Four presentations at the workshop examined the broad issues that affect any effort to use technology to reduce health disparities. These issues range widely, from health literacy to community engagement, from the role of theory in program development to the key role of design in technology development. The overall picture that emerged from the presentations is of a complex but navigable path toward greater health equity.
“Why do we think digital strategies will make a difference in addressing health disparities?” asked Ivor Horn, medical director of the Center for Diversity and Health Equity and professor of pediatrics at the University of Washington School of Medicine. “There are so many interventions that have been developed over the years, and yet health disparities persist in racial and ethnic minority communities across the country. How is this different from the behavioral interventions that we have done?” she asked.
First, she said, mobile technologies are different than other innovations in health care. As observed in the keynote addresses (see Chapter 2), racial and ethnic minorities have been early adopters of mobile technologies. With smartphones in particular, racial and ethnic minorities have overtaken whites in their use of the technology.
Second, racial and ethnic minorities are leaders in the online community. For example, African Americans spend more than twice as much time per person on Web-hosting sites compared with the average of all consumers. “Racial and ethnic minorities are not just consumers,” said Horn, “They are actually producers of information,” which translates into an ability to influence the content of the discussion in mainstream media.
Third, racial and ethnic minorities have a clear interest in using digital health. In a recent survey done by Horn and her colleagues of African American parents in an urban clinic, 90 percent said that they want to receive health information online. Other studies have shown that solid majorities of parents in urban pediatric clinics use smartphones as their primary way of getting online, Horn said, explaining, “Not only are they there, not only are they making a difference, but they are really interested in having a health discussion.”
Despite the tremendous opportunity to use digital strategies to address health disparities, progress has been slow. One major problem, Horn observed, is that there is a disconnection between where digital strategies are developed and where the needs are. “The motivation for the development of digital strategies is primarily driven by financial gain rather than improving health outcomes. As a result, there is very little motivation to commit to the sometimes lengthy process . . . of developing an evidence base to determine which tools move the needle. Without evidence, providers are not going to adopt them,” she explained.
Another barrier to progress is that the effort to reduce health disparities does not present a common enemy against which multiple stakeholders can rally. Action against climate change faces a similar obstacle, Horn noted. People can take many different approaches to address climate change, from driving an electric car to recycling and composting, but the policy framework and infrastructure do not exist to focus and magnify these efforts.
New technologies are driving a shift in communication, said Horn. In health care, providers are trained to communicate with each other. Now, technology is forcing providers to learn how to communicate with patients and with the community in ways they have not done before. She explained:
To do this, we have to go to the patient and to the community to understand what the need is, because now they are communicating back to us. We have not necessarily been trained to listen. We have been trained to hear, but we do not listen. We hear what we want to hear through our lens of what we are trained to do and to put it into our differential diagnosis, but then we do not hear the relevant thing that we need to hear because we have already moved on while they are still communicating. We need to start figuring out how to make that paradigm shift.
Digital health strategies need to be grounded in the appropriate infrastructure and policies if they are to be effective, Horn said. Also, to be sustainable, these strategies need to be connected to grassroots efforts, organizations, and infrastructure. “Rosa Parks was not just a seamstress who sat down on the bus and decided she did not want to get up,” said Horn. “She was part of a strategic grassroots movement connected to a plan to influence policy and make changes.”
Digital health strategies present a unique opportunity to address health disparities, but “we need to leverage the trusted leadership of online racial and ethnic minority communities to draw attention to the injustice of health inequities,” Horn stated. Doing so would create a call to action and enable people to become learners in the context of a community where leaders already exist.
As an example, Horn described a series of focus groups on technology use done with sickle cell patients who were transitioning out of pediatric care facilities into adult care facilities. When asked how they already were using technology in their day-to-day lives, focus group members said they were, for example, texting their parents to get lists of medications, or using games to distract themselves when they were having a pain crisis. “Whatever you create needs to be hackable,” said Horn, “meaning that the patient who has it needs to be able to make it their own.” Horn also works with children as young as 2 years old who are creating digital narratives using a tablet. “They are teaching us how to utilize that tool if we listen, if we look, if we pay attention,” she said.
In taking on this kind of work, several important questions need to be answered, Horn noted. Where are the communities online? Who are the leaders in those communities? How can the health care community work with those leaders to ensure the information people receive is the right information?
Partnering with racial and ethnic minority developers needs to be
paired with outreach to offline organizations and systems, such as churches and community organizations that can push for better access to health care, Horn said. The Affordable Care Act has increased coverage in particular communities, such as the Hispanic community and states that are offering increased coverage. “We need to encourage patients and communities to drive and push for that and fight for that. Without these tangible, sustainable movements for change, we will just be wearing more stylish Google glasses and quantifying ourselves while people continue to prematurely die all around us,” she concluded.
“Evidence, as important as it is, is not enough,” said Misha Pavel, professor of practice jointly appointed between the College of Computer and Information Sciences and the Bouvé College of Health Sciences at Northeastern University. “It is not enough to know that something worked or it did not work. We need to know why, because if we don’t, we don’t know what to change,” he said.
Technology has the potential to create important changes in behavior over the next decade, and, as Pavel said, “Behaviors are killing us.” At least 40 percent of premature mortality is the result of people’s behaviors. The problem is that behaviors are hard to change. Sexual behaviors, alcohol and drug use, smoking, a sedentary lifestyle, and even sleeping patterns—which are closely related to stress and its negative consequences—are deeply engrained behaviors.
Human behaviors produce health states and biological indicators that can be measured. However, a connection needs to be drawn between what is being measured and what is of concern, Pavel stated. For example, blood pressure is of interest for what it signifies about the health of the cardiovascular system. “We need to have some way of transforming what we measure to what we want to know,” said Pavel, “then we can close the loop and intervene in an optimal way.”
The proliferation of mobile technologies can help greatly both with the measures of health indicators and with interventions. Indicators that can be measured range all the way from glucose levels to a person’s mood. Mobile technologies also can sense the context a person is in and initiate questions or interventions at the appropriate time. For example, they can remind someone to take a medicine or make a measurement when that person is in the right circumstance.
However, the optimal use of technologies requires having a good theory about why and where they work. A good theory “is going to not only drive your design, but also the questions that you ask,” said Pavel.
Even when a theory is wrong, demonstrating that it is wrong can result in progress.
Mobile technologies can be used to do much more than has been done to date, Pavel observed. For example, he described a randomized controlled trial on the use of mobile technologies to change people’s weight. One group got a traditional weight control intervention, while another group received an intervention that used mobile technologies. “What we do is build these two process models to characterize outreach performance. Once you have a model, then you can develop the most optimal intervention that maintains the weight,” he explained.
A major conclusion was that the use of mobile technologies worked for some people and not for others. The challenge thus has evolved from understanding the behavior of groups to modeling the behaviors of individuals to be able to predict which individuals will benefit from an intervention. Pavel explained, “What we need to do is model [the behavior of] individuals and be able to predict what they will do.”
Older people can pose unique problems, Pavel noted in response to a question, because many of them “approach technology with suspicion.” For example, one of older patients’ highest concerns is cognitive decline. In response, “We use games and cognitive games that enable them to practice various cognitive skills,” he noted. Also, using Skype, older patients can find each other and build communities, adding “We would have never been able to anticipate that.”
In response to another comment, Pavel pointed to the difficulty of keeping up with technology. A study done of a technology, like personal digital assistants, can be obsolete by the time the study is published because the technology is no longer being used. At the same time, mobile technology provides very rapid access to information and data. “We can collect data in a way we have never been able to do before. As a result, we are beginning to think about randomized controlled trials and evidence gathering as a much more dynamic process. . . . But there is also always going to be a bit of a gap,” he noted.
A fundamental need, Pavel concluded, is for a vision of what health care will look like in 20 years. Then steps can be taken to help the system evolve toward that vision. Developing the basic principles that will drive progress entails making difficult social and economic decisions, but “we have to face it,” he concluded.
“How do you engage providers and patients in digital strategies to promote health equity and reduce disparities?” asked Ruth Parker, professor of medicine, pediatrics, and public health at the Emory University School of
Medicine. The research on this question is nascent, Parker noted, and many questions need to be answered, including the following:
- Which digital strategies would do this?
- What are patient-centered, health-literate digital strategies?
- What is known about the value of strategies, in terms of both quality and cost, as the affordability of health care delivery is addressed?
- How do trust, civility, and social impact—for example, the effects of digital technologies on the environment—come into play?
Parker focused mostly on issues of health literacy during her talk, since, as the Institute of Medicine has observed, “Efforts to improve quality, to reduce costs, and to reduce disparities cannot proceed without simultaneous improvements in health literacy” (IOM, 2004, p. xiv). In an effort to define health literacy, a group of authors in the discussion paper Ten Attributes of Health Literate Health Care Organizations identified the following qualities (Brach et al., 2012):
- Has leadership that makes health literacy integral to its mission, structure, and operations.
- Integrates health literacy into planning, evaluation measures, patient safety, and quality improvement.
- Prepares the workforce to be health literate and monitors progress.
- Includes populations served in the design, implementation, and evaluation of health information and services.
- Meets the needs of populations with a range of health literacy skills while avoiding stigmatization.
- Uses health literacy strategies in interpersonal communications, and confirms understanding at all points of contact.
- Provides easy access to health information and services and navigation assistance.
- Designs and distributes print, audiovisual, and social media content that is easy to understand and act on.
- Addresses health literacy in high-risk situations, including care transitions and communications about medicines.
- Communicates clearly what health plans cover and what individuals will have to pay for services.
Many of these attributes have direct parallels with the health literacy of individuals. As an example, Parker cited the final attribute on coverage and cost. “A lot of the transformation of health care organizations will
come with cost transparency,” she said, because people do not want to waste money.
The main lesson that Parker drew from this list is that patients are the real experts. As one of Parker’s patients once told her, “I know what I know. I also know more than I think I do.” Therefore, to work effectively with patients, the health care system will need to partner with them, Parker explained. Health care systems also need to find ways to engage patients more effectively, she said later in response to a question. “I have never heard anybody say, ‘I get to go to the doctor today.’ It’s ‘I have to go.’ Anything we can do to try to make it engaging, to try to make it more fun, that is a good thing,” she said.
She described several projects in which partnerships and engagement led to important advances. During the surge of H1N1 cases in the United States, she and her colleagues worked with the Centers for Disease Control and Prevention to develop an online self-triaging flu website. With input from experts and from interviews with potential users, the developers put together a list of questions that would help someone know whether to go to an emergency room or stay home and treat their aches and fever. “Everybody on my team got the flu, by the way, [but] I feel that you really do have to work with the target population to figure out what is going on,” she explained. These conversations require learning to listen. Listening to people “is the hardest part of communication,” Parker said, “because we do not always hear what we want to hear.”
Another example was text4baby, in which the major mobile carriers in the United States waived the fees for text messages sent to pregnant women and mothers with babies in the first year of life. About 650,000 women signed up to use the system, which is still ongoing. To develop the messages, Parker and her colleagues went into the field and tested the messages with potential users. For example, the original message, “The flu can be dangerous for pregnant women & their babies. Talk to your doctor about seasonal flu & H1N1 flu shots,” was changed on the basis of this input to “If you get the flu while you are pregnant, you and your baby can get very sick. Ask your doc if you need a flu shot.” Besides changing the message from passive to active voice, the new message provided a more specific action step by asking about getting a flu shot. “We tested all these early messages to get input from the ground before they were launched,” said Parker.
Evaluations of the text4baby project have found that text messaging is more common among the high-risk population than among other populations (Gazmararian et al., 2012, 2014; Poorman et al., 2014). In addition, successful enrollment was more likely with higher health literacy.
However, these evaluations also identified several major issues that need to be addressed, including a lack of good theories of behavior change, small sample sizes, high drop-off rates, missing control groups, reliance
on self-reports, and a failure to tie outcomes to content. Future research, said Parker, needs to look at such issues as navigating mHealth initiatives despite lower health literacy, the use of prepaid phone cards, the delivery of supplemental information (such as videos, icons, or verbal narratives), better enrollment strategies, the effect on behaviors, and the relative effect of the program at different literacy levels.
Parker also discussed a study of Internet and mobile phone use among older Americans. In a study of about 430 subjects drawn from an academic medical center and four community primary care clinics, a significantly larger fraction of those patients with adequate health literacy had Internet access in their home compared to those with limited health literacy. The same is true for access to the Internet anywhere other than the home. These findings provide many opportunities for continued exploration, Parker noted.
Summing up, Parker pointed to several new directions for research and interventions that could help reduce health disparities. Both high-tech and low-tech solutions need to be evaluated, she said. Patient input and engagement need to be maximized to reflect patients’ needs, she added, while more practical metrics of technology use and its impact are developed. Health care providers also need to be educated about patient-centered, health-literate digital strategies, she emphasized, with identification of best practices and the competencies that providers should have. “I have never found a health officer or a medical student who regularly advises patients on where to go for good evidence-based and useful information on the Internet or on their phone,” she said.
During the discussion session, Parker pointed briefly to some of the privacy issues associated with digital technologies. For example, her elderly patients often tell her that they would like to send and receive photographs of their medications to know what they should be taking. But health systems are concerned that doing so will raise privacy concerns. Photographing medications and taking pictures of clinical wounds or conditions such as wounds or rashes is “a hot zone for action,” she said, “because patients like it and it is user friendly. But systems do not know what to do with that.”
She concluded by quoting Kelly Johnson’s advice to “keep it simple, stupid.” “This is still the hallmark of great design,” she said. “Simplicity should be the key goal.”
When the World Wide Web first became widely available in the 1990s, Cameron Norman, principal of CENSE Research + Design and an adjunct professor in the Dalla Lana School of Public Health at the University of Toronto, began working on eHealth literacy, which he defined as “the ability to seek, find, understand, and appraise health information from
electronic sources and apply the knowledge gained to addressing or solving a health problem” (Norman and Skinner, 2006, p. e9). But this was early in the history of the Web, he added. Since then, his views have evolved. In a 2011 paper, he wrote:
Just as the field of eHealth is dynamic and evolving, so too is the context where eHealth literacy is applied and understood. The original [model] of eHealth literacy and the scale used to assess it were developed at a time when the first generation of Web tools gained prominence before the rise of social media. The rapid shifts in the informational landscape created by Web 2.0 tools and environments suggests it might be time to revisit the concept of eHealth literacy and consider what a second release might look like. (Norman, 2011, p. e125)
The Web originally was designed to communicate information to the world. In the health context, the idea was that framing and delivering information the right way would get people to change their behavior. But technologies and platforms such as mobile computing, interactive video, Facebook, and Twitter have changed the Web, Norman said. “Now we realize that it is much more about being able to create part of the conversation and lead the conversation.” Many interesting conversations happening today are not being led by experts but by members of the public. The one-way push of information remains part of the picture, but it is just a small part of a much more complicated world.
In an earlier era, communicating was about knowing an audience and addressing it directly. The problem with that view is that addressing an audience requires accepting at the outset the basic premises that unite the audience. “Artists do not address themselves to audiences,” Norman said, citing the work of the anthropologist Edmund Snow Carpenter. “They create audiences. The artist talks to himself out loud. If what he has to say is significant, others hear and are affected,” he explained.
In the context of health disparities, Norman said that digital media now provide the ability for people to “take ownership of the issues and say, ‘If others are not going to do it for us, we will do it for ourselves. We will create our own audiences.’” This is a powerful possibility, said Norman, but it also presents challenges.
First, the volume of content that these technologies can provide is overwhelming. “Does anybody in the room here think that they do not have enough e-mail?” he asked. And e-mail is just one channel. Text messages, Facebook posts, tweets, and other electronic intrusions all compete for one’s time.
Second, the systems created around information technologies frame the context of information. Norman quoted the historian of technology Melvin
Kranzberg, who established as his first law of technology that “technology is neither good nor bad; nor is it neutral” (Kranzberg, 1995, p. 5). People are expected to make complex health decisions using sophisticated information that is coming at them at light speed. As people are overwhelmed by the amount and complexity of information they receive, they can be expected to question some of the systems enabled by digital technologies. People need to think about what they are gaining and what they are giving up through their use of technology, Norman observed. “What are we hiring this technology to do?” he asked.
The systems created around technologies also have distinct drawbacks. An example is the ad content and other distractions, such as click bait, transmitted with electronic information to support those communications. Norman explained that “your life is time. The more time you are spending on this, [the more] it is part of your life. But is that the life you want?”
Finally, technology is inserting itself into parts of people’s lives that were once technology free, “like walking a dog, or walking across the street with a friend, or spending time with your friends or loved ones, or having lunch, or going on a trip and visiting the sites, or exercising,” he noted. New technologies can lock people into systems, Norman stated, and people can no longer think about systemic structures when they are locked into those systems. Health disparities originate in systemic structures and patterns of behaviors, said Norman. These structures include how workplaces, communities, and social systems are set up. Technology has the potential to change these structures and patterns of behavior, he said, “but only if we are thinking about it in those terms.”
The only way to change systemic problems is to think and act systemically, Norman explained. That requires individuals making individual choices collectively in collaboration with one another. “We need to be thinking about what are individuals doing, but we cannot do that divorced from the idea of what happens within the networks in which people interact,” he noted. What is important is to pay attention to what is adding value, Norman said. People may use a technology in a way that was not planned but that nevertheless adds value to their lives. Also, technologies need to fit into the fabric of everyday life and not be simply an add-on if their use is to be sustainable.
Systemic change also involves the concept of design, said Norman, which he views as a way of creating new things. As Nobel laureate Herbert Simon once observed, “A designer is anyone who devises a course of action that seeks to change current situations into preferred ones” (Simon, 1981, p. 129). Technologies can be designed for the collective well-being, or they can continue to focus people on events. “We need to be clear about our intentions,” said Norman, because design is about “making intent come alive.”
He closed with an observation that is attributed to Paul Batalden, Professor Emeritus of the Geisel School of Medicine at Dartmouth: “Every system is perfectly designed to get the results it gets.” To change the system, he said, “We have to listen, and then we have to lead.”
This page intentionally left blank.