident reporting contacts. The emergency operations plan supplement it described repeated the requirement for at least two distinct methods of communication.

The two-method rule is easy to dismiss and important. A cyber incident can take down the systems people use to find each other: email, the paging platform, the directory. A contact chart that exists only on the network is unavailable in the incident it is meant for. Printing it, and testing that the second method works, is the kind of task that falls between departments unless someone owns it.

Defining when a device problem becomes an incident

The playbook asked hospitals to define a medical device cybersecurity incident in their emergency operations plan supplement, and it posed the questions that definition has to answer. The definition has to say when a device cybersecurity issue counts as an incident, which scenarios trigger the response, whether a vulnerability with an available patch is an incident, and whether and when alerts from external bodies, such as the regional coalition, federal agencies or the health sector information sharing center, establish incident status.

Those questions are about staffing as much as definitions. Every trigger that activates the incident team pulls people from their normal work. A definition that is too broad wears the team out on routine patch notices. One that is too narrow leaves a real event to be handled as routine maintenance. The playbook also suggested defining the triggers for activating device cybersecurity members of the team, how situational awareness is maintained, when to move to and from business continuity plans, and the internal and external communication requirements.

It added a practical note on cyber insurance: hospitals with a policy might want to be aware of its terms and have access to it during an incident. That is one more document that should be reachable when the network is not.

The region: capacity beyond the hospital’s walls

The playbook’s most distinctive idea was in its title. It argued that hospital preparedness for device cybersecurity incidents could be strengthened through regional outreach, by pooling limited resources and expertise before, during and after an incident. A region, it said, could be geographic, such as a state or a tri-state area, or organizational, such as hospitals in the same system. It was defined less by geography than by the incident response structure that best fit the participating hospitals.

It listed examples of regional partners: state and local health departments, state and local emergency response agencies, state or regional cybersecurity centers, regional health care coalitions, hospital trade associations, fusion centers, the local InfraGard chapter, information sharing organizations, regional testing laboratories and peer hospitals. Building trust relationships with those partners, it said, was the first step in device cybersecurity preparedness.

The preparedness activities it proposed included sharing practices, plans and policies; developing mutual aid agreements for device cybersecurity, or supplements to broader mutual aid agreements, to include loaner devices, diverting patients to a facility with operational devices and incident response assistance; exchanging points of contact; making sure key staff receive alerts through the regional health emergency communication system, such as a state health alert network; joint exercises; and identifying a primary and backup regional coordination center.

For response, it listed incident notification, informal sharing such as confirming whether others were seeing the same activity, formal sharing of indicators and lessons, backup communication means, requests for technical assistance and execution of mutual aid agreements. It acknowledged a barrier: hospitals may hesitate to share during an incident because of media concerns or nondisclosure agreements with manufacturers, and it suggested agreements among regional partners to protect sensitive information.

The rule that already pointed this way

The regional idea was not new to hospital emergency managers. The Medicare emergency preparedness condition of participation for hospitals, at 42 CFR 482.15, requires a plan based on a documented facility-based and community-based risk assessment, using an all-hazards approach. The plan must include a process for cooperation and collaboration with local, tribal, regional, state and federal emergency preparedness officials. The hospital’s policies must address arrangements with other hospitals and providers to receive patients if operations are limited or stop. The hospital must conduct exercises to test the emergency plan at least twice per year, including an annual community-based full-scale exercise or, where that is not accessible, an annual facility-based functional exercise. This summary follows the rule as the eCFR showed it on November 1, 2026.

The playbook cited the CMS emergency management rule and suggested including device cybersecurity incidents in the hospital’s hazard vulnerability analysis. Read together, the two documents give a hospital a way to staff the regional job with people it already has. The emergency manager who maintains patient transfer agreements and coalition relationships can add device incidents to them. The exercise the hospital must run anyway can include a device cybersecurity scenario. What the emergency manager needs from security and clinical engineering is the content: which devices, which scenarios, what loaner or diversion arrangements would make sense.

Loaner devices and diversion are logistics, not security

Mutual aid for devices is a supply chain problem. A neighbor can only lend a device it has spare, of a model the borrowing hospital’s staff can use safely, with configuration and connectivity that will not carry the same vulnerability. Diverting patients needs capacity at the receiving hospital and transport between them. Both depend on arrangements that take time to negotiate, and on knowledge the inventory job supplies. A hospital that cannot say which devices it has cannot ask a neighbor for the right ones.

Agreeing in advance what can be shared

The playbook’s communications section asked hospitals to specify incident sharing expectations for everyone in the response communications plan: what information can and cannot be shared, with whom and under what circumstances, by what mechanism, and when, including whether any circumstance would prevent sharing during an incident or any reporting timetable applies. It also asked whether a designated regional command center exists and how the hospital would take part.

Those decisions involve legal, privacy, public affairs and contracts staff as well as security. A hospital that has not made them will make them under pressure, and the default under pressure is to share nothing. The regional job therefore includes a prepared set of sharing rules, agreed with counsel, that the incident team can apply without waiting for a meeting. For health systems with several campuses, the same rules are needed internally, since the playbook allowed that a system’s own campuses could form its region.

The manufacturer: duties written at purchase

The playbook began its preparedness section with procurement. It said incorporating cybersecurity into device procurement could strengthen incident response, and it named four things a hospital might seek at purchase.

On costs, it suggested building into the purchase or maintenance fees the cost of mitigating device vulnerabilities, possibly including the availability of spare devices during an incident, because covering unforeseen costs during an incident slows the response. On exercises, it suggested securing a commitment from the manufacturer to participate in the hospital’s cybersecurity exercises, so that the two could define roles and practice sharing information about scope, impact, product-specific actions, patches and status. On components, it said requesting a software bill of materials would let the hospital identify and address vulnerable components and prioritize across its inventory. On access, it suggested arranging a cybersecurity preparedness user account that gives service-layer access during an incident.

Each of those is a contract term, and each is easiest to get before the purchase order is signed. After installation, the hospital’s bargaining power drops. The playbook also suggested that procurement might require the manufacturer to provide both a software bill of materials and a query capability to help maintain the inventory, which ties the manufacturer job back to the inventory job.

Who carries the requirement into the contract

Device purchasing in a hospital involves the clinical department that wants the device, the value analysis or supply chain team that negotiates, clinical engineering that will maintain it and IT that will connect it. The playbook’s procurement terms need someone in that chain whose job is to ask for them. Without that, each party assumes another raised security. Among the resources the playbook listed for a cybersecurity hazard analysis were the Manufacturer Disclosure Statement for Medical Device Security and the Department of Veterans Affairs’ pre-procurement assessment form for medical devices and systems.

We would make the requirement concrete. A standard device cybersecurity schedule, attached to every purchase of a connected device, covering the four items the playbook named. A rule that a connected device cannot be received until the inventory fields are filled in. And a list, maintained by supply chain, of which suppliers have agreed to exercise participation, so that the emergency manager knows whom to invite.

The manufacturer job continues after purchase. The playbook said that when a cybersecurity incident involving a device is identified, the hospital should initiate outreach first to the manufacturer. That instruction only works if the hospital knows whom to call and the manufacturer has agreed to answer. The contact established at purchase has to be kept current through staff changes on both sides, which is a small recurring task for whoever owns supplier relationships.

There is also a limit to what the manufacturer job can achieve. Some devices in service are no longer supported. The playbook’s inventory fields included whether a device is still supported by its manufacturer, and its mitigations included segmenting legacy devices that cannot be easily secured. For those devices, the hospital carries the response alone, and the other three jobs, inventory, command and region, carry more of the weight.

Training the people who notice first

The playbook gave training its own section, and its reasoning places the front line of detection with clinical staff. It said device users, from clinicians to IT help desk staff and healthcare technology management professionals, should be aware of potential device cybersecurity incidents, their impacts and appropriate responses. It said user awareness is particularly important in discovery, because device cybersecurity issues can first show up as unusual device behavior, and regular training would help ensure that cybersecurity is considered as a possible cause of any device peculiarity.

The point concerns the workforce, and it has a practical consequence. The nurse who notices that a pump or monitor is behaving oddly will report it as a malfunction to clinical engineering. Whether anyone then asks if the malfunction could be a cyber event depends on whether the report path includes that question. The playbook’s detection section made the same point from the other side: unlike natural disasters, cybersecurity incidents can be hard to identify and may masquerade as malfunctions or go unnoticed. It suggested validation questions, including how the potential incident arose and whether regional partners had seen anything similar.

The playbook asked hospitals to identify device cybersecurity points of contact, familiarize users with the incident classification and prioritization system, and fold awareness training into broader emergency preparedness or device user training. That last suggestion matters for staffing: it adds a module to training staff already receive, instead of creating a new course competing for time on the units.

On exercises, the playbook said hospitals already run all-hazards preparedness exercises, and that cybersecurity could be integrated into them or exercised separately. It asked for participation from across the hospital, including the emergency response organization, the healthcare technology management and IT departments, manufacturers and maintenance contractors. After an exercise, it said, the hospital should update its plans with lessons learned, improve communication channels between units, define policy and process, create new groups if necessary, obtain buy-in from senior leadership and affected units, and identify the individuals who will take part in incident response. That list is, in effect, a staffing plan produced by rehearsal.

What changes for each department

Seen as staffing, the playbook asked something specific of each department that touches devices. The playbook does not set out these assignments in this form; they are our reading of where its tasks land.

Clinical engineering, or healthcare technology management, becomes the custodian of the device side of the inventory and supplies the device liaison during an incident. Its staff need enough security knowledge to recognize when a device fault may be a cyber event and to talk with manufacturers’ security teams.

IT security supplies the information security lead for the incident team, keeps the network fields of the inventory current and runs the reconciliation against what the network sees. Its staff need enough clinical context to understand that isolating a device fleet has consequences on the units.

Emergency management folds device incidents into the hazard vulnerability analysis, the emergency operations plan, the coalition relationships, transfer agreements and the exercises the hospital already runs. It does not need deep technical knowledge, but it needs a standing relationship with the other two.

Clinical leadership, through the chief medical information officer or nursing leadership, owns the patient-care side of the decision to curtail device use, and makes sure frontline staff know how to report odd device behavior and how to work without a device if it is pulled.

Supply chain carries the cybersecurity schedule into each purchase and keeps the list of supplier commitments. Senior leadership signs the policy that delegates authority and decides which gaps the hazard analysis identifies are worth funding.

Objections: optional guidance, small hospitals, rare events

The first objection is that the playbook was not guidance and created no obligation. The playbook said so plainly. Its weight came from two other places. It built on the emergency preparedness rule that participating hospitals must already meet, and it recorded what the FDA had heard from its stakeholders after real attacks. A hospital that ignored it was not breaking a rule. It was declining a ready-made way to meet one.

The second objection is that small and rural hospitals cannot staff four new jobs. The playbook anticipated this. It said it presented target capabilities that not every hospital would be able to execute fully, that smaller, less-resourced hospitals may benefit more from the deeper bench regional collaboration offers, and that collaborating with regional peers or outsourcing might be needed for specialized skills. The regional job is, in part, the answer to the staffing problem in the other three.

The third objection is that device cyber incidents affecting patients were rare. The FDA’s statement said the agency was not aware of any reports of an unauthorized user exploiting a vulnerability in a device in use by a patient. But the playbook’s scope included incidents that affect device function at scale, and its background cited WannaCry, which it said showed how vulnerable devices may be compromised whether or not an attack targets health care. The FDA’s own statement made the same point: even when devices are not deliberately targeted, those connected to a hospital network may be affected. The playbook’s own risk mitigations included placing devices, especially legacy devices that cannot easily be secured, on dedicated and protected network segments.

A fourth objection is that the inventory alone would be enough. It would solve the first hour. It would not decide who may pull devices from service, find loaners, or get a manufacturer to answer. The four jobs depend on each other, but none substitutes for another.

A fifth objection is that the regional job asks competing hospitals to cooperate, share incident details and lend each other equipment, and that competition and liability concerns will stop them. The playbook acknowledged the hesitation, citing concerns about media attention and nondisclosure agreements with manufacturers, and suggested agreements among regional partners and an information sharing organization as an initial conduit. Hospitals already cooperate on patient transfers and surge capacity through coalitions and the arrangements the emergency preparedness rule requires. Device incidents can be added to those arrangements without building trust from nothing.

A sixth objection is that clinical engineering and IT security belong to different reporting lines with different priorities, and that a playbook cannot make them work together. We think it is the strongest objection, and it is why we read the playbook as a staffing document. The playbook could describe the roles. Only hospital leadership can assign them, fund them and require the two departments to share an inventory and an exercise calendar.

Staffing the plan: who does what on an ordinary Monday

The playbook was written for the incident. Its demands are met, or not, on ordinary days. A hospital that wants to know whether it could follow the playbook can check each job against what someone does in a normal week.

  • Inventory: someone reconciles the device inventory against the network at a set interval, and each field has a source team that updates it when devices are added, moved, patched or retired.
  • Command roles: the emergency operations plan names the medical-technical specialists, the information security lead and the device liaison for a device incident, with two contact methods each, and a policy signed by a senior leader delegates the authority to curtail device use.
  • Region: the emergency manager’s coalition and transfer agreements mention device incidents, at least one exercise each year includes a device cybersecurity scenario, and the hospital knows which neighbors use the same critical device models.
  • Manufacturer: every new connected device purchase includes the hospital’s cybersecurity schedule, and supply chain keeps a list of manufacturers that have agreed to exercise participation, provided a software bill of materials and named an incident contact.

Each of those is a routine task with an owner. None requires a new department. The playbook’s contribution was to name the tasks and show where they fit in structures hospitals already run: incident command, emergency planning, coalitions and purchasing. In our view, a hospital that staffs those four tasks has done the core of what the playbook asked, whatever tools it uses.

With hindsight

The FDA’s device cybersecurity work continued along the premarket track that the October 2018 statement previewed. According to the FDA’s cybersecurity page, content current as of November 10, 2026, the agency issued final guidance on cybersecurity in medical devices and premarket submissions on September 27, 2023. It issued draft guidance on select updates for section 524B of the Federal Food, Drug, and Cosmetic Act, which concerns cyber devices, on March 13, 2024. On June 27, 2025, it issued final guidance that added a section on its recommendations under section 524B and superseded the 2023 guidance. On February 2, 2026, it issued a revised final guidance aligned with the amendments to 21 CFR 820, the quality management system regulation. The same page lists FDA-contracted MITRE work on managing legacy device cybersecurity risks, dated November 15, 2023, and a white paper on software bill of materials data normalization, dated October 24, 2024.

The playbook itself was updated. The later edition, prepared by MITRE under contract with the FDA, states that the first version was published in 2018 and that cyber incidents in the sector had grown since. It keeps the regional structure, adds partners such as regional health information exchanges, ASPR regional emergency coordinators and regional offices of the Cybersecurity and Infrastructure Security Agency (CISA), and asks hospitals to consider the clinical and operational impact of extended device downtimes, dependencies between devices and IT systems, and cloud dependencies. The FDA page also lists a video for health care facilities on cybersecurity incident preparedness and response, launched May 1, 2023. The updated playbook also notes that, as of its publication, reporting to CISA was voluntary, and that the Cyber Incident Reporting for Critical Infrastructure Act of 2022 requires covered entities to report covered cyber incidents and ransomware payments to CISA once CISA completes its rulemaking.

For hospital operations leaders, the record since 2018 splits along the same line as the playbook. The manufacturer side gained detailed premarket guidance, including the FDA’s recommendations under section 524B for cyber devices. The hospital side, the inventory, delegated authority, regional arrangements and purchase terms, stayed a matter of each hospital’s own staffing and planning, guided by an updated playbook that remained, like the first, a resource rather than a requirement.

The inventory job got a concrete test on January 30, 2025, when the FDA issued a safety communication about Contec CMS8000 patient monitors and Epsimed MN-120 monitors, which it said were relabeled CMS8000s. The FDA said the monitors’ software included a backdoor, that once connected to the internet they could send patient data outside the care environment, and that no software patch was available. It told facilities to use only local monitoring features, to review CISA’s mitigations for the related vulnerabilities, to check for unusual functioning such as inconsistencies between displayed vital signs and the patient’s actual state, to unplug the network cable and disable wireless where the device did not rely on remote monitoring, and to stop using the device where it did. Following that advice required knowing which monitors a hospital had, where and how each was connected. The FDA said it was not aware of any cybersecurity incidents, injuries or deaths related to those vulnerabilities at that time.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • medical devices
  • incident response
  • FDA
  • emergency preparedness
  • clinical engineering
  • SBOM