Technology for a Healthier World

associated symptoms, behaviors, and outcomes)

• To measure the correlation between the known epidemi­ology (patterns, causes, and effects) of these conditions and those reported via Twitter

• To conduct a randomized controlled trial to determine which Twitter messages resonate with patients with heart disease and can help to manage the disease.

Previous projects at the intersection of social media and health care have done things like flag flu outbreaks, track sentiment about the Affordable Care Act, and predict heart disease or depression based on tweets, but, says Merchant, “This project is really focused on using Twitter more for intervention, not just passive learning.”

Lyle Ungar, Ph.D., a professor of computer and information science in Penn’s School of Engineering and Applied Science, studies the use of social media to understand the psychology of individuals and communities. “We tend to think of health problems as being treated by drugs,” he says, “but people can do a lot to improve their health through behaviors.” In the short term, says Unger, who is also a member of the Graduate Group in Genomics and Computational Biology in the medi­cal school, the focus has been on collecting data (that means culling through billions of tweets) and building the systems to automatically look for patterns. The long-term goal, however, is intervention — helping people improve their health. “The key to treatment is changing behaviors.”

Answering the Call to Participate

The growth of social media – with its potential as a public health tool – is only a small example of the broad impact of connected devices on medicine. According to the National In­stitutes of Health, there are more than 285 million wireless sub­scribers in the United States alone, and an estimated 67.6 per­cent of adults worldwide own cell phones. Given the laundry list of things we can do with a few swipes or taps of a finger – send Facebook messages to long-lost high-school friends, do our grocery shopping online, or conduct banking transactions – it may seem obvious to use mobile technology for recruiting research participants. But only recently has this been made a reality. Working with a team of doctors and software engineers across the country, Kathryn Schmitz, Ph.D., M.P.H., a professor of epidemiology whose interests include prevention, public health, and physical activity, is exploring new ways of conduct­ing larger, more accurate research studies – and faster.

“We’ve all seen the flyers people use when they are trying to sell furniture or looking for a new roommate – the ones with the pull-off tabs with a phone number to call if you’re inter­ested. That’s what we’ve been using to recruit research partic­ipants for decades,” says Schmitz. “In today’s world where we have technology in the palm of our hand that lets us watch live TV, set DVRs, and control home-security systems from anywhere in the world, it’s crazy that we’re still using paper flyers to conduct what is basically a series of simple surveys and questionnaires.” 

ResearchKit was unveiled earlier this year by executives at Apple, Inc. Built as an open-source software framework, the kit makes it easy for researchers and developers to create apps that could greatly facilitate medical studies. Already the frame­work is being used by leading institutions across the country to study aspects of diseases, including Parkinson’s disease, cardiovascular disease, and breast cancer. With ResearchKit, researchers can create apps that take advantage of mobile phone features to gather new types of data on a scale never available before. What’s more, many of the apps built with ResearchKit will enable study participants to track their own data and potentially discover correlations between symptoms and daily actions such as diet or exercise.

The breast cancer app, which was developed with guidance from a panel of experts that included Schmitz, is called Share the Journey. The app allows women living with breast cancer and survivors to answer questions about their health, perform activities related to clinical trials, and generate real-time data wherever they are, providing a source of information that is more objective than ever before. The goal of the study is to understand the symptoms and late effects that can follow treatment for breast cancer, why these symptoms vary over time, and what can be done to improve them. 

For an earlier research study on breast cancer, Schmitz’s team sent out more than 60,000 letters to recruit participants. Out of that total, they were able to enroll only 351 women. In contrast, with Share the Journey, there were more than 13,000 downloads within three days; 1,060 women consented to par­ticipate, and 1,020 of them are still enrolled in the study.

For Schmitz, however, the picture is much bigger than a single study. “It’s not about a specific app or what we have chosen to track in a particular app,” she recently said in an interview with CNBC. “It’s about the platform, and the fact that researchers will be able to collaborate to be able to build whatever they want to track in patients. That’s what’s really exciting.”

There’s no denying that more research is needed on the ef­ficacy, accuracy, and ethics of using mobile apps to collect data, but for now, most experts see the continuing develop­ment of these software frameworks as a much-needed and long-awaited big break in the way research is conducted. Ac­cording to William Hanson, M.D. ’83, vice president and chief medical information officer of the University of Pennsylvania Health System, studies that are done well and that permit pa­tients to use smartphones as vehicles to enter and/or collect data “have the potential to broaden the scope and reach of research considerably.”

“When we monitor and gather information in real time – whether it’s for a research study or to coordinate care – pa­tients see the value in that.” But, he adds, he does not believe that most providers are yet prepared for “the data dump from everything their patients will be able to track.”

“As we move toward digital data, the big challenge is to un­derstand how to take all of the info that’s now digitized and immediately available, and transform an industry which has historically been very paper-centric,” says Hanson.

With the advent of electronic medical records (EMRs), pa­tient information became as close as the nearest computer. But like research studies conducted on paper, securely logging in to a desktop and navigating EMR screens can be a lengthy process. As Roy Rosin, M.B.A., Penn Medicine’s chief innova­tion officer, explains, EMRs were not designed to collect or share data as part of population health management, which makes it difficult for multidisciplinary care teams to collabo­rate and deliver streamlined care. 

“There’s great value in having EMRs as an organized, single point of health history, but it’s still too hard to get information in or out of them,” says Rosin. “One of the things we’re work­ing on is developing flexible systems that will allow us to cap­ture information and share it directly with the team making the decisions about a patient’s care.” 

First introduced to Penn Medicine late last year, Carelign (formerly known as Connexus) uses the latest advancements in technology to take clinical data to a new level. The “web app,” created by Information Services in partnership with the Office of the Chief Medical Information Officer and the Cen­ter for Health Care Innovation, provides real-time patient in­formation in a mobile-friendly view. According to Hanson, “The site frees the provider from the need to log onto a com­puter or depend on out-of-date information from printouts produced hours earlier.”

Carelign also presents the data in a format that “we as clini­cians need to make medical decisions,” says Subha Airan-Javia, M.D., G.M.E. ’07, of Internal Medicine, who worked with Glenn Fala, senior director for Software Development, and the Information Services team to create the application. “I can easily see trends, such as a sudden spike in a value.” And, as she told Technically Philly last year, it’s 22 times faster to use the website on your phone than it is to log in to a computer!

Carelign is a mobile web application, not an app, which means there is nothing to download. It is essentially a mo­bile-friendly web page that can be accessed from any mobile device that is connected to the Health System’s network. That connection ensures that the patient’s protected information displayed is secured behind the UPHS firewall. “We’re in the process of deploying smartphones to all clinical care faculty and staff across the health system, making it an essential part of the medical toolkit, analogous to a stethoscope,” says Han­son. He adds that via Carelign, providers will now be able to securely access patient data from anywhere.

Providers can view a variety of clinical information on Carelign, including up-to-date vital signs, labs, radiology stud­ies, and medical history reports. But its use goes well beyond just accessing EMR data. According to Airan-Javia, one of its best features is the ability to quickly find the names of each member of a patient’s care-provider team and to keep every­one is in the loop, sharing information and up-to-date on evolving care situations. Carelign maps out the names of the patient’s attending physician, primary intern, covering intern, and nurse, as well as any other specialty services that are in­volved in the patient’s care. Even better: the clinician can click on any name on the list to immediately call or e-mail. 

Carelign has already impressed outsiders. In December, while still known as Connexus, it received a Fierce Innovation Award in the category of Clinical Information Management.

“There’s a massive inefficiency in health care right now,” says Hanson. “Patients often need to go from one venue to the next, having tests repeated because information can’t be transferred efficiently. A lot of the money we spend on health care goes toward the administrative burden of getting records, information, etc. What we’re working on is essentially provid­ing zone coverage.” If a patient needs to see different doctors, he notes, their information must be available. “Health care IT has the potential to address a lot of that.”

Unlimited Opportunities

It’s no surprise that Carelign is one of many endeavors Fala and his team are tackling. Over the past 15 years they have created more than 75 apps. 

The role of technology in health care clearly goes beyond apps and smartphones. Teams across the health system are working on a range of innovations. Sensor technology, such as sensors embedded in “smart” pill bottles to stream data on medication compliance and enable new high-impact interven­tions, is already showing great promise in improving health care efficiency and patient outcomes. In addition, as Rosin points out, new technology can limit the inevitable bias that has always been a problem in surveys of patients: “There’s a great value in subjective questions – what kind of pain they are experiencing, or what they are worried about.” But sensors, for example, are able to register patients’ compliance with tak­ing medications, their activity, or their daily weights accurately and without requiring them to expend any additional effort. 

Researchers are also taking predictive software systems – similar to those used by Netflix and Amazon to suggest movies or books a user might enjoy – and integrating it in the doctor’s office. Once in place, these systems will allow care providers to better assess patients at risk for various conditions based on their current health as compared to similar populations. 

Regardless of the modality, it’s clear that technology is giv­ing the health-care industry a much-needed upgrade. And the process will not slow down any time soon. As Hanson em­phasizes, “Genomics will be transformative.” With the new and more precise information, textbooks will be rewritten. Today, he continues, we use terms like “heart failure,” which is a broad way of characterizing medical conditions. In the fu­ture, “we’ll be able to be even more precise, using genomics to understand what causes cancer, etc., to be able to intervene much earlier. You need health-care IT and computers to do anything with genomics because the information is massive.” Advances in computer science and genomics: in Hanson’s view, “all interwoven.”