Healthcare organizations are facing an evolving cybersecurity threat landscape, with accelerating threats in the wake of recent, massive attacks on organizations worldwide. The need for stronger cybersecurity is now on most healthcare organizations’ radars, more so than even just a few years ago.
Chief information security officers (CISOs) at patient care organizations face complex and challenging issues with respect to information security, including rapidly evolving malware threats, insider data breaches and the increasing use of medical Internet of Things (IoT) devices across their organizations.
At the same time, while the CISO position requires technical expertise in security, the role continues to evolve as healthcare CISOs have now become a part of the broader senior leadership team. Top healthcare security leaders from across the country will explore many of these topics at Healthcare Informatics’ San Diego Health IT Summit on Feb. 1-2. Brian Kreitzer, CISO, information services and solutions at UCLA Health, leads cybersecurity strategy at the four-hospital, 2,000-physician integrated health system in Los Angeles. Kreitzer recently spoke with Healthcare Informatics’ Associate Editor Heather Landi about the evolving role of the CISO, the challenge of strengthening legacy systems and the security threats he is most concerned about. Below are excerpts from that interview.
Legacy systems, from electronic health records to billing platforms, can pose a security risk. What is your perspective on strengthening legacy systems?
My thoughts are to get rid of them. I say that jokingly, but there is some truth to that as well. It’s not necessarily getting rid of them, but getting rid of them in their current state. If they are old enough to where they have security vulnerabilities, an organization needs to ensure that they have proper planning to be able to upgrade those systems, so that they can still make use of those feature and functionalities in a more secure way. So, typically that means upgrading both the infrastructure and application, because the application may rely on older versions of other third-party software, like JAVA as an example of one that is always vulnerable. The way that access management is handled is another possible security vulnerability in legacy applications and whether they can integrate well with some of the latest and greatest security best practices in that space, such as multi-factor authentication.
How would you describe the state of healthcare cybersecurity right now?
We are making progress as an industry from what I’m seeing, but there is still a long way to go and we’re still not where we need to be from a straight security perspective. But if you’re taking a risk-based approach then there are certainly some arguments that could be made that perhaps we are where we need to be because the more security controls you put in place then the less likely it is that you do get breached. And, it’s about trying to figure out what is that percentage and what is that real estimated value there, which nobody can do, and that’s the problem. Nobody knows what that expected value is so we keep addressing what we consider high and critical and hopefully some medium vulnerabilities and threats. I think we’re making progress, and we’re much more secure as an industry than we were probably 12 months ago.
Within the healthcare industry, are you seeing more engagement and support for cybersecurity strategies from senior executive and board-level leaders?
It seems like that is very company-dependent on the level of exposure and influence or engagement of boards and/or senior leadership. I’ve seen some organizations take a more check-the-box, compliance approach. But, others may have the CISO reporting directly to the CEO. And so, to me, the reporting structure definitely goes a long way and says a lot about how seriously a company is taking security.
How do you see the reporting structure impacting an organization’s security strategy?
It really depends, partly, on the level of engagement and the knowledge of those people in those reporting relationships. What I mean by that is I, personally, think it’s okay for the CISO to report into the CIO, but only if the CIO has a good understanding of risk and finances and the rest of the business to be able to measure that and make those risk-based decisions. If the CIO is strictly technical, then that might not the best place for a CISO and maybe they should be more on the risk side of the house, or even under the CFO. Obviously, if you are trying to drive change, then the higher in the organization you put that person, the more influence they can have around driving that change. If you’re organization is in a spot where you need a change agent to come in and shake things up a little bit, then reporting to the CEO, or maybe even directly to the board, might be the way to go.
How is the role of the CISO evolving?
What I’ve seen in the past, and what I hear from my colleagues, is that many times, the CISO has been much more technical in nature, so they will go in there and solve technical problems. Obviously, the entire landscape and architecture of our IT solutions that we put in place has completely changed over the past 10, 20, 30 years. It used to be good enough to where you used the M&M approach and protect the outer shell, throw up some firewalls, put some anti-virus on and it’s done. But now, in the technical space, the whole landscape has changed. Our users are everywhere; they want to bring in every device, they are using cloud, they are on the network, using the same devices going back and forth on and off your network, and they want a seamless user experience. The CISO needs to have enough of a technical understanding to be able to work all that out, and measure the pros and cons to be able to make those risk-based decisions based on how much a new solution is going to cost versus the expected value—i.e. how much it is really going to move the needle on your security posture.
But, more importantly, I feel that the CISO needs to be engaged at that higher level. The CISO needs to have that business acumen and understanding so that they can make those risk-based decisions based not only on the new technological widget that might barely move the needle, but they need to be considering business impact, user impact, operational impact on our IT groups, the financial impact with cost versus savings, all of that to be able to work with the rest of the business and really be true business partners.
AT UCLA Health, what is your approach to utilizing technology as part of the security strategy?
There’s absolutely a huge technical piece to the role, and in my mind, that’s a must. The CISO must have that level of knowledge because there are certainly different ways that you can approach security and you can approach it through administrative controls, but those only go so far and they are only as effective as the sanctions that are around those controls. If you want a truly effective approach to security then you enforce it through technical means, so you’re not just relying on your users to say that they’ll do things in the right way, but you’re enforcing it and ensuring that they cannot do it the wrong way. There is of course a risk that users will find their own workarounds, so it is also critically important to offer secure options that enable the business.
Specifically, how are we using technology? We’re basically looking at every different type of security technical control out there and measuring it in a risk-based approach on how much do we think it’s really going to move the needle versus what’s the cost of the solution. So, yes, we’re looking at SIEMS, we’re looking at user behavior analytics and DLP, you name it.
Breaches due to insiders, including phishing attacks, are also a risk. What is your perspective on addressing insider breaches and building a culture of cybersecurity?
I probably have a unique perspective on phishing, as I’m not sure how much phishing training actually helps, because no matter how much you train, there are still people who are going to click on bad links and get phished. To me, it really should boil back down to the finances on those phishing training solutions and how much that really saves you on the back end with your incident response teams. To me, it’s a straight finance and productivity type equation. With that said, the importance of ensuring that we have a security culture in the organization is absolutely, critically important. We should continue with at least annual cybersecurity training to makes sure that’s on the forefront of everybody’s minds for at least some period of time. When you go live with a new security tool, you should absolutely ensure that you’re doing everything you can to work with your users and ensure that you have deep communication plans in place and walk them through it. And partner with them as well so they can be part of the solution and you can find workable solutions together, instead of having your users feel like security is being jammed down their throat and causing disruptions. That engaged and partner security culture is important because you don’t want people going around your controls and breaking policy. Policy, when enforced correctly and with the appropriate sanctions, can also help your security culture.
It’s also critically important to have those C-level folks, executive leadership, on board with your program. Again, it goes back to why the reporting structure is so important because culture is not something that any one group of people can change. It has to come from everywhere. You can have change agents coming in from a security team but if you don’t have executive leadership on board and if you’re not doing grassroots type of things like annual training, then a culture change never really takes effect. You have to approach it from all angles.
What do you see as the biggest cyber threat? What approach is your organization taking to address it?
Medical devices are definitely the biggest vulnerability that we have. It’s really about working with vendors, at the end of the day, and coming up with partnerships that you can move forward with that support a secure environment. Historically, medical devices don’t typically come with a lot of security built into them so your only remedy is to quarantine them off in some corner and not let them talk to the rest of your network, or just the minimum. So that’s one approach, but that doesn’t solve the problem of physical harm for that one patient that is hooked up to that medical device. You need to work with your vendors. We as an industry need to do a better job of saying no to contracts that don’t have the ability to have security worked in in a meaningful way.
Are you seeing improvements in that area, with vendors more focused on security?
I have seen some improvement, but it’s a relative term. It feels like we’ve moved from 0 percent to 1 percent, so we’re still a long way from where we need to be.