eBriefing

Health 2.0: Digital Technology in Clinical Care

Health 2.0: gital Technology in Clinical Care
Reported by
Michael Linde

Posted May 10, 2013

Overview

Digital technology has started to transform health care, impacting how patients and providers interact, care is delivered, and medical professionals are trained. New mobile applications, assistive technologies, education tools, and supercomputer-based physician decision support programs promise to facilitate high-quality, value-based health care that is also more accessible. But the use of digital technology in health care settings is a new development, and providers should consider how it will be implemented. At the March 22, 2013, Health 2.0: Digital Technology in Clinical Care conference, sponsored by the New York State Department of Health AIDS Institute, the Josiah Macy Jr. Foundation, and the New York Academy of Sciences, stakeholders convened at the Academy to discuss the technical, legal, and ethical implications of health care technology innovation. As part of a NYSDOH AIDS Institute initiative to examine the use of technology in HIV services in New York, the meeting also focused on specific digital advances in HIV/AIDS care.

Use the tabs above to find a meeting report and multimedia from this event.

 

Presentations available from:
Jessica S. Ancker, PhD (Weill Medical College of Cornell University)
Barbara Barry, PhD (Northeastern University)
Herbert Chase, MD (Columbia University)
Curtis M. Coomes, JD (RTI International)
Michael C. Gibbons, MD (Johns Hopkins Urban Health Institute)
Joseph C. Kvedar, MD (Center for Connected Health, Partners Healthcare)
Debra A. Lieberman, PhD (University of California, Santa Barbara)
Roberto Martinez, MD (New York State Department of Health)
Deven McGraw, JD (Center for Democracy & Technology)
John O. Moore, MD (Massachusetts Institute of Technology, MIT Media Lab)
Jean-Luc Neptune, MD (Health 2.0 LLC)
William R. Rodriguez, MD (Daktari Diagnostics Inc.)
Nirav R. Shah, MD (New York State Department of Health)
Iana Simeonov (University of California, San Francisco School of Medicine)
Marc M. Triola, MD (New York University School of Medicine)
Jennifer D. Uhrig, PhD (RTI International)


Presented by

  • Josiah Macy Jr. Foundation
  • The New York Academy of Sciences

Josiah Macy Jr. Foundation
This meeting is part of our Translational Medicine Initiative, sponsored by the Josiah Macy Jr. Foundation.

 

Connected Health: Empowering Patients and Providers


Joseph C. Kvedar (Center for Connected Health, Partners Healthcare)
  • 00:01
    1. Introduction
  • 03:18
    2. The Center for Connected Health; Connected health components
  • 10:10
    3. Connected cardiac care; Diabetes and blood pressure; Medication adherence
  • 16:56
    4. Texting and messaging; Genetic risk; Summary and conclusio

A New Wave in Patient-centered Care: Apprenticeship in the Management of Chronic Disease


John O. Moore (MIT Media Lab)
  • 00:01
    1. Introduction; Patient virtual care platform
  • 08:47
    2. Asynchronous interaction; A health ecosystem; Scaffolding
  • 16:05
    3. Open source; Going forward; Conclusion

Relational Agents: Digital Coaches and Caregivers


Barbara Barry (Northeastern University)
  • 00:01
    1. Introduction
  • 02:07
    2. Relational agents and their use in healthcare
  • 04:36
    3. Agent demonstration; Current projects
  • 12:00
    4. Personalization; Agents for medication adherence and substance abuse screenings
  • 17:44
    5. Forthcoming projects; Conclusio

Panel Discussion: Personalized, Collaborative Medicine


  • 00:01
    1. Avatars and mental illness; Reshaping the system
  • 09:50
    2. Health coaching; Adherence beyond bottles; Health education
  • 19:34
    3. Family studies; Conclusio

Using Digital Technology to Reduce Healthcare Disparities


Michael C. Gibbons (Johns Hopkins Urban Health Institute)
  • 00:01
    1. Introduction; A communication crisis
  • 08:23
    2. Technology opportunties; Health IT and social media
  • 17:31
    3. Patient, consumer, and caregiver decision support; Big data
  • 19:33
    4. Other challenges and new opportunities; Conclusio

Privacy, Security, and Professionalism in Electronic Communications


Deven McGraw (Center for Democracy & Technology)
  • 00:01
    1. Introduction; Privacy and security considerations
  • 05:34
    2. Provider to patient communications; ePHI; Security rule and transmissions to patients
  • 14:00
    3. Professionalism; Tips for using social media; Conclusio

Clinical Research Apps and the Next Generation of Wearable Sensors


Iana Simeonov (University of California, San Francisco School of Medicine)
  • 00:01
    1. Introduction; The UCSF mHealth group
  • 05:17
    2. Care on any mobile device; Mobile application process; User-centered design
  • 09:49
    3. Learning from advertising; UCSF apps; Examples
  • 19:00
    4. Personalizing the app; Conclusio

Evaluation in Digital Health: New Challenges, New Techniques


Jessica S. Ancker (Weill Medical College of Cornell University)
  • 00:01
    1. Introduction; Standard approaches to intervention
  • 07:05
    2. Some leaner approaches; Novelty effects; Generalizability
  • 15:12
    3. Public health impact; Integration with health IT systems; Conclusio

Conversation: Promoting Translation and Expanded Use of Digital Technology


Moderator: Jean-Luc Neptune (Health 2.0 LLC)
  • 00:01
    1. Introductions; What medical professionals need
  • 12:32
    2. Developers vs. designers; User input during development; Multifactorial issues
  • 22:24
    3. Evaluating technologies; Parsing big data; Conclusio

Point-of-Care Diagnostic Devices for HIV


William R. Rodriguez (Daktari Diagnostics Inc.)
  • 00:01
    1. Introduction
  • 05:37
    2. HIV and mobile phones in Africa; Need-driven change; Progress in CD4 counting
  • 10:28
    3. The commercial future; Global markets; Connectivity; Conclusio

Tailored Text Messages to Promote Knowledge, Prevention, Social Support, and Medication Adherence for People Living with HIV


Jennifer D. Uhrig and Curtis M. Coomes (RTI International)
  • 00:01
    1. Introduction; SMS-based intervention study
  • 05:55
    2. Evaluation design; Participant receptivity; Outcomes
  • 14:40
    3. Limitations; Future directions; The UCARE4LIFE program; Conclusio

Game Changer: Using Digital Games to Motivate Patient Behavior Change and Support Clinical Care


Debra A. Lieberman (University of California, Santa Barbara)
  • 00:01
    1. Introduction and overview
  • 03:00
    2. Benefits of gaming
  • 08:30
    3. Social aspects; Mobile devices and sensors; Technological options
  • 14:28
    4. Characters and avatars; Health games; Study results
  • 20:21
    5. Storytelling; Forthcoming projects; Conclusio

The Virtual, BioDigital Patient and Other Innovative Technologies for Medical Education


Marc M. Triola (New York University School of Medicine)
  • 00:01
    1. Introduction; Curriculum for the 21st century
  • 06:34
    2. Educational technologies; The emerging ecosystem; Biodigital human
  • 14:10
    3. Education data warehouse; Curriculum mapping; The future; Conclusio

The Emerging Role of Artificial Intelligence in Medical Decision-making


Herbert Chase (Columbia University)
  • 00:01
    1. Introduction
  • 04:35
    2. Cognitive levels; Diagnostic systems in action; Clinical reasoning bias
  • 13:00
    3. Prioritization; Analysis and judgment
  • 21:01
    4. Summary and conclusio

Introduction


Nirav R. Shah (New York State Department of Health)

Resources

Transforming provider roles and health care delivery

McNair S, Checchi K, Rubin A, et al. A pilot study of a computer-based relational agent to screen for substance-use problems in primary care. Society of Behavioral Medicine 34th Annual Meeting and Scientific Sessions. San Francisco, CA. March 20-23, 2013. 58-59.

Moore JO, Hardy H, Skolnik PR, Moss FH. A collaborative awareness system for chronic disease medication adherence applied to HIV infection. Conf Proc IEEE Eng Med Biol Soc. 2011;2011:1523-1527.

Moore JO, Boyer EW, Safren S, et al. Designing interventions to overcome poor numeracy and improve medication adherence in chronic illness, including HIV/AIDS. J Med Toxicol. 2011;7(2):133-138.

Partners Healthcare. Connected Cardiac Care.

Pfeifer L, Bickmore T. Longitudinal remote follow-up by intelligent conversational agents for post-hospitalization care. Association for the Advancement of Artificial Intelligence Spring Symposium on Artificial Intelligence in Health Communication. Palo Alto, CA. March 21-23, 2011.

Steventon A, Bardsley M, Billings J, et al. Effect of telehealth on use of secondary care and mortality: findings from the Whole System Demonstrator cluster randomized trial. BMJ. 2012;344:e3874.

Equity and ethics

Federation of State Medical Boards. Model policy guidelines for the appropriate use of social media and social networking in medical practice. Euless, TX. 2012.

Gibbons MC, Casale CR. Reducing disparities in health care quality: the role of health IT in underresourced settings. Med Care Res Rev. 2010;67(5 Suppl):155S-162S.

Gibbons MC, Fleisher L, Slamon RE, et al. Exploring the potential of Web 2.0 to address health disparities. J Health Commun. 2011;16 Suppl 1:77-89.

Hive Strategies. Eight steps to launch a successful social media strategy (A guide for health care). 2011.

Icahn School of Medicine at Mount Sinai. Mount Sinai Medical Center Social Media Guideline. 2013.

Peek ME, Quinn MT, Gorawara-Bhat R, at al. How is shared decision-making defined among African-Americans with diabetes? Patient Educ Couns. 2008;72(3):450-458.

Rose LE, Kim MT, Dennison CR, Hill MN. The contexts of adherence for African Americans with high blood pressure. J Adv Nurs. 2000;32(3):587-594.

Shin HB, Kominski RA. Language use in the United States: 2007. American Community Survey Reports, ACS-12. U.S. Census Bureau. Washington, DC. 2010.

Research and evaluation at digital speed

Ancker JS, Kern LM, Abramson E, Kaushal R. The triangle model for evaluating the effect of health information technology on healthcare quality and safety. J Am Med Inform Assoc. 2012;19(1):61-65.

Dallery J, Cassidy RN, Raiff BR. Single-case experimental designs to evaluate novel technology-based interventions. J Med Internet Res. 2013;15(2):e22.

Glasgow RE, McKay HG, Piette JD, Reynolds KD. The RE-AIM framework for evaluating interventions: What can it tell us about approaches to chronic illness management? Patient Educ Couns. 2001;44(2):119-127.

University of California, San Francisco. mHealth.

Tools that change point of care and point of view

Coomes C, Lewis MA, Uhrig JD, et al. Beyond reminders: a conceptual framework for using SMS to promote prevention and improve health care quality and outcomes for patients living with HIV. AIDS Care. 2012;24(3):348-357.

Furberg RD, Uhrig JD, Bann CM, et al. Technical implementation of a multi-component, text message-based intervention for persons living with HIV. J Med Internet Res Prot. 2012;1(2):e17.

Harris JL, Furberg RD, Martin N, et al. Implementing an SMS-based intervention for persons living with HIV. J Pub Health Manag Pract. 2013;19(2):E9-E16.

Lewis MA, Uhrig JD, Bann CM, et al. Tailored text messaging intervention for HIV adherence: a proof-of-concept study. Health Psychol. 2013;32(3):248-253.

Lieberman DA. Video games for diabetes self-management: examples and design strategies. J Diabetes Sci Technol. 2012;6(4):802-6.

Lieberman DA. Management of chronic pediatric diseases with interactive health games: theory and research findings. J Ambul Care Manage. 2001;24(1):26-38.

Rodriguez WR, Christodoulides N, Floriano PN, et al. A microchip CD4 counting method for HIV monitoring in resource-poor settings. PLoS Med. 2005;2(7):e182.

Uhrig JD, Harris J, Furberg R, et al. Communication-focused technologies: health messages for HIV-positive men who have sex with men;—final report. AHRQ Publication No. 11-0063-EF. Rockville, MD: Agency for Healthcare Research and Quality. 2011.

Uhrig JD, Lewis MA, Bann CM, et al. Addressing HIV knowledge, risk reduction, social support, and patient involvement using SMS: results of a proof-of-concept study. J Health Comm. 2012;17(Sup1):128-145.

Medical education and clinical decision making

Bowen JL. Educational strategies to promote clinical diagnostic reasoning. N Engl J Med. 2006;355(21):2217-25.

Li Y, Salmasian H, Harpaz R, et al. Determining the reasons for medication prescriptions in the EHR using knowledge and natural language processing. AMIA Annu Symp Proc. 2011;2011:768-76.

Sondhi P, Sun J, Zhai C, et al. Leveraging medical thesauri and physician feedback for improving medical literature retrieval for case queries. J Am Med Inform Assoc. 2012;19(5):851-858.

Triola MM, Pusic MV. The education data warehouse: a transformative tool for health education research. J Grad Med Educ. 2012;4(1):113-5.

Triola MM, Friedman E, Cimino C, et al. Health information technology and the medical school curriculum. Am J Manag Care. 2010;16(12 Suppl HIT):SP54-6.

Wang X, Chase H, Markatou M, et al. Selecting information in electronic health records for knowledge acquisition. J Biomed Inform. 2010;43(4):595-601.

Organizers

Johanne Morne

New York State Department of Health AIDS Institute
e-mail | website

Sonja Noring

New York State Department of Health AIDS Institute
e-mail | website | publications

Cheryl Smith, MD

New York State Department of Health AIDS Institute
website | publications

AIDS Institute Social Media Workgroup

New York State Department of Health AIDS Institute

Brooke Grindlinger, PhD

The New York Academy of Sciences
e-mail

Kerstin Hofmeyer, PhD

The New York Academy of Sciences
e-mail


Speakers

Jessica S. Ancker, PhD

Weill Medical College of Cornell University
e-mail | website | publications

Barbara Barry, PhD

Northeastern University
e-mail | website

Herbert Chase, MD

Columbia University
e-mail | website | publications

Curtis M. Coomes, JD

RTI International
e-mail | website | publications

Humberto Cruz

New York State Department of Health AIDS Institute
e-mail | website | publications

Michael C. Gibbons, MD

Johns Hopkins Urban Health Institute
e-mail | website | publications

Miguel Gomez

U.S. Department of Health and Human Services; AIDS.gov
e-mail | website | publications

Martin S. Kohn, MD

IBM Thomas J. Watson Research Center
e-mail | website | publications

Joseph C. Kvedar, MD

Center for Connected Health, Partners Healthcare
e-mail | website | publications

Debra A. Lieberman, PhD

University of California, Santa Barbara
e-mail | website | publications

Roberto Martinez, MD

New York State Department of Health
e-mail | website

Deven McGraw, JD

Center for Democracy & Technology
e-mail | website | publications

John O. Moore, MD

Massachusetts Institute of Technology, MIT Media Lab
e-mail | website | publications

Jean-Luc Neptune, MD

Health 2.0 LLC
e-mail | website

William R. Rodriguez, MD

Daktari Diagnostics Inc.
e-mail | website | publications

Nirav R. Shah, MD

New York State Department of Health
website

Iana Simeonov

University of California, San Francisco School of Medicine
e-mail | website

Marc M. Triola, MD

New York University School of Medicine
e-mail | website | publications

Jennifer D. Uhrig, PhD

RTI International
e-mail | website | publications


Michael Linde

Michael Linde is a Denver-based medical writer. He specializes in HIV/AIDS, but has written about diverse medical and scientific topics, including health care delivery, oncology, diabetes, neurology, urology, end of life care, and immunology. His most recent article The Conserved Set of Host Proteins Incorporated into HIV-1 Virions Suggests a Common Egress Pathway in Multiple Cell Types appeared in the Journal of Proteome Research. He holds an MS is in biochemistry and molecular biology from University of Southern California and is a PhD candidate in immunology at Johns Hopkins University. He can be reached at mlinde@lindemedicalwriting.com.

Sponsors

Presented by

  • Josiah Macy Jr. Foundation
  • The New York Academy of Sciences

Video Address:
Nirav R. Shah, New York State Department of Health
 
Conference Chair:
Miguel Gomez, U.S. Department of Health and Human Services; AIDS.gov

Digital technology has transformed the way people interact throughout the world, but there remains a need to integrate digital advances into health care. Compared to other industries health care lags behind the digital boom, but this is beginning to change. New initiatives, collectively termed Health 2.0, implement technologies like mobile applications, digital medical education, electronic collaboration, information sharing, and decision support to improve patient care.

New York is a leading region in information and communication technology and health care innovation, said Nirav R. Shah from New York State Department of Health. In one program, in Rochester, use of electronic health records by emergency departments decreased admission rates by 6%. Initiatives like this are projected to save the state $52 million annually. Digital technology is also being implemented across the world, and large trials are underway to explore and validate eHealth tactics. In the UK, the Whole System Demonstrator (WSD) project, a 6000-subject trial, is designed to assess telemedicine in rural, suburban, and urban environments. Early results show a 45% decrease in mortality among participants, demonstrating the potential for digital interventions to improve patient outcomes.

However, there are significant barriers to the use of digital technologies in health care settings. Garnering participation is a challenge: in the WSD trial, 80% of patients declined to participate. Indeed, uptake of new technologies will be difficult to achieve if digital tools are not designed with user input; as Humberto Cruz from the New York State Department of Health AIDS Institute noted, "There is always a human being behind technology." Patients and providers should be prepared for a shift in the way health care is distributed, with an increased focus on team-based care delivered outside traditional settings. The hope is that this shift in the distribution of care will empower patients and providers.

The New York State Department of Health (NYSDOH) AIDS Institute is harnessing technology and social media to improve patient care. This shift is particularly important in HIV care because the patient base is expanding while the number of providers is shrinking. The institute has developed a strategic plan that promotes HIV prevention and care, integrating digital technologies to improve patient outcomes and reduce costs.

Digital technology platforms are also well suited for implementing personalized-medicine strategies. With access to patient data, providers can create treatment plans that are specific and focused. Because patient responses to digital technologies are varied, there is a push to develop a wide range of eHealth interventions and to make these interventions user-responsive. By developing flexible digital resources that react to user responses and usage habits, developers hope to increase uptake and to personalize interventions. Similarly, the variety of interventions that can be offered, including games, adherence support, point-of-care technologies, and relational agent computer programs, allows providers to create patient-focused eHealth programs.

As the health care industry has just begun to incorporate digital technology, providers should consider how it will be used to deliver care and facilitate education. There is also a need to examine the legal and ethical implications of new technologies, as well as to validate the effectiveness of digital tools. The Health 2.0: Digital Technology in Clinical Care conference examined how provider roles and health care delivery are changing, considered emerging questions of equity and ethics, and presented new research and evaluation strategies and digital tools for consumers and providers. It addressed how to achieve real-world impacts with eHealth advances, to transform health care from a volume-based to a value-based system.

Speakers:
Joseph C. Kvedar, Center for Connected Health, Partners Healthcare
John O. Moore, Massachusetts Institute of Technology, MIT Media Lab
Barbara Barry, Northeastern University

Highlights

  • Home-based digital technologies are changing the interaction between patients and providers by facilitating daily monitoring and as-needed interventions.
  • Mobile applications can promote self-maintenance and change how patients perceive and consume health care.
  • Tools such as relational agents, a computer agent designed to form long-term social and emotional relationships with users in face-to-face conversations, are being tested to aid in the transition from hospital to home care.

Connected health: empowering patients and providers

Joseph C. Kvedar from the Center for Connected Health at Partners Healthcare discussed technology that is moving health care from a volume-based hospital model to a value-based patient-centered model. Health information exchanges, patient-centered medical homes, and accountable-care organizations challenge providers to improve health care delivery. Kvedar noted that digital technology can facilitate a new model of care—just-in-time care (care where the patient is, when the patient needs it)—that is not intended to replace provider roles, but instead to extend care through technology.

At Partners Healthcare the Center for Connected Health uses digital technology to gather and report accurate patient data and improve outcomes. The information patients self-report can often be inaccurate because patients may tell their doctor or nurse what they want to hear, in order to please providers. Data garnered through digital technology is more accurate, can be reported in real-time, and can be aggregated to give a more complete and timely picture of a patient's condition. Data can also be presented back to the patient, in what Kvedar calls a feedback loop, to help educate and motivate individuals to make healthier lifestyle choices and better manage their health.

Motivational tools help to keep patients engaged in their health and wellness. In the Connected Cardiac Care Program (CCCP), a digital tool developed by the Center for Connected Health, heart failure patients use a home-based blood pressure cuff, oximeter, scale, and touch screen device to monitor and transmit their daily vital signs, which are monitored by a telemonitoring nurse. The program allows health care personnel to intervene before clinical events occur; for example, nurses call patients who do not check and report their vital signs at scheduled times, or when their vital signs are out of the parameters set by their provider. It is cost-effective, with a ratio of one nurse for every 100 patients, and has decreased hospitalization, with a 50% decline in recidivism. A similar program for patients with diabetes and hypertension lowered HbA1C levels and decreased blood pressure by 69%, respectively.

The Center for Connected Health has also implemented a program using electronic pill-bottle caps that indicate when medications should be taken, which was shown to increase adherence by 68%. The center is exploring the use of "wear and forget" vital-sign monitoring devices that will provide automatic feedback so treatment can be adjusted based on real-time data. By using digital technology to change the interaction between patients and providers, the center seeks to create in-the-moment interventions to prevent medical complications. The goal is also to increase patient engagement, improving both patient experience and outcomes.

Hospitalization rates for patients enrolled in the Partners Healthcare Connected Cardiac Care Program. (Image courtesy of Joseph C. Kvedar)

Partners Healthcare has also implemented a program using GlowCaps, electronic pill-bottle caps that indicate when medications should be taken, which has increased patient adherence by 68%. They are exploring the use of "wear and forget" vital-sign monitoring devices that will provide automatic feedback so treatment can be adjusted based on real-time data. By using digital technology to change the interaction between patients and providers, Partners Healthcare seeks to create in-the-moment interventions to prevent medical complications. The goal is also to increase patient engagement, improving both patient experience and outcomes.

A new wave in patient-centered care: apprenticeship in the management of chronic disease

Patient engagement is central for providers who take a patient-apprenticeship approach to health care, in which patients work with health coaches, providers, and family members to improve disease management. This model necessitates interacting with patients outside traditional health care settings. John O. Moore and his colleagues at Massachusetts Institute of Technology have designed tools to improve patient engagement and deliver actionable data. One example is a game that teaches patients about how non-adherence affects HIV mutation.

Other virtual tools developed at MIT accommodate bilateral control by both patients and providers. Bilateral control coupled with shared decision support allows patients to co-navigate health care and manage their own data. The apprenticeship model can also function as a societal intervention, changing the way people perceive health and health management, since most interactions occur outside traditional settings and involve peers and family members. As their knowledge and health management skills improve, patients can become health coaches and assist others.

Thus, digital applications can serve as scaffolding tools—scaffolding is a support technique that builds on prior knowledge to teach new concepts—and provide flexible platforms for patient–provider interaction. Digital platforms give providers more opportunities to interact with patients in real time. This model has been employed to intervene in crises, such as to control a diabetic attack, and researchers at MIT plan to adapt it to other chronic diseases and conditions, such as to control blood pressure and cholesterol.

Relational agents: digital coaches and caregivers

Technology is also changing the relationship between patients and providers. Project RED (Re-engineered Hospital Discharge), a program designed to improve the transition from hospital to home care, implemented a virtual nurse pilot program at Boston Medical Center. The virtual nurse is a type of technology called a relational agent—a computer agent designed to form long-term social and emotional relationships with users in face-to-face conversations. According to Barbara Barry from Northeastern University, relational agents can coach patients in behavioral change, deliver health information, collect data, and act as a scaffold for patients with low health literacy. They also serve as a connection between patients, providers, and caregivers.

The virtual transition nurse informs patients about important care transition steps. It employs a dynamic interface, with a script that changes based on the patient's reactions, logged through a series of touch screen-based questions. The touch screen interface has a low technology barrier for patients and garners instant feedback; it is also an accessible system that can be used to build patients' trust. A trial using the virtual transition nurse with approximately 750 patients, which examined patient satisfaction, uptake, and knowledge retention, found that the relational agent tool works well for older patients and those with low health literacy, and that it is preferred for interactions that involve sensitive topics, such as sexual health. Further studies are needed to demonstrate its efficacy.

The program is expanding to include a hospital buddy system that will accompany patients during their hospital stay. The relational agent will be equipped with sound and motion detectors to provide information about the environment, such as describing room alarms to the patient, and will have a built-in query list to answer patients' questions. It will also give providers information; for example, the hospital buddy can monitor the room environment and determine if the patient is sleeping well.

The hospital buddy relational agent. (Image courtesy of Barbara Barry)

These tools are still in development. Questions remain about privacy and optimization: How will personalization and uptake work for different disease states? How will these tools help patients living in unstable situations, who may not have access to the Internet? Will these tools be accessible for marginalized and low-income patients, who may not have access to smart phones and other digital technologies?

Speakers:
Michael C. Gibbons, Johns Hopkins Urban Health Institute
Deven McGraw, Center for Democracy & Technology

Highlights

  • Communication differences can be a significant barrier to care, leading to health disparities. Thus, new technologies should be designed to ease communication and ensure equitable use.
  • Providers must consider patient privacy when communicating with patients. HIPPA provides guidance on the use of protected information.
  • There is little governmental oversight of health applications, and new laws will likely lag behind technological innovations.

Harnessing digital technology to reduce health disparities

New eHealth technologies raise questions about ethics and equitability, said Michael C. Gibbons from Johns Hopkins Urban Health Institute. One potential benefit is to aid those with low health literacy. Many patients are unable to understand or be understood because of either low health literacy or poor English language fluency. In the U.S., there are approximately 23 million non-English speakers; while many patients speak Spanish or Chinese, a large proportion speak other languages, such as French, Korean, and Vietnamese. The number of patients with poor English language fluency is likely to increase; since 1980, there has been a greater than 200% increase in the number of people speaking Spanish, Russian, Persian, Chinese, Korean, or Tagalog in the U.S. The number of Vietnamese speakers has increased over 500% during this period. Jargon and dialects can also create communication barriers. There are over 25 million people in the U.S. whose English language fluency is poor, and the number is rapidly increasing. Communication problems affect patient trust, engagement, and care satisfaction.

Proportion of non-English speakers in the United States. (Image courtesy of Michael C. Gibbons)

According to Gibbons, the issue goes beyond patient understanding—it also impacts how providers understand their patients. Patients need to be heard: shared decision-making and informed health care choices cannot occur without a framework of mutual communication. In the absence of this framework, patients might make poor decisions, such as non-adherence, simply to assert control. Patient engagement and satisfaction can also impact metrics such as Healthcare Effectiveness Data and Information Set (HEDIS), and thus have significant cost and reimbursement impacts.

Health care providers need a variety of tools to tackle the growing communication crisis. Digital application developers should consider communication differences when building the user interface. Should a new digital tool be video-based or graphic-based? Will a video-based tool be used equitably? Will these tools be culturally appropriate? Electronic resources should be designed to increase equitable health care access.

Privacy, security, and professionalism in electronic communications

As digital health care tools are disseminated, privacy concerns arise. There is a need to craft a balanced and workable model to establish patient trust in data security. The Health Insurance Portability and Accountability Act (HIPAA) provides guidance on the use of protected health information, and applies to communications in paper and electronic forms. HIPAA mandates that providers supply secure electronic protected health information (ePHI). Notably, ePHI extends beyond clinical information; any communication from a health care provider to a patient could be considered ePHI. Regardless of whether a communication is classified as ePHI, professional and ethical obligations apply to its use.

Fortunately, HIPAA is a permissive law, according to Deven McGraw from the Center for Democracy & Technology. While HIPAA dictates which information providers can share and how to protect patient privacy, it also has optional components. Often, HIPAA does not expressly prohibit sharing a particular set of information, but organizations that are risk-averse may decide to comply with the optional guidelines and restrict information. Developers and users should determine which regulations are applicable. For example, a tweet or text message between a provider and patient may need to be documented in the medical record, but this depends on the content and context. Providers should always document cases in which they decide not to comply with optional HIPAA regulations.

In some scenarios, providers and developers may be permitted to disseminate information through less-secure communication channels. Patients have a legal right to receive communication in their preferred form, but should be informed if they request to receive information in a manner that is not secure.

McGraw noted that "the rules are always behind the technology," and that much of the legality concerning digital media is not yet defined. Therefore, developers need to be mindful of how their products might be regulated. McGraw also recommended that providers consider the potential implications before posting on social media or public websites. Disclaimers can be helpful in mitigating risks, but they may not always absolve providers and developers from responsibility. Developers should pay attention to privacy, considering both state and federal regulations, and take steps to protect health information.

FDA guidance on digital technologies. (Image courtesy of Deven McGraw)

Speakers:
Jessica S. Ancker, Weill Medical College of Cornell University
Iana Simeonov, University of California, San Francisco School of Medicine
Roberto Martinez, New York State Department of Health
 
Moderator:
Jean-Luc Neptune, Health 2.0 LLC
 

Highlights

  • New technologies must be validated in clinical trials.
  • Crossover or multiple-baseline trials may be more useful than traditional models.
  • User feedback should be built into the trial design to maximize utility.

Evaluation in digital health: new technologies, new challenges

Although many digital technologies promise to improve health outcomes, the benefits of developing new procedures need to be validated. As noted by Jessica S. Ancker from Weill Medical College of Cornell University, a digital advancement that promises to improve the quality of care may not be effective in all instances. Thus, new digital products must be clinically evaluated. With approximately 14 000 applications in the market, this is a daunting project.

Traditional methods to test interventions include observational and randomized clinical trials. Observational tools can provide insights from intervention testing; however, these types of studies have some drawbacks, such as self-selection bias. Randomized clinical trials can eliminate many biases, but these trials are time consuming and expensive. As a result, newer trial models are being developed to require fewer patients, take less time, and include internal controls.

One example is a crossover trial design, in which the control and the experimental groups switch after a period of time, and then may switch again. This creates both a control group and an internal control, allows for measurement of behavior changes over time, and requires fewer subjects. In a multiple baseline study, several groups receive the intervention for different periods of time; for example, one may receive the intervention for 4 weeks, while another receives it for 24 weeks. This design helps mitigate time trends, such as changes in cell phone plan costs or mobile operating systems.

Multiple-baseline trial design. (Image courtesy of Jessica S. Ancker)

An important consideration is that any intervention that includes monitoring will result in a change in subjects' behavior, a phenomenon known as the Hawthorne Effect. With digital technology, trial design should also rule out a novelty effect, in which an intervention may affect change for a while but become significantly less effective over time. Thus, improvements shown in shorter trials may not be observed in longer trials of the same intervention.

Instant research and clinical apps: the next generation of wearable sensing devices

Iana Simeonov from University of California, San Francisco School of Medicine described the university's new mobile health research and development program, which aims to build digital products that provide a cohesive experience across mobile health applications. The group builds applications—which extend beyond just phones and can include wearable sensing devices and other digital interfaces—on a framework that can be adapted for different diseases. Using information-management tools, the group aims to make patient data easy to parse, and thereby increase the usefulness of an application. Effective data management is especially important for the University of California, San Francisco community, as most applications are developed for research purposes, not for consumer use.

Development process schematic for user-centered design. (Image courtesy of Iana Simeonov)

Whatever the intended use of an application, its interface must be user-oriented. The department's applications have a feedback loop to evaluate usability. The intent is to develop dynamic tools, with features that are user-responsive and based on user feedback, to increase uptake and patient engagement.

The department's J2ME (Java 2, Micro Edition)-based framework allows the designers to move user-centered refinements from one application to another. For example, goal-setting features from an application for depression can "cross-pollinate" with a similar app designed for post-traumatic stress disorder, said Simeonov. Applications are made available in multiple languages, or without language, to suit various audiences. These features are also developed in conjunction with providers and are designed to be easy for them to use. Providers help designers to manage information standards, which will likely be critical when the applications are evaluated. As providers are increasingly mandated to use electronic records systems, ease of use is an important consideration.

The assessment and development of digital technologies must address impact and scalability. Determining the public-health impacts of new technologies goes beyond assessing efficacy in controlled clinical trials; the reach, adoption, implementation, and maintenance of the tools must also be assessed in clinical settings. The ease with which a digital tool can be integrated with other electronic resources may in fact determine its sustainability, as it is more difficult, and less likely, for providers to implement stand-alone technologies.

Promoting translation and expanding the use of digital technology

Digital technologies for health care should be developed to address real-world issues. One of the main concerns providers face today is how information should be shared, according to Roberto Martinez from the New York State Department of Health. This was the topic of a panel discussion led by Jean-Luc Neptune of Health 2.0 LLC, featuring Ancker and Martinez. The New York State Health Information Network allows providers to exchange health information with patients, researchers, and government public health agencies in order to make information accessible at the point of care.

Costs are another real-world concern. Many digital health care advances are driven by cost–savings incentives. As the federal government is the single-largest payer in the U.S., the Centers for Medicare and Medicaid Services is a major driver for the development and direction of health technologies. Thus, digital technologies such as electronic medical records must meet their criteria and, in order to do so, must have built-in interoperability.

Ease of use is perhaps the most important factor when it comes to encouraging uptake of new technologies. Technologies designed for providers are often not intuitive, noted Neptune. This may happen when designers—distinct from developers—are not involved in the early stages of development. Digital technologies must fit into and improve providers' workflows to be effective. Technologies designed for patients must consider the end user, particularly if the technology is designed for an older population.

Speakers:
William R. Rodriguez, Daktari Diagnostics Inc.
Jennifer D. Uhrig, RTI International
Curtis M. Coomes, RTI International
Debra A. Lieberman, University of California, Santa Barbara

Highlights

  • Digital tools have the potential to enable low-cost point-of-care diagnostics, such as HIV viral load and CD4+ T cell count testing.
  • Text messaging adherence and health reminders have demonstrated success in helping HIV-positive patients on antiretroviral therapy to maintain undetectable viral loads.
  • Heath-oriented games are a promising avenue to increase patient education, self-maintenance, and health-seeking behaviors.

Point-of-care diagnostic devices for HIV

Digital technology has the potential to change both health care delivery at the point of care and how people access health care. William R. Rodriguez from Daktari Diagnostics Inc. explained that this is especially relevant in locations where the traditional health care model is not sustainable, such as in resource-limited regions and areas with few providers. Need-driven change will spur technology advances and adoption in these regions.

The HIV/AIDS epidemic in sub-Saharan Africa is one example. Daktari Diagnostics Inc. is one of a number of companies developing technology to improve HIV/AIDS care in this region. In South Africa and other countries in sub-Saharan Africa, there is a need for HIV RNA and CD4 T-cell diagnostics, and these companies are developing small, inexpensive point-of-care diagnostic tools to meet patient and provider needs.

In addition to reducing the lag time between patient visit and test results, these diagnostic tools are also spurring changes in the way patient data is managed. For example, Kenya mandates that diagnostic tests have wireless capabilities to send results directly to patients, providers, and relevant databases. Providers will need to adjust to the idea that they are no longer needed as patient-data gatekeepers. Advances such as these will lead to more rapid interventions and less expensive—or even free—diagnostic test results.

Tailored text messages promote knowledge, prevention, social support, and adherence

Changing the point of care was the centerpiece of a Short Message Service (SMS) text-based intervention at RTI International, funded by the Agency for Healthcare Research & Quality (AHRQ). Curtis M. Coomes and Jennifer D. Uhrig, from RTI, described how the intervention was designed to remind HIV-positive patients to take antiretroviral medication and access care for comorbidities. The messages addressed medication adherence, sexual risk reduction, substance-use risk reduction, general health and well-being, social support, and patient involvement.

Rather than institute a static program, this pilot trial incorporated dynamic elements, such as changes in the timing, frequency, and interactivity of messages, personalized for each patient. The 46-subject intervention also included prevention-themed messages on topics such as sexual safety and substance-use risk reduction. Subjects underwent a baseline evaluation and survey; their answers were used to create the initial tailored text scripts for the program, which changed during the course of the study and became more personalized, based on subjects' reactions.

Sample text messages from the RTI International trial. (Image courtesy of Jennifer D. Uhrig and Curtis M. Coomes)

In addition to assessing whether the intervention impacted medical parameters such as viral load and CD4 T-cell counts, the trial also evaluated changes in targeted risk behaviors, social support, patient involvement, self-reported medication adherence, and knowledge, attitudes, and beliefs. The bidirectional communication program increased patients' knowledge, reported adherence, and perceived social support, and reduced risk behaviors. The overwhelming majority of subjects reported that the messages were easy to understand (98%) and gave good advice (82%) and that they trusted the information (89%). Notably, there was a significant decrease in viral load and a significant increase in CD4 T-cell count. While these results are encouraging, larger and longer trials are needed to assess the long-term impact of this type of program. RTI International is implementing and evaluating the UCARE4LIFE program, funded by the Health Resources and Services Administration (HRSA), which aims to increase primary care retention rates for racial and ethnic minority youth aged 15 to 24 living with HIV for at least 9 months. The randomized, controlled public–private partnership trial will develop, test, and maintain a text message library addressing HIV disease management.

Game changer: using digital games to motivate patient behavior change and support clinical care

Digital technologies also promise to change how people receive and perceive health information. Debra A. Lieberman from University of California, Santa Barbara discussed the potential for gaming to increase health literacy and self-maintenance. Media evoke emotional responses and cultivate our expectations of the real world. Because we learn by observation, action, and interaction, games offer an attractive rule-base medium to promote behavioral outcomes.

Studies have shown an increase in positive health-related behavior when digital games are adapted to health-related contexts. For example, seniors with access to cyber cycling—stationary bikes attached to screens that allow the rider to explore areas or compete—showed both physical and cognitive improvements. Other games target health education: in the Super Nintendo game Bronkie the Bronchiasaurus, players emulate real-life asthma self-maintenance behaviors to win the game.

The merger of health and games, if done successfully, may have a significant audience—health and games are two of the top three reasons people access the Internet. However, the games must not only serve as a springboard to affect positive health-behavior change: they must also be fun. Providers and patient can access a database of health-related games at the Health Games Research Database.

Disease states represented in the Health Games Research Database. (Image courtesy of Debra A. Lieberman)

Speakers:
Marc M. Triola, New York University School of Medicine
Herbert Chase, Columbia University
Martin S. Kohn, IBM Thomas J. Watson Research Center

Highlights

  • Medical education is incorporating digital technologies to transform and expand training methods.
  • Artificial intelligence programs are being developed as clinical decision support systems to aid in diagnosis and treatment.
  • The Watson supercomputer, which can rapidly process large amounts of language-based information, is being adapted to help with clinical decision support.

The virtual biodigital patient and other technologies for medical education

Marc M. Triola from New York University School of Medicine explained how digital technologies promise to change how providers learn about and practice medicine. Medical students in hospital-based curricula have traditionally had fragmented exposure to learning tools. But students now communicate, collaborate, and learn very differently. Providers have many sources of information and need to synthesize care in a team-based environment that may extend beyond the traditional setting. Medical schools need to change their teaching methods to prepare new practitioners for modern health care-delivery models.

Medical education in the 21st century includes an integrated curriculum of basic science and clinical education, with an increased focus on educational technology. New York University School of Medicine has changed its curriculum and now offers a three-year program designed for students to transfer directly into the NYU residency program in addition to the traditional four-year medical degree. The school also offers an extended five-year program in which students have the opportunity to concurrently pursue an MBA, MPA, MSCI, MPH, or bioethics Master's degree. The goal of this flexibility is to tailor education to students' varied career goals.

The school has also introduced educational technologies that uncouple the time and place of learning. To expand student access to learning tools, NYU has partnered with Biodigital Systems to give students access to a biodigital human, a high resolution 3D computer simulation that students can virtually examine and dissect and can use to study anatomy and procedures outside the laboratory. The biodigital human provides continuous access to a learning tool that was previously only accessible during a cadaver study period. This program is part of a larger NYU School of Medicine initiative to create an educational data warehouse, in which the curriculum is mapped to SNOMED and MeSH medical databases, allowing students to easily find educational tools.

The NYU School of Medicine educational data warehouse. (Image courtesy of Marc M. Triola)


 

The emerging role of artificial intelligence in medical decision making

Although medical education now incorporates new technologies to improve the learning process and medical data is managed through online platforms, there is still an unmanageable amount of information for medical professionals. According to Herbert Chase from Columbia University, there are approximately 13 000 diseases, 6000 drugs, and 4000 medical and surgical procedures that providers need to understand. In addition, one electronic health record can have over 10 000 data points. The enormous and ever-expanding amount of data available to providers can result in diagnostic errors, failure to follow common guidelines, and suboptimal treatment decisions. Coupled with the significant time demands already placed on providers, this excess of data necessitates development of clinical decision support systems to assist providers.

Artificial intelligence (AI) systems can help with the decision-making process. At a basic level, AI systems can search for drug–drug interactions, recognize dose adjustments, and retrieve relevant diagnosis and treatment information. They can also assist with diagnosis, either through active provider input or through automated searches of electronic health records. For example, clinical decision support systems could extract unstructured data from clinical notes and identify potential diagnoses.

Using clinical decision support to help correctly diagnose patients. (Image courtesy of Herbert Chase)

Digital clinical decision support has several other advantages. Importantly, it eliminates a number of human biases. Anchor bias occurs when the provider becomes attached to a particular diagnosis early in the decision-making process. Flaw of availability bias occurs when the provider makes a decision based on the information that he or she has available and not the total knowledge base. Immediate experience bias occurs when a provider leans toward a specific diagnosis because he or she has recently encountered it clinically. The support system can create decision trees to identify potential outcomes, find probabilities, prioritize decision choices, and modify decisions based on a patient's unique characteristics, such as comorbidities, polypharmacy, family history, and genetic profile. These systems can also assess patient progress and predict disease progression.

IBM's Watson joins the health care team: digital clinical decision support

Developers are using the Watson supercomputer to transform clinical decision support systems. Martin S. Kohn from the IBM Thomas J. Watson Research Center explained that structured data is relatively easy for artificial intelligence to process, but language is more difficult. The Watson supercomputer was designed as a knowledge-driven support system that processes language. The supercomputing system came into the public eye when it was featured as a contestant on Jeopardy!, competing (and winning) against two of the most successful contestants from the show. Because Jeopardy! uses complex language, the Watson team used the game show to test the computer's language-processing capacity. Watson, which has the ability to read 65 million pages per second, would be well suited for use as a clinical decision support system, as the medical literature also uses an arcane vocabulary. The Watson team is working with Memorial Sloan-Kettering Cancer Center and WellPoint to develop Watson as a clinical decision support system.

In addition to processing large amounts of language-based data, Watson also acts as a discovery tool: "it is the opposite of a search engine," explained Kohn, because it not only prioritizes results but also lists multiple decision options. The Watson reasoning process is valuable because it is unbiased and able to learn. Whereas humans find it difficult to process very large amounts of data without biases, Watson can manage this data and identify missing information. With responses weighted by likelihood, Watson's interactive process helps providers identify appropriate decisions. Thus, providers are responsible for the decision, but Watson helps them to sift through information and focus on making the correct decision or diagnosis. These capabilities represent only the beginning of clinical decision support systems in medicine.

Health care providers are increasingly using new technologies that are transforming the patient experience, including moving the point of care outside traditional health care settings. Patients will have more access and control over their information and health care, and providers will be able to use technology to manage data and deliver real-time care. Just as digital advances have changed many aspects of modern life, technological progress promises to transform how we give and receive health care.

How should existing technologies be validated? Which advances improve patient outcomes?

Will new technologies affect the cost of care? How will new technologies fit into the health care system?

How can barriers to uptake of technologies be overcome? How can we ensure equitable use of technology?

How can designers follow market trends and incorporate user input in the development process to maximize utility?

How should legal and ethical questions surrounding health care advances be evaluated? How should consumer applications be regulated? How can patient privacy be protected?