Tomorrow’s Medicine, Today: A CIO's Vision for Enabling a Better Future
For Penn Medicine CIO Michael Restuccia, coordinating gene maps, big data and electronic medical records isn’t the future — it’s here and now.
Imagine a healthcare future in which your physician runs a sample of your blood through a genetic sequencer to create a profile of your genes and identify your propensity for certain diseases or how you metabolize certain medications. Using that information, your physician of the future then treats you with a customized plan optimized for your unique genetic profile.
No need to wait. That future is taking shape now at the University of Pennsylvania Health System, better known as Penn Medicine.
Blending big data, gene sequencing, electronic medical records and patient care, Penn Medicine houses its nearly 22,000 employees on a campus of 10 buildings — including hospitals, medical schools and research facilities — with three others under construction. It boasts a central IT staff of 500 and an annual budget of about $130 million.
The following is an edited version of the discussion with CIO Michael Restuccia.
How do you approach IT management? I'm a big believer in the three C's: common systems, centrally managed and collaborative in nature. We used to have many decentralized IS [information systems] groups; now we've essentially rolled them into one central IT department, so it's all centrally managed. Our thinking: To support a common vision in a world that's very expensive, the best way to keep everybody in sync from an operations perspective is to be centrally managed and migrating toward common systems. We recognized a long time ago that when you start spending $100 million-plus each year, you'd better have it pointed in a direction that everyone agrees on. One key step has been our development of a strong IT- and operational-governance structure. Our senior and midlevel IT leadership makes virtually no decisions on its own. We do everything collaboratively with operations. We have a senior-level oversight group that sets the direction for IT, as well as other groups. The combination of these groups and their subcommittees provides the sponsorship and executive support for our IT group.
You mentioned common systems. What are you doing there? We're moving to one common platform, at least on the health-systems side. When completed in 2017, it will put all of our hospitals, ambulatory practices and homecare on the same IS platform. That will support our goal of providing integrated data for integrated patient care. Any caregiver will be able to see a patient's data, no matter where the patient has been seen in our system previously. Getting to a common system wasn't an easy decision. In a large place like this, there are always people with other preferences. If that decision had been made just by IS, there would have been a lot of conflict. But with the decision made by a committee of physicians, nurses, administrators, researchers and others, it's a much easier sell. Over on the research side, we've made huge strides by centralizing our IT support and moving to a common data warehouse. That warehouse, in turn, integrates with our patient-care systems. We can port what we've learned on the research side to our common health-systems side. The goal: what we call precision medicine.
Precision medicine? What's that? Precision medicine is the use of an individual's genetic profile to guide decisions about the prevention, diagnosis and treatment of diseases. This means our clinicians can treat patients based on their genetic profiles. A patient can give some blood or tissue, and we'll run it through a sequencer that generates up to a half a terabyte of data. With that, a researcher can identify genetic markers that might indicate a propensity toward, say, prostate cancer, heart disease, a unique drug targeted at the patient's specific tumor type, whatever it might be. These genetic markers are not absolutes at this point, but they are indicators. We believe that the better informed our clinicians are at the point of care, the better decisions they can make, and the better care they can provide. What an academic medical center brings is the combination of patient care, research and education. So we can take the research component — especially around genomics — and meld it with patient-care data. That's where the special sauce is, and ultimately the new discoveries.
But these discoveries come at a high price, don't they? Yes, and with these kinds of systems, the barrier to entry is difficult. You can't be a small hospital without a research arm and afford to buy into genetic-sequencing. We've stood up a high-performance computing center made up of 2,000 core servers and several petabytes of storage. You simply can't buy the combination of technology, science and process off the shelf. And for our researchers, this facility far exceeds what they could get with their own resources. You also need a strong electronic medical records (EMR) system on the health system side. You need strong adoption of that EMR across your clinicians. You need good governance to make sure you're effectively using those systems. Then, once you get all that, you need to tie it together with a research computing group that's willing to play along. That's what we have here, and that's what's really unique.
How do you derive value from data? One example: We've implemented an early-warning system that continuously surveils patient data to identify potential problematic changes in the patient's vital signs or other clinically relevant indicators that would indicate the patient may be developing a sepsis infection. The major benefit of this surveillance is that the earlier you detect such an infection, the quicker and more easily it can be addressed. Or the system might compare a [newly prescribed] dose of medicine with the patient's previous doses. It could give a warning: "That's a potentially dangerous dose. Did you mean 0.15 mg instead of 1.5 mg?"
Given this collaborative setting, how have the requirements of your CIO position changed? I now need to be much more consultative. It's no longer this dictatorial IT guy sitting in the hospital basement. Now, the role is much more of a team-builder, someone who creates strong relationships with operations, recruits and retains high-caliber personnel, communicates powerfully and isn't afraid to take a position. You must also have a knowledge of many topics that's broad yet also deep enough to understand and drive them. I have to worry about mobility, security, functionality, vendors and more. If I'm simply an administrator, that's not good enough. I need to really understand what the issues are, and where the opportunities are.
How important is mobile technology? Extremely important. From the patient-care perspective, mobile must follow what I call the "five rights": providing the right patient data to the right decision maker at the right time in the right place via the right device. We have many different types of settings. There's an ambulatory setting, where you have patients going into an examination room. There's a hospital setting with patient rooms, intensive-care units, surgery rooms and more. Each of those settings might require different types of data and perhaps even different types of mobile devices.
How about the cloud? Not as much as some other industries. The issue is the privacy, security and confidentiality of our patients' medical records. Compared with other industries, healthcare hasn't become comfortable enough with the cloud yet. Also, we're still investing in applications that are mainly server-based, so we're not going to be changing rapidly.
Speaking of security, what special challenges do you face? The concerns around security in healthcare are huge. But another big challenge is maintaining a balance between privacy, security and confidentiality on the one hand, and facilitating and enabling operations on the other. That is, we have to maintain the right level of security, in an environment where timeliness and accessibility to data are paramount. Yet if we make things too restrictive, it'll become so cumbersome that we'll be unable to treat as many patients as necessary. It's a balance we struggle to maintain all the time.
Earlier, you mentioned getting non-IT people involved with IT planning. But what if your healthcare professionals are already stretched for time? That's one of our biggest challenges. Say you're a surgeon, and people's lives depend on your hands; you're much more inclined to be treating patients in the operating room than to be sitting in one of my meetings. We know that from an IT perspective, it's people, process and technology all the time. We have really good technology. We have great governance and process. But getting enough of people's time — the right people's time — remains a challenge.
DXC has been working with Penn Medicine since 2007, providing hosting services and storage management.