HICP sized health organizations by who could do the work

HHS’s December 2018 cybersecurity practices sorted health organizations into tiers defined largely by IT and security staffing. That made staffing the first decision in any HICP program, and later federal goals were written in the middle tier’s terms.

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

HHS’s December 2018 cybersecurity practices sorted health organizations into tiers defined largely by IT and security staffing. That made staffing the first decision in any HICP program, and later federal goals were written in the middle tier’s terms.

  1. The week At the end of December 2018, the Department of Health and Human Services published Health Industry Cybersecurity Practices: Managing Threats and Protecting Patients, known as HICP.
  2. How the publication was built The task group worked from threats to practices.
  3. The small tier: security through someone else Technical Volume 1 said that small organizations generally do not have dedicated IT and security staff to implement cybersecurity practices, because of limited resources.
  4. The medium tier: IT staff carrying security The best-fit table described medium organizations as having dedicated IT resources on staff but no or limited dedicated security resources.
  5. The large tier: specialists and coordination Large organizations, in the best-fit table, had a security leader with dedicated staff, a dedicated budget with a strategic roadmap, and a significant number of exchange partners or partners with less rigorous standards.
  6. The connections between tiers The main document warned that choosing a size tier was harder than it looked.
  7. The strongest objections One objection is that the tiers were plainly size categories, with bed counts and physician counts, and that reading them as staffing levels overstates one row of a table.
  8. Turning the tiers into staffing decisions The first decision is which volume to follow.
  9. With hindsight HICP lasted, and its tiers lasted with it.

The week

At the end of December 2018, the Department of Health and Human Services published Health Industry Cybersecurity Practices: Managing Threats and Protecting Patients, known as HICP. The page that presented it on the public health emergency website of HHS’s Office of the Assistant Secretary for Preparedness and Response was last reviewed on December 28, 2018. An Internet Archive capture of the main document is dated the same day.

The publication answered a mandate in the Cybersecurity Act of 2015. Section 405(d) of that law, titled Aligning Health Care Industry Security Approaches, called for common voluntary guidelines for the health sector. HHS convened a task group in May 2017 through the sector’s critical infrastructure partnership, working with the Health Sector Coordinating Council. HHS said the group drew on more than 150 healthcare and cybersecurity experts, with roles ranging from chief executives and security officers to doctors and nurses.

The group decided it could not address every cybersecurity problem in the industry in one document. It focused instead on five threats: e-mail phishing attacks, ransomware attacks, loss or theft of equipment or data, insider accidental or intentional data loss, and attacks against connected medical devices that may affect patient safety. Against them it set ten practices: e-mail protection, endpoint protection, access management, data protection and loss prevention, asset management, network management, vulnerability management, incident response, medical device security and cybersecurity policies.

The publication had four parts. A main document set out the threats and a call to action aimed at executives. Technical Volume 1 covered the ten practices for small organizations. Technical Volume 2 covered them for medium-sized and large organizations. A resources and templates volume supplemented them. Across the volumes, the ten practices broke down into 88 sub-practices.

The defining design choice was size. The co-leads’ foreword said the task group had determined it was critical to tailor practices to an organization’s size: small, medium-sized or large. A table titled Selecting the Best Fit for Your Organization helped readers choose. It used provider counts and hospital beds, but also staffing, budget and connections. A small organization had no dedicated IT professionals on staff, with IT possibly outsourced on a break-fix or project basis. A medium organization had dedicated IT staff but no or limited dedicated security staff. A large organization had a chief information security officer or dedicated security leader with security staff.

The publication was explicit about its status. Its disclaimer said use of the document was neither required by nor a guarantee of compliance with federal, state or local law. The foreword said the authors did not expect the practices to become a de facto set of requirements. It also aligned the practices with the NIST Cybersecurity Framework rather than creating a new framework.

The main document addressed executives and clinicians as well as technical staff. It compared cyber hygiene to hand washing before patient care. It said that, like hand washing, a culture of cyber awareness did not have to be complicated or expensive for a small organization. It cited a 2016 ransomware attack in which a private hospital reverted to pen and paper and transferred some patients.

HICP is usually described as a list: five threats and ten practices. That description is accurate and misses its most consequential choice. The task group sorted organizations into three tiers, and the tiers were defined as much by staffing as by size. A small organization was one without its own IT professionals. A medium organization had IT staff but little or no security staff. A large one had a security leader and a security team.

Our argument is that HICP sized health organizations by who could do the work, and that this made staffing the first decision in any HICP program. The practices for each tier were written for the people that tier was assumed to have. An organization that read the right volume for its bed count, but did not have the people that volume assumed, held a plan it could not carry out. Later federal baselines, as the hindsight section shows, were written in the vocabulary of the middle tier.

We work through four parts, after a short account of how the publication was built: the small tier, where security work was done through someone else; the medium tier, where IT staff carried security alongside everything else; the large tier, where security staff existed but faced a coordination burden; and the connections between tiers, which the publication itself said could make a small organization’s risk larger than its size. A sixth section tests the argument against the strongest objections. The last turns it into staffing decisions.

The sector is healthcare because HICP was written for the Healthcare and Public Health sector. It was voluntary guidance. Covered entities and business associates have separate obligations under the HIPAA Security Rule, which they should confirm with counsel.

How the publication was built

Threats first

The task group worked from threats to practices. Its stated approach was to examine current threats to the sector, identify the weaknesses that make organizations vulnerable to them, and select the practices experts ranked most effective. The main document then translated each threat into vulnerabilities, impact and practices, with a table of suggested practices for each threat.

That structure suited its audience. The main document spoke to executives and clinicians, using medical comparisons such as hand hygiene. The technical volumes spoke to IT and security professionals. A practice manager could read the threats and understand why. An IT provider could read the volume and understand what.

Sized by attributes

The best-fit table listed attributes in several groups. Under common attributes, it compared health information exchange partners, IT capability and cybersecurity investment. Under provider attributes, it compared physician counts, provider counts for acute and post-acute care, hospital beds and complexity. It also listed other organization types, such as practice management organizations, health plans and device manufacturers.

The provider numbers set rough boundaries. A small organization had one to ten physicians, or one to 25 providers in acute or post-acute care, or one to 50 hospital beds. A medium organization had 11 to 50 physicians, 26 to 500 providers or 51 to 299 beds. A large one had more than 50 physicians, more than 500 providers or more than 300 beds. The table presented these as guides to best fit, not rules.

How the volumes used the tiers

Technical Volume 1 presented one set of sub-practices for small organizations. Technical Volume 2 presented each practice twice, once for medium and once for large organizations. Medium organizations were advised to start with their own sub-practices. Large organizations were advised to review both sets. Medium organizations were encouraged to adopt large-tier sub-practices where they fit.

The publication also offered a way to prioritize. An assessment method and toolkit in its appendices was meant to help organizations decide which sub-practices mattered most, based on the threats that concerned them. The main document said not all sub-practices would be effective for every organization.

What the publication said it was not

The authors drew careful limits. The foreword said a dogmatic approach would not work, given how fast threats and technology change. It said the authors did not guarantee that the practices would help organizations meet compliance or reporting obligations. It said the document did not create new frameworks or rewrite specifications, and instead introduced the NIST Cybersecurity Framework’s terms to health sector professionals.

Those limits matter for staffing. A framework that tells an organization what to achieve leaves it free to decide who does the work. HICP went further than that. By writing separate volumes for organizations with different staff, it told each organization what its people could reasonably be expected to do. This was a more useful kind of guidance, and it carries an assumption about headcount in every line.

What the tiers really measured

Read closely, the tiers measured three things that go together: who was on staff, what money was set aside, and how many outside parties the organization exchanged data with. Bed counts and physician counts were proxies for those. A 40-bed hospital with an IT director and a security analyst would fit the medium tier on staffing while fitting the small tier on beds. The table allowed for that by asking for the best fit across attributes.

That is the mechanism behind our argument. Every sub-practice assumes someone to perform it. When the tiers are read as staffing levels, the volumes become a description of what each staffing level can be expected to do.

The small tier: security through someone else

What the small volume assumed

Technical Volume 1 said that small organizations generally do not have dedicated IT and security staff to implement cybersecurity practices, because of limited resources. It said many small practices use third-party IT support and cloud service providers. It said those providers should be capable of reviewing the publication’s practices to decide which apply. The main document described the volume as intended for IT and IT security professionals, and as a guide to what a small organization should ask its IT team or vendors.

The sub-practices followed that assumption. The e-mail section noted that most small practices use outsourced e-mail providers and told them to check with the provider that controls were in place. The network section said that even when a third-party IT vendor managed the network, the organization had to understand key aspects of network management and make sure they were in its contracts. Vulnerability scanning was described as a capability often provided by an electronic health record or IT support vendor.

What the small organization kept

Some work could not be handed to a vendor. The small volume’s data protection practice called for policies, procedures and education. Its asset management practice asked the organization to keep an inventory and to assign responsibility for recording new assets to the person who buys them. Its decommissioning practice told small organizations to record each device’s disposal, and to record the certification of destruction when a service provider did the work.

Incident response was the clearest case. The volume said small organizations were often challenged by incident response, partly because procedures might not exist. It said employees who rarely see attacks might not remember what to do, and managers might not know whom to contact. It said that, in many cases, small organizations had no dedicated security professionals, which increased reliance on the IT department. It advised deciding, before any incident, who would lead the investigation and making sure that person had full authority.

Training as the small tier’s main control

Much of the small volume relied on the workforce itself. The e-mail practice asked small organizations to keep a training program with a section on phishing, so that all users could recognize common techniques. It listed them: checking embedded links, looking for spoofed sender addresses and being cautious with urgent requests. Its phishing simulation sub-practice recommended regular campaigns, monthly or quarterly, with real-time training.

It also told small organizations to direct their IT specialist to send a test phishing e-mail to all staff and track who opened it. That instruction assumed the organization had an IT specialist to direct. HHS’s later how-to guide for small organizations gave the same instruction to the appointed IT personnel or an administrative specialist. The shift is small and telling. In the smallest practices, the person running security awareness may not be in IT at all.

Contracts in place of staff

Where a small organization had no staff to perform a practice, the volume often turned it into a contract term. It said contracts with electronic health record vendors should require medical data to be encrypted at rest and in transit. It said cloud backup was acceptable if the organization had an agreement with the cloud vendor and had verified the security of the vendor’s systems. It said network management requirements should be in the contracts for those services.

Each of those instructions moves work from the organization’s staff to its vendor, and moves a different kind of work back. Someone has to write the term, check that it is met and act when it is not. Contract management is a staffing task. In the small tier, it falls to whoever handles purchasing, often without technical support.

Where the small volume pointed elsewhere

On one practice, the small volume did not try at all. Its medical device sub-practice said that an organization connecting medical devices to a network should consider the device practice in Technical Volume 2. In effect, a small organization with networked devices was told to use the guidance written for organizations with IT staff. The volume was admitting that some risks do not scale down with the organization.

The staffing gap in the small tier

Put together, the small volume asked a small organization to do three kinds of work. It had to direct a vendor. It had to run a few processes itself, such as inventory, policy and training. And it had to lead in a crisis. None of those requires a security engineer. All of them require a named person with time and authority.

The volume did not say who that person should be. In a two-physician practice, the likely candidates are the practice manager or one of the physicians. Each already has a full job. The practices were achievable. The time to perform them was not budgeted anywhere in the document, because the document could not budget it.

The incident nobody leads

The small volume’s incident response advice deserves a second look. It said a common concern in small organizations was fear of regulatory penalties for acknowledging and fixing a security incident. That fear affects staffing as much as policy. A practice manager who is unsure whether to call anyone may wait, and in a ransomware attack delay can be costly.

The volume’s answer was to decide in advance who leads and to give that person full authority. For a small organization, the lead does not need to be technical. The lead needs to know the plan, the vendor contact, the outside help to call and the authority to spend money and shut systems down. A sample incident response plan in the main document’s appendices gave a starting point.

What it takes to run the small tier

A small organization that wanted to follow Volume 1 would need, at minimum:

  • a named internal owner for security, with a few hours a week set aside and authority to direct the IT vendor;
  • a contract with the IT vendor that names the HICP practices the vendor performs and how it reports on them;
  • a short list of internal tasks the owner performs, such as inventory updates, access reviews when staff leave and the annual training cycle;
  • an incident plan that names the leader, the vendor contact and the outside help the organization would call.

That list is modest. Its cost is mostly time, and time in a small practice is clinical time or management time diverted. The tier was defined by the absence of staff, and the work still needed someone.

The medium tier: IT staff carrying security

What the medium tier assumed

The best-fit table described medium organizations as having dedicated IT resources on staff but no or limited dedicated security resources. Funding was allocated for specific initiatives, future funding was potentially limited, and cybersecurity and IT budgets were blended. Medium organizations had several exchange partners and could have multiple sites across a wide area.

The medium sub-practices were substantially more demanding than the small ones. They included multifactor authentication for remote e-mail access, e-mail encryption, identity and provisioning procedures, classification of data, backup strategies and data loss prevention. Network management included firewalls, segmentation, intrusion prevention and web proxy protection. Vulnerability management included host and server scanning, web application scanning and patch, configuration and change management. Incident response included a security operations center and information sharing.

Why this tier carries the most strain

In our view, the medium tier is where the gap between practices and people is widest. The sub-practices assume work that security specialists usually do, such as running a security operations center or tuning data loss prevention. The staffing assumption is that IT staff do it, alongside keeping the network, the clinical applications and the help desk running. The blended budget means every security project competes with every IT project.

A small organization has an obvious answer to its gap: buy the service. A large organization has security staff. A medium organization often has enough IT staff to believe it can do the work, and not enough to do it while also running operations. Good plans stall there.

The budget problem

The best-fit table described the medium tier’s money as funding for specific initiatives, with potentially limited future allocations, and with security and IT budgets blended. That description points at a specific failure. A project can be funded once: a firewall replacement, a data loss prevention rollout or a scanning tool. The staff time to run it afterward is an operating cost, and in a blended budget it competes with every other operating need.

Many of the medium sub-practices are ongoing activities rather than projects. Patch and configuration management, scanning, monitoring and access reviews never finish. An organization that funds them as projects buys the tool and not the hours. The tier’s own funding description predicts that outcome.

Choosing what to staff and what to buy

A medium organization following Volume 2 has to decide, sub-practice by sub-practice, which work its own staff will do and which it will buy. Some work suits internal staff because it depends on knowing the organization: identity and provisioning, data classification and asset inventory. Some suits a service because it needs round-the-clock coverage or specialist skills: monitoring and the security operations center function, and some scanning.

The decision has consequences beyond cost. Work that is bought still needs an internal owner who reads the reports and acts on them. A security operations center service that sends alerts to a mailbox nobody watches out of hours provides little. The medium tier’s real staffing need is a security lead, even a part-time one, whose job is to own that boundary.

Information sharing as a force multiplier

One medium sub-practice offered help from outside: information sharing through information sharing and analysis centers and organizations. Even the small volume included participation in such groups. For a medium organization with thin security staffing, sharing extends what a small team can know. It gives early warning of campaigns and practical advice from peers. It also takes time to read, filter and act on what is shared, which brings the question back to who has that time.

Identity work belongs inside

Of all the medium sub-practices, the identity group most needs internal staff. Identity, provisioning, transfers and deprovisioning procedures, authentication and multifactor authentication for remote access all depend on knowing who works for the organization today, in what role, and who left yesterday. That knowledge sits in human resources, medical staff offices and department managers. An outside provider can run the tools. Only internal staff can keep the identity records true.

For that reason, a medium organization that can afford only a small security staff should consider putting it first on identity and access, and buying monitoring as a service. The reverse arrangement, internal monitoring and outsourced identity, leaves the outsider guessing who should have access.

Signs the medium tier is overstretched

Leaders can watch a few indicators. They include the age of the oldest unapplied critical patch, the share of departed staff whose access was removed on time, the time between a security alert and a human review, and the number of security projects started but not finished. Each measures whether the people assumed by the tier are actually doing the work. Rising numbers mean the plan has outrun the staff.

The large tier: specialists and coordination

What the large tier assumed

Large organizations, in the best-fit table, had a security leader with dedicated staff, a dedicated budget with a strategic roadmap, and a significant number of exchange partners or partners with less rigorous standards. They included integrated delivery networks and participants in accountable care organizations or clinically integrated networks.

The large sub-practices assumed specialists. They included advanced e-mail tooling and analytics-driven education, endpoint detection and response, application whitelisting and micro-segmentation, federated identity and access governance, mapping of data flows, network access control and anomaly analytics. Incident response included advanced security operations centers, orchestration, user behavior analytics and deception technologies. Medical device security included procurement and security evaluations and contacting the FDA.

Where the large tier’s difficulty lies

For a large organization, the staffing question is less about headcount than about reach. Security staff exist, but the systems they protect are spread across many hospitals, clinics, departments and vendors. Many decisions that affect security are made by people outside the security team: clinical engineering for medical devices, procurement for new systems, and department leaders for new applications.

The large sub-practices recognize that. Procurement and security evaluation for medical devices, mapping data flows and access governance all depend on other teams doing their part. The security team’s job in this tier is as much coordination as operation.

Medical devices as a staffing seam

Medical devices show the coordination problem most clearly. The fifth threat in the main document was attacks against connected medical devices that may affect patient safety. The large sub-practices for devices included vulnerability management, security operations and incident response, procurement and security evaluations, and contacting the FDA.

Each of those crosses team lines. Devices are often bought by clinical departments, maintained by clinical engineering and supported by manufacturers. The small volume itself noted that many devices are managed remotely by third-party vendors, which increases the attack footprint. A security team cannot run device vulnerability management without the people who know which devices are where and who may change them. The staffing need is a working link, with named people on both sides.

What a large organization needs to staff

A large organization following Volume 2 needs specialist roles for the advanced sub-practices. It also needs people whose job is to connect security to the rest of the system. That includes security representation in procurement decisions, a working relationship with clinical engineering on devices, and a way to bring department leaders into risk decisions about their own systems. Without those links, the specialists defend what they can see, and the rest of the organization adds what they cannot.

The connections between tiers

What the publication said about coupling

The main document warned that choosing a size tier was harder than it looked. It said a small practice with one or two providers and no affiliations or exchanges was not as common as it used to be. It said even the smallest organizations might be tightly coupled with others, sharing information about common patients, establishing health exchanges and affiliating with larger health systems.

The best-fit table built that into its attributes. Exchange partners were the first common attribute listed. The large tier included partners with less rigorous standards or requirements.

Exchange as an attribute of size

Using exchange partners as a size attribute was an unusual choice. Most size measures count what an organization has. This one counted whom it connected to. A practice with one or two exchange partners was small on that measure. A system with partners holding less rigorous standards was large, partly because of the weakness of those partners. The measure treated connections as a source of complexity and risk, which is how attackers treat them too.

Why coupling matters for staffing

Coupling moves risk across tiers. A small practice connected to a large health system’s record system is defended to the small tier’s standard, staffed by the small tier’s people, and connected to the large tier’s data. A large system’s security staff may never see the practice’s systems. The practice’s IT vendor may never see the system’s security requirements.

The large tier’s mention of partners with less rigorous standards points at exactly this. A large organization’s exposure depends partly on the staffing of its smallest partners. Its own security team cannot fix that by working harder inside its own network.

Organizations that are not providers

The best-fit table also placed organizations that are not care providers. It put practice management organizations, managed service organizations, smaller device makers, smaller drug companies and smaller payers in the small and medium tiers. It put health plans, large device makers and large pharmaceutical companies in the large tier.

Several of those organizations supply services to providers. A managed service organization that runs systems for many practices sits at a junction of many couplings. Its staffing determines the security of every practice it serves. The best-fit table placed it by its own size, but its effect reaches every client it supports.

What organizations on each side can do

The larger party can set connection standards and help partners meet them. It can require multifactor authentication for partner access, limit what each partner connection can reach, and monitor partner activity. It can offer templates, training or shared services to small partners, subject to the legal limits on what it may give them, which it should check with counsel.

The smaller party can treat each connection as a security decision. It can ask the larger partner what it requires, ask its IT vendor to meet those requirements, and name the person who manages the relationship. A connection that nobody owns is the most likely place for a gap between tiers.

Both sides should write the arrangement down: what each connection reaches, who approves changes to it, who is called when something looks wrong, and how either side can shut it off quickly.

The strongest objections

The tiers were about size, not staff

One objection is that the tiers were plainly size categories, with bed counts and physician counts, and that reading them as staffing levels overstates one row of a table. The table, though, put IT capability and cybersecurity investment in its first group of attributes, ahead of provider counts. The small volume began by saying small organizations generally lack dedicated IT and security staff. The 2023 edition, as the hindsight section shows, made the staffing link more explicit still.

Outsourcing solves the staffing problem

A second objection is that an organization without staff can buy everything. HICP itself suggested vendors could review and implement practices. Outsourcing does solve the skills problem for much of the work. It does not remove the need for a person inside the organization who directs the vendor, reads its reports, owns the contract and leads in a crisis. The small volume’s own incident response advice assumed that person existed.

Voluntary guidance does not drive staffing

A third objection is that HICP was voluntary, so its assumptions did not bind anyone. The point holds for the 2018 document. It became less right over time. A 2021 amendment to the HITECH Act named the approaches promulgated under section 405(d) among the recognized security practices HHS must consider in enforcement, and HHS’s later performance goals linked back to HICP sub-practices. The tier assumptions traveled with the practices.

The threats changed, so the tiers are dated

A fourth objection is that the 2018 threats and practices were overtaken, so the tier structure belongs to history. The 2023 edition did change the threats and practices. It kept the three tiers, the best-fit table and the split between the two technical volumes. The staffing assumptions survived the update, and were stated more directly.

Where the argument is weaker

The argument is weaker for organizations that sit inside larger systems and receive security services from them. A clinic owned by a health system may be small in size and large in security staffing, because the parent’s team covers it. For that clinic, the best-fit answer is the parent’s tier, and the staffing question becomes how far the parent’s team actually reaches.

Turning the tiers into staffing decisions

Pick the tier by staff, then check by size

The first decision is which volume to follow. An organization should start with the staffing and budget rows of the best-fit table, not the bed count. If those rows say small, Volume 1 is the realistic plan, whatever the size. If they say medium and the size says large, the medium sub-practices are the floor and the gap to the large tier is a staffing plan.

Name the owners before the controls

The second decision is to assign each of the ten practices to a named owner, internal or contracted, before choosing tools. An owner who is contracted still needs an internal counterpart. A practice with no owner is not in place, whatever the policy says.

Budget the time as well as the tools

The third decision is to budget hours. For a small organization, that means protected time for the internal security owner. For a medium organization, it means separating security time from IT operations time, so that security work is not the first thing dropped in a busy month. For a large organization, it means funding the coordination roles that connect security to procurement, clinical engineering and departments.

Write the roles from the volumes

The technical volumes can double as a source for role descriptions. For a small organization, the security owner’s role can be written from Volume 1’s internal tasks: inventory, policy, training, access removal and incident leadership. For a medium organization, the security lead’s role can be written from the medium sub-practices it keeps in house, with the rest listed as vendor oversight. For a large organization, the coordination roles can be written from the sub-practices that cross teams, such as device procurement, data flow mapping and access governance.

Roles written this way have two benefits. They make the hours visible, because each task can be estimated. And they survive staff turnover, because the next person inherits a defined job rather than a predecessor’s habits. A role description also gives leadership something concrete to fund, which a list of sub-practices rarely does.

Measure the staffing alongside the controls

Organizations usually report security as a set of controls in place. A staffing view adds different measures: hours spent on security work in the period by the people assigned to it, the share of the ten practices with a named owner, the number of practices where the owner is a vendor without an internal counterpart, and turnover in security-related roles. Those numbers show whether the tier the organization chose is one its people can sustain.

Revisit when the organization changes

Tiers change when organizations grow, affiliate or lose staff. An acquisition, a new exchange partner or the departure of the one person who handled security can move an organization between tiers in a month. The best-fit decision should be revisited whenever one of those things happens, and at least each year.

With hindsight

HICP lasted, and its tiers lasted with it. In January 2021, Public Law 116-321 amended the HITECH Act. It required HHS, when deciding fines, audits and other remedies under the HIPAA Security Rule, to consider whether an entity had recognized security practices in place for at least the previous 12 months. The law defined recognized security practices to include the approaches promulgated under section 405(d) of the Cybersecurity Act of 2015. It said the entity would determine its own practices, consistent with the Security Rule, and that it created no liability for choosing not to adopt them. It also said HHS gained no authority to increase fines or audits because an entity lacked such practices. The law thus rewarded the practices without requiring them. For an organization deciding how to staff, it made evidence of practice over the prior year worth something in enforcement, and that evidence had to be produced by whoever did the work.

The task group began updating the publication in 2019, and HHS issued a 2023 edition. It kept the five-threat, ten-practice structure, the three tiers and the two technical volumes. It renamed the first threat social engineering, rewrote the medical device practice, and turned the policies practice into cybersecurity oversight and governance. It added sub-practices on attack simulations, cyber insurance and risk assessment. It also added a section on workforce training and the role each person plays.

The 2023 edition made the staffing link explicit. It said the technical volumes were intended for IT staff, cybersecurity staff or managed service providers. It told small organizations with a provider to hand over the main document and Volume 1, and those without one to designate an individual to assess current practices and build a plan. Its best-fit table described a medium organization’s IT as on staff, co-managed with outsourcing or fully outsourced. It told readers that an organization might have elements of several sizes, and should pick the one that best represents it. HHS also published how-to guides, including one on cyber workforce training for small organizations. As checked on November 14, 2026, HHS’s cyber gateway site still offered the main document and both technical volumes.

HHS then published voluntary healthcare Cybersecurity Performance Goals, divided into essential and enhanced goals. The page describing them said they were built on CISA’s goals and informed by HICP and the NIST Cybersecurity Framework, and that they addressed common attack vectors against hospitals identified in a 2023 analysis of hospital cyber resiliency. As checked on November 14, 2026, the HHS page that lists them linked each goal to HICP sub-practices. Every essential goal linked to medium-tier sub-practices, and the enhanced goals to medium- and large-tier ones. None linked to a small-tier sub-practice.

What held up was the tiering itself, and the honesty of the small volume about who would do the work. What changed was the floor. The voluntary baseline HHS later published for healthcare organizations was written in the language of the medium tier, which assumes IT staff. For organizations staffed like HICP’s small tier, meeting that baseline was a staffing question before it was a technical one.

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

  • HICP
  • 405(d)
  • cybersecurity staffing
  • small practices
  • HHS