In May, the University of California San Francisco (UCSF) announced it was creating a Center for Digital Health Innovation (CDHI) to lead the transformation of healthcare delivery and discovery from empiric, generalized, disease-based diagnostic and treatment approaches to the era of individualized precision medicine.

Precision medicine is an emerging field that aims to harness the wealth of data available from the human genome and research into the molecular basis of disease and integrate it on both a personal and global level with information on environmental factors and patients’ electronic health records (EHRs). The practice of precision medicine would allow scientists to share emerging research findings faster, drug companies to develop more precise therapies, and clinicians and patients to make more informed decisions about treatments that would ultimately improve care, save lives, and reduce healthcare costs.

As such, the focus of the CDHI will be to develop new technologies, apps, and systems that, along with the explosion of social media, will generate enormous new data sets. In recent years, UCSF has invested heavily in the development of a variety of information technology and management resources to give healthcare providers, educators, scientists and students the tools to succeed in the rapidly evolving digital age.

UCSF Chief Medical Information Officer (CMIO) Michael Blum, M.D., has been tapped to lead the CDHI in the new position of associate vice chancellor for informatics. In his new role, Blum, a cardiologist and clinical professor of medicine, will coordinate and leverage UCSF’s information technology assets. Recently, HCI Assistant Editor Rajiv Leventhal had a chance to speak with Blum about the goals of the CDHI, projects they are working on, how to best manage the social media digital deluge, and the key in advancing precision medicine. Below are excerpts from that interview.

What are the main goals of the CDHI?

In terms of the overall goal, it’s three-fold. One is to create a front door where internal faculty innovators can bring in their new ideas, concepts, and creations, and get supported with development and technology expertise. There has always been the issue of, “I have this great technology, how do I wedge it into healthcare?” vs.  “I have a healthcare problem and have an idea on how technology can help fix it.” Historically, people have had good ideas but fear that the university is an ogre. And when that happened, the inventor would go outside the university to talk to a brother or cousin who would set them up with a developer and that would usually lead to failure since it would get developed in someone’s garage. In reality, yes , the university does technically own the intellectual property since they pay you to work here, but at the same time, the university is very flexible and progressive in forming agreements with the inventors of the property that allow everyone to benefit from it. So creating that front door is the first big piece.

The second aim is about validating the functionality and accuracy of new digital health devices, sensors, and technologies, and evaluate whether they bring value to patients and the healthcare system. We get approached by external investors who tell us they have a great new digital technology to revolutionize healthcare. We’ll then ask them, “How do you know it does what you say it does?” They’ll respond by saying, “Well, we had world-class engineers build it.” Then we will ask, “Were they healthcare engineers?” You have to validate that it measures what you’re saying it measures,—that’s the first step. I’m not interested in new technology that increases the cost of care. Instead, I want to decrease the cost of care and bring better outcomes. If it doesn’t bring true value to the healthcare system, then no thanks.

And the third aim is to incubate important new digital technologies, apps, sensors, and systems, and bring them to market via collaborations with start-ups and industry and capital partners. We’re good at developing proof of concept, we’re good at bringing things to the pilot study level and piloting them internally, but we don’t fancy ourselves as industry partners.

The CDHI is working on several digital health projects. Can you explain some of them?

First is GreenDot, which collects data from various diabetes-related devices into one location, allowing both web and mobile applications to leverage the data for analytics and better visual displays for patients and physicians. This was co-founded by faculty and technologists, who were personally interested in juvenile diabetes because they came from juvenile diabetes. Typically, kids with juvenile diabetes have visits four times a year with providers, and there isn’t much that can be done with these reams of data in just 15 minutes. But now, you’re getting deeper data analytics that provide information, so there can be a discussion in those visits. And this is something that will be worked into the clinical workflow— building the connections to the EHRs so you can integrate it into the provider workflow is a big piece. Providers need information and knowledge; what they don’t want is more data being thrown at them.

Next is Health eHeart, a social media-based cardiovascular study in which we are developing a scalable social media clinical trials platform that integrates with the campuses’ clinical and research information resources. The Health eHeart Study expands on the Framingham Heart Study, [which tracked 5,209 men and women in Framingham, Mass., starting in 1948]. That study tracked participants in one city; think about that model on a global scale and taking the collected data to treat heart patients so precisely that we can account for their gender, age, ethnicity and lifestyle factors.

Then there is CareWeb, which is a collaborative, team-based clinical communications platform. Think of this as changing the game from an individual sport (patient-doctor relationship; doctor-nurse relationship, etc.) to a team sport, where there is a whole team caring for you. Think about Facebook for healthcare—no one wants to use Facebook for privacy reasons obviously, but imagine a secure platform that integrates with the EHR so the entire care team could see everything.

Expanding on that, how do you plan to manage the social media digital deluge?

While healthcare has not experienced the full force of the social media revolution, it will shortly. It has already changed the way we fundamentally communicate as a society. But how is that going to play out in healthcare? Well healthcare isn’t typical or traditional communication, so no one is putting his or her healthcare information on Facebook, right? But what they are doing is using Facebook to talk about it and gather information. So the question is, “How can we use the fact that people are communicating about healthcare and looking for information to get them to engage in healthcare more effectively?”

We are also studying what creates stickiness in social media around healthcare. People in social media tend to be very in and out and very transient, and that’s not how you want to do healthcare. Transient doesn’t maintain wellness. It’s about stickiness, about persistence, and about true lifestyle change. We need to learn how to assess which apps, systems, and sensors will be reliable, persistent data sources, and which will be a flash in the pan. Some of these tools will be incredibly valuable and will change the way we understand and deliver care, but most will not. There are a whole variety of ways in which social media can change things, but you need to understand its place in healthcare.

Ultimately, what is the key to advancing precision medicine?

We need to leverage our digital assets to build the knowledge network and information commons that are the foundations of precision medicine. The knowledge networks not only link sources together but then bring this knowledge to providers at the time they’re providing care.

So when I have a patient in front of me now, I can go to different risk calculators, and say, “Based on your family history and risk factors, you have a five percent chance of having a heart attack in the next 10 years.” Although that is better than nothing and sometimes it might motivate a patient to change behavior, it’s ultimately not enough. What I really need to be able to tell them, based on what we know, based on the censored data, the biomarker data, and the 100,000 patients similar to you—genetically and life wise—your risk is actually 50 percent, or your risk is actually a quarter of a percent. We need to make those strong predictions, and prescribe the right drugs that they’ll respond to. That’s the real move to precision medicine.