Tying Tech to Care
People can do an incredible range of things remotely today. From a chair at the office or under an umbrella at the beach, they can adjust lights and appliance settings at home, monitor visitors on their doorstep, and check in on their teenagers’ whereabouts. When it comes to health care, however, there’s still a long way to go before patients can get even simple care advice without having to make an appointment and trudge into the clinic or doctor’s office.
It’s about time for that to change, according to Wendy Nilsen, director of the Smart and Connected Health Program at the National Science Foundation (NSF). Jointly funded by the NSF and the National Institutes of Health (NIH), the program is providing US$23–$25 million in grant funding every year to innovative researchers with projects that will help push patient care into the digital age.
IEEE Pulse caught up with Dr. Nilsen to learn more about her vision for the Smart and Connected Health Program and how it can accelerate the transition of the industry into one that lets health professionals do what they do best, while giving patients the kind of responsive and excellent care they desire.
IEEE Pulse: How do you describe connected health?
Wendy Nilsen: Connected health is the new term for digital health but with an ecosystem view where things work together. Mobile is one piece of that, but it also includes a wide range of other things that happen in a technology-enhanced world and that might include mining social-media data to figure out information on adverse events or data analytics within an electronic health record (EHR) where the data is irregular, may or may not be correct, or may have other issues that make it a mess from a traditional analytics point of view.
Beyond that, connected health also means thinking about how we give patients devices that will help them understand their health in real time. That doesn’t mean that all of the mobile health data has to come back to the health care system. For instance, if you’re a patient with diabetes, you wouldn’t constantly feed every bit of information about your diabetes management to your doctor. There’s not enough time for a doctor to look at how you’re getting your meals every day, if you’re getting the right amount of exercise, and if you’re varying your medication dose based on your activities. That would be a full-time job itself. Instead, the doctor would only get certain information, such as if the patient was over some threshold.
Too often when we talk about connected health, we tend to focus on how the data get to the health care system and how the system itself deals with that data. While that’s important, I’m much more excited about how we can start with the patient and work outward from there.
IEEE Pulse: So you’re saying that the ultimate advantage of connected health is not to make the health care system work well from an institutional standpoint but to make sure the health care system works well for the patient.
IEEE Pulse: Would you provide some examples of projects the Smart and Connected Health Program has funded?
Nilsen: Last year, the NSF funded an early-career award for a computer scientist who was looking at using microfluidic impedance sensing to develop in-home sickle-cell testing for children. The idea was to give parents a tool so they could understand when their child’s cells were starting to change, because pain and sickling are related. The tool would support the parents, who then could do something immediately, which is a big improvement over having this poor child curled up in his or her parent’s arms screaming in pain while waiting to see the doctor.
Another project we funded last year through the NIH—the NSF funds some and the NIH funds some—was a teeny microsensor that is injected into a tumor to tell in real time whether the tumor is responding to treatment. So instead of the doctor having to run a full 12-week course of chemo and then say, “I’m sorry, it didn’t have an effect,” the sensor could potentially indicate very early on that the tumor is showing no response at the cellular level, and the doctor can rethink your treatment much more quickly. And since cancer treatments are toxic—they kill off cells—wouldn’t you like to know that if you had cancer?
IEEE Pulse: Why do we need the Smart and Connected Health Program to encourage these kinds of projects?
Nilsen: When the NIH and NSF first partnered on this, what we were finding was that people had great ideas and were doing great science, but they were often looking at something personal. The other day, for example, I was talking to a researcher who was doing a cancer project based on a problem her mother had but that actually might not be a common problem, or it might be a problem that science has solved already. These are good researchers, but they don’t know what they don’t know. So what was happening was that researchers were often doing these projects, building the prototypes, testing them on small groups of people, and then the projects would die because no company or health care system would pick them up.
What should happen is that researchers collaborate very early on in the process. The Smart and Connected Health Program helps to provide that bridge between researchers in different scientific communities. Here’s an example. We funded a project last year for a really interesting sweat sensor. It’s a fouryear project, and we’re working with the researcher to help her make connections in different fields so that, at the end of that four years, she’ll have something with real applications, and she’ll know how to take the prototype to the next step. If she does need help, the NSF can then perhaps make some introductions so she can bring it to the NIH or the Department of Defense to work with them or give whatever other assistance she needs. The point is that there’s a pathway so she can go forward. This is how the great ideas that are coming from computing, engineering, and information sciences are going to find their way into health care.
IEEE Pulse: People have been talking about the benefits of multidisciplinary teams for a while now, but it’s only recently that both the science side and the clinical side are coming together. Is that a key to good connected health projects?
Nilsen: For me, that is where they become connected health projects. What I often found in mobile projects was that the engineers would use text messaging as the communication mode because they knew how to do that, but that’s a really bad reason to do something. It’s better to have a really good partner and build the tool you need versus building a tool because you can. My goal now is to bring people together so we have the science, the clinical understanding, the relationships, and the tools we need to really move all this forward.
IEEE Pulse: What are the barriers to that?
Nilsen: There are barriers at so many levels. For one thing, there’s a language barrier. We had a multidisciplinary meeting a while back, and a cardiologist started talking about cabbages. All of my computing friends looked very confused as to why we’d transitioned to vegetables in our conversation, so I finally asked if he could explain what cabbage is for those of us who eat cabbage? And he said “Oh, it’s coronary artery bypass surgery: CABG. It’s an acronym.” You don’t realize the number of acronyms we have and how they limit projects, so we have to foster understanding across disciplines.
Another barrier involves respect. I think the health community often sees people in computing, engineering, and information technology strictly as technologists who program things or who build stuff, but they don’t see them as scientists. And I think the opposite is true, with the computing and engineering people seeing the health community as solely clinicians rather than researchers. It all comes down to understanding what each member of the group brings to the relationship. When you do that, you realize it’s totally worth the effort because you can see your initial great idea turn into something that will make a real difference.
IEEE Pulse: Are you seeing any trends in connected health?
Nilsen: We’re seeing a lot of interest in bringing multiple sensing modalities together now. Last year, for example, we funded a rehab device for older adults who had suffered stroke. In the past, people had been creating these interesting algorithms using Kinect (a motion-sensing device designed for video gaming), which would let health care professionals monitor remotely how well a patient was doing his or her exercises at home.
IEEE Pulse: So the patient is wearing sensors while following a video exercise routine or a game, and the information is sent to the health care professional in real time.
Nilsen: Right. The problem was that adult children would see that Mom or Dad was unstable, so they’d help, putting a hand on Mom’s or Dad’s elbow to stabilize them or even actually pushing their arms to help them do an exercise. The rehab person who is monitoring, however, doesn’t know that 1) somebody else is there with the patient and 2) what that person did to help. To correct that, this new rehab device is using capacitive sensing and two-person imaging in Kinect that not only knows there’s somebody else in the room with the patient but also includes special gloves for the second person that can sense what they are doing and can feel how much pressure they’re exerting. That allows the rehab person to ask the second person to perhaps back off a bit or to let go for a few seconds and therefore provide a better idea of what the patient is actually able to do. It’s a much more complex system that integrates information from multiple sensors and provides data that we can provide feedback on.
We’re seeing this interest in creating cyberphysical systems using sensing technology, so we sense things, provide feedback or intervention, and continue to monitor, which would change health care into more of a closed-loop system. That also means we have a big need for advances in analytics and signal processing so we can work with this data in real time. I think in the mobile space, specifically, we have not spent enough time thinking about how you analyze that often-messy real-world data.
IEEE Pulse: Would you provide an example of a closed-loop cyberphysical system?
Nilsen: Here’s a great one that shows how mobile health can come full circle. A couple of years ago, we funded a project to create a dynamic system model to promote physical activity. It would start monitoring and collecting data when a person began walking and then run small experiments to determine what types of variables had an impact on how much that person walked: Was it the weather? Was it appointments in that person’s Google calendar? Was it encouraging messages, like “good job”? All that information fed into this dynamic system model to figure out what makes that particular person more active, so it’s a closed loop.
Looking forward, the artificial pancreas is coming, and, in time, it should be a perfect closed-loop cyberphysical system. It will not only monitor your glucose and inject insulin or glycogen as needed, but I think we’re going to start seeing that it will take into account what the patient is doing because, if you’re running a marathon, that’s hugely different for your insulin needs than watching TV for an afternoon. That will make it a closed-loop system that incorporates all these different variables, because that’s how your body works.
IEEE Pulse: What is it going to take to fuel this evolution of connected health?
Nilsen: It’s going to require the funding to be in place, it’s going to require the conversations and the relationships to be built, and it will take work. But we have people entering the research world now who are coming up as digital natives. Connected health is just going to make sense.
I always tell the story of my son talking to a nurse at an insurance company, and he just wanted to text her a photo of the lump he had on his eye to see what he should do. There was this long silence at the other end, so he said, “OK, instead of that, give me your e-mail and I’ll e-mail you the picture.” And the woman said she couldn’t accept an e-mail from him because she was in a different state and she wasn’t licensed in the state he was in. And he said, “Are you serious? How can you tell me what to do if you can’t see it?” So she suggested he go to emergency care. He gets there and says, “I’m outta here. These people are really ill, and I’m going to get sick.” So he left. It turned out that he just had a stye on his eye. He should have been able to send the nurse a picture, and she could tell him how to take care of it at home. But, instead, he ends up checking into an emergency room. The health care system simply cannot sustain care like that.
The question is, how do we use technology for a lot of the retrieval, coordination, organization, and visualization of information so that the people with talents in health care aren’t bogged down and can really focus on what they do best?
IEEE Pulse: Where will connected health have its biggest impacts?
Nilsen: I think it will have a huge impact on health care everywhere and especially in rural areas. The issue isn’t that rural areas are necessarily poor in terms of money, but they are poor in terms of resources and, specifically, in the number of experts available. In fact, one of our projects in rural Peru was looking at patient X-ray images for tuberculosis. In some areas, there aren’t specialists to read those X-rays, so that project used an iPhone to take a picture of the X-rays, upload them to the cloud, and then developed an algorithm that uses machine learning to identify tuberculosis from the picture.
Here’s another example. I used to have a friend who was the only pediatric cardiologist for the Indian Health Service (a division of the U.S. Department of Health and Human Services), and he obviously couldn’t be everywhere he was needed at once. How do you have the ability to call him into a team meeting? How can you have him kind of in your head as you do surgery? These are the kinds of things that smart connected health help us do. We can provide the right digital technology, the right information, and the right visualization to providers and patients to really help them make the best out of the resources they have.
This is an exciting time for smart and connected health. I think we have a research community that’s ready for it and really wants to do great science that changes the world.