Cybersecurity: Its clinical, too
Imagine this text message shows up on your smartphone: “Dear John, we have complete control of your pacemaker software. Please remit $10,000 to our offshore bank immediately or we will drain your battery.”
Some might say this is perhaps a bit far-fetched. But many in health care now must consider possibilities along these lines every day. With ransomware attacks such as WannaCry — which hit Britain's National Health Service — frequently in the news, and with hospitals' clinical technology becoming more and more connected to information technology systems, the cybersecurity risks associated with medical devices are constantly growing. So, while connected medical devices provide numerous ways to better coordinate patient care, with all of its benefits, those connections simultaneously expose us to new risks, which we now must manage.
With increased exposure, experts worry a hacker could connect remotely to a hospital network and the medical devices connected to it. We must all better understand these risks so we can reduce their possible effect on our patients.
In many ways, this risk has been building over the past 20 years as medical devices of all different categories have incorporated software and software systems into their design. More recently, though, the risks have escalated as more devices connect to hospital networks, outpatient networks and even home networks. More technology means more risk. And more connected technology means more connected risk.
In today’s hospital, according to our research, there are approximately 15–17 devices per bed, and about 1 out of 4 of those bedside devices are networked. With more medical devices like physiologic monitoring systems connecting via hospital networks to electronic health records and other information systems, medical device cybersecurity vulnerabilities extend beyond the patient bedside. These connected devices could serve as entry points into a hospital’s network, placing hospital operations, medical information, patient identity and patient financial information at risk. Older devices can bring more risk: Many medical devices have long lifespans and, consequently, these systems may have older operating systems that are more vulnerable.
We should also note that while the Food and Drug Administration is acutely aware of medical device cybersecurity and has issued guidelines and recommendations, it does not test devices for these vulnerabilities. The FDA describes the perception that it does so as a “myth” in a fact sheet it published. “The FDA does not conduct premarket testing for medical products," the sheet reads. "Testing is the responsibility of the medical product manufacturer.” We have a robust medical device industry with thousands of manufacturers, however, and many of these manufacturers are relatively small, with limited resources for making their devices cybersecure.
As trustees, then, what should you keep in mind to help health care management address the device cybersecurity risk? We recommend some basic but important steps: prioritize, identify, protect.
Prioritize: As a first step, make sure you consider medical device cybersecurity in the context of other cybersecurity risks within your organization. While no patient harm has occurred to date as a result of hacked medical devices, several facilities have experienced loss or ransom of financial information and/or patient identity, often through relatively simple email phishing scams. Then, work with management to prioritize which types of connected medical devices need attention first. For example, make sure you are protecting patients by focusing on life-critical devices, such as ventilators and infusion pumps, over less risky devices. And protect your organization’s information by focusing on devices and systems that contain "Protected Health Information" under the Health Insurance Portability and Accountability Act.
Identify: Now that your management team knows which device types to start with, step two requires that your organization create an inventory of all equipment in the device types. This inventory should include key information such as the exact software version, network configuration settings, and which information systems or devices they are routinely connected to. Though this sounds simple, it is not. Without knowing exactly what software is running and what connections exist, it is impossible to establish and maintain good cybersecurity practices.
Protect: Step three is the active practice of maintaining and improving your medical device cybersecurity. Work with device vendors to make sure your organization is getting updated software to patch identified vulnerabilities. Make sure your wired and wireless networks are using appropriate security methods and that your networked devices can support these methods. Use safeguards like firewalls or private networks around less-secure equipment to reduce its risk, or, if a device can’t be secured adequately, plan for its replacement.
Collaboration and growth
While medical device cybersecurity gives us all reason to worry, two additional key questions for trustees to ask will help mitigate risk: (1) Who exactly are the people responsible for medical device cybersecurity in our facility? And (2), what type of ongoing medical device cybersecurity educational training are those team members receiving?
These questions are important because medical device cybersecurity requires a blended knowledge of information technology and medical technology. The team responsible for protecting patient care and information must have expertise in both areas.
Often, medical device cybersecurity requires collaboration between IT, clinical engineering, risk management, legal, sourcing, compliance and audit departments or functions. These departments may be working together for the first time in coordinating your preparations for medical device cybersecurity across your health systems. Like any collaborative endeavor, we suggest making sure that there are clear management processes in place.
For example, because many medical device cybersecurity risks arise suddenly — as with the WannaCry attack — each institution must set policy on both responsibility (who is responsible for what aspects of the system) as well as escalation and handoff (how to transfer aspects of a multi-department issue to the right responsible party) in advance.
Furthermore, because we are suggesting that cybersecurity policies and procedures address the specific threat related to medical devices, not simply IT systems, your organization must build the medical device cybersecurity risk assessment program into the basic IT security program or parallel to it. In particular, the program must not only include a risk assessment but must also create a program of proactive application of manufacturer-validated software patches for medical devices.
In addition, training must be ongoing because medical technology and IT change so rapidly and malicious actors become increasingly more sophisticated. While the field of cybersecurity for medical devices grows, trustees should help ensure that appropriate personnel, time and money are available for staff to access and participate in the many different organizations trying to help keep the health care community well-informed of potential risks.
Connected technology has obvious advantages, but as with progress in many areas, it brings new risks. If we understand these new cybersecurity risks and take the steps above, we will create a more protected network and a more secure patient experience. The price of connectedness is eternal vigilance, but the value of connectedness is truly better care. So, be vigilant.
Anthony J. Montagnolo, M.S. (email@example.com) is executive vice president and chief operating officer of ECRI Institute in Plymouth Meeting, Pa.
A number of organizations are working to inform and ultimately protect the health care field and the public in the area of medical device cybersecurity. A number of publications, alerts, seminars and tools are available from groups including: