Commentary: Pediatric Digital Health Supported by the National Institutes of Health

A recent Pew report finds that 95% of teens (ages 13–17 years) report owning or having access to a smartphone, with 45% reporting being online on a “near-constant” daily basis (Pew Research Center, 2018). Digital health technologies have the potential to provide considerable benefits to child health and human development, especially when harnessed for targeted health behavior assessment and intervention, but the effects of considerable exposure to these technologies also raises concerns regarding child health and development (Hurst-Della Pietra, 2017).

The maturing field of research on pediatric eHealth and mHealth interventions addresses a wide range of issues (Wu, Steele, Connelly, Palermo, & Ritterband, 2014), and the National Institutes of Health (NIH) funds a diverse portfolio of basic, translational, and applied research across the lifespan. Using the search terms (eHealth, mHealth, “mobile health” or “digital health”) AND (children, adolescents, or pediatrics), the NIH RePORT (https://report.nih.gov) shows that 116 pediatric digital health grant projects or subprojects were funded by NIH in fiscal year (FY) 2018, totaling $65.12 million. These grant projects and subprojects were funded by 17 different institutes and centers. Expectedly, National Institute of Child Health and Human Development (NICHD) had the largest number funded in FY18 (19), but National Institute of Mental Health (NIMH), National Cancer Institute (NCI), National Institute on Drug Abuse (NIDA), and National Institute of Diabetes and Digestive and Kidney Diseases funded 10 or more pediatric digital health grants in FY18. Despite the interests of multiple NIH institutes and centers in pediatric digital health research, pediatric research represents only 18.8% of grant projects/subprojects and 20.8% of the total funding by NIH in digital health research for FY18.

Given the pervasive and persistent nature of digital technology use by children and adolescents, and the interests of multiple NIH institutes and centers in funding pediatric digital health research, this commentary encourages researchers to consider a broad range of pediatric digital health research grant applications relevant to the missions of various NIH institutes and centers. To do so, we first highlight some of the trends in pediatric digital health research based on recent research funded by the NIH. Second, we briefly describe some of the challenges and opportunities that NIH program staff and grant reviewers have considered in this area of research.

Highlights of NIH Funding in Pediatric Digital Health Technologies

Research involving digital health technologies is increasingly included in the priorities of the various NIH institutes and centers. Recent digital health advances are among the transformative opportunities that influenced the Office of Behavioral and Social Sciences Research (OBSSR) Strategic Plan, particularly for advancing measurement and methods, and for encouraging dissemination and implementation of effective interventions (Collins and Riley, 2016). Many NIH institutes and centers increasingly address digital health technologies in their strategic plans and priorities. For example, the priorities of the National Center for Medical Rehabilitation Research (NCMRR) include a subset specifically addressing technology use and development (Frontera, Bean, & Damiano, 2017). Recently, the National Mental Health Advisory Council (NAMHC) convened a Council Workgroup on Behavioral and Social Science Research to explore opportunities and challenges of using new information technologies to study human behaviors relevant to the NIMH mission (NAMHC Workgroup, 2016). The workgroup report provides an overview of NIMH research priorities, aligns these priorities with the NIMH Strategic Plan, and highlights key areas of opportunity. Technology-supported opportunities from this report include: (a) real-time automated assessment of behavior in natural environments (e.g., via sensors), (b) integration of technology data with other data sources (e.g., clinical or family history), and (c) digitally derived behavioral phenotypes to predict illness course/trajectories, identify at-risk youth for early intervention/prevention, and improve disease management. Since the onset of mental disorders is typically during childhood or adolescence, with origins prior to the onset of symptoms and associated impairments (Jones, 2013; Kessler et al., 2005), many of these workgroup recommendations are relevant to pediatric digital health. Researchers are encouraged to consider how their pediatric digital health research proposal fits with the strategic plans and priorities of the institute or center being considered for funding.

The NIH supports a growing portfolio of pediatric digital health research across a range of platforms. Health communication research is increasingly leveraging the online social media environment to examine trends in messages, sentiment analysis, and the use of social media platforms for interventions. For example, the NCI supports research on the surveillance of evolving themes and factors related to messages about e-cigarettes on social media platforms. Data gleaned from social media can inform the design of strategies to reduce diffusion of e-cigarette messages, identify particularly vulnerable population segments, and obtain these surveillance data in emerging areas much more rapidly than traditional survey-based surveillance approaches.

Social media also provides a platform for interventions delivered to youth. For example, NCI has funded research on a social media message campaign to address the interface between current parental-permission laws for indoor tanning, parent permissiveness, and indoor tanning behaviors in teen girls. The social media campaign seeks to reduce mothers’ indoor tanning permissiveness to maximize the effectiveness of extant parental-permission laws by developing messages around indoor tanning risk, decreasing parental permissiveness for tanning, and examining the prevalence of indoor tanning among mothers and daughters (Pagoto et al., 2016).

Gaming is another potential intervention platform for youth, including educational games, “gaming” competitions, and exergaming (Sween et al., 2014). To promote children’s (ages 6–10) physical activity, research funded by NHLBI relayed data from a commercial actigraphy device to a virtual pet that reinforced the child and also reported the child’s physical activity to the parents who could text encouraging messages to the child (Johnsen et al., 2014). An NIMH grant is using virtual reality and robotics to address the core social and attentional deficits in autism spectrum disorder (5R33MH103518-05). Health information and warnings also can be embedded within the commercial virtual gaming environment (e.g., graphic health warning for tobacco use) to influence health behaviors. However, determining the balance of providing sufficient health information to influence behavior without negatively impacting the inherent engagement of videogames requires further research.

Health apps are among the most frequently downloaded apps (Krebs & Duncan, 2015), and various smartphone apps have been developed and tested for a range of pediatric health goals including physical activity promotion, medication adherence, chronic disease management, stress and coping skills, psychosocial support, and HPV vaccination uptake. The use of digital technologies for dietary and physical activity monitoring and interventions has been a significant component of the research on childhood obesity funded by the NHLBI, NICHD, and other NIH institutes (Darling & Sato, 2017). In one funded study on the management of sickle cell disease, children (ages 10–17 years) reported that daily pain through an e-diary app remotely connected to a nurse practitioner who monitored pain reports and responded medically as needed, and to a counselor who provided coping strategies, peer support, and study website resources (Jacob, Duran, Stinson, Lewis, & Zeltzer, 2013). The NIMH supports a grant evaluating a smartphone intervention to augment behavioral parent training for conduct disorders and promote in vivo skills practice and mastery between sessions (5R01MH100377-04) as well as a tablet-based application to guide delivery and improve quality of treatment in child mental health settings (R01MH110620). An NINR grant is evaluating a mHealth intervention as part of a stepped care approach to improve medication adherence among children with epilepsy (1R01NR017794-01A1).

Sensor technologies are increasingly used in pediatric health research. For example, to address secondhand smoke risk, an NHLBI funded study evaluated a home-based intervention that included air quality monitors to alert families of exceeding fine particle concentration thresholds and to provide graphical feedback and social reinforcement for improving the air quality (Klepeis et al., 2013). An NINR grant is studying novel computer vision and wearable physiology sensor technologies to estimate pain severity in children (5R01NR013500-05).

The more recent NIH funded pediatric digital health research grants are integrating assessment and intervention components across platforms (i.e., multimodal) and across levels of influence (i.e., multilevel), and leveraging technologies to facilitate the research procedures as well. NIDA funds a Center for Excellence that aims to enhance the research that combines behavior change interventions with technologies to develop, evaluate, and disseminate technology-based interventions targeting substance use disorders (5P30DA029926-08). The Grow Right Onto Wellness (GROW) trial was a multilevel, family, and community-based trial to prevent childhood obesity (ages 3–5) that integrated study website with Facebook to recruit parent–child dyads, communicate key health messages, facilitate support among fellow participants, and (with GIS data) monitor and encourage use of the built environment (Barkin et al., 2018). In a study with asthmatic children (ages 12–17), iPads were used to provide consent and assent procedures for daily reporting of children’s asthma symptoms and to receive Facetime coaching on how to use a portable spirometer (Blake et al., 2015). An NINR grant is examining the application of a specialized technology software tool that collects electronic Patient-Reported Outcomes (e-PROMS) and generates feedback reports among children ≥2 years old with advanced cancer to family members and the oncologist (5R01NR016720-02).

Digital technologies are increasingly being used in pediatric surveillance and cohort studies. The NCI Childhood Cancer Survivor Study (CCSS; Brinkman et al., 2016) has recently adopted mobile technology to systematically assess 24,000 participants nationwide. The CCSS will use cell phones to collect health-related data (i.e., blood pressure, electrocardiograms, weight, and physical activity) from geographically dispersed childhood cancer survivors. NIDA and its NIH partners, to understand better the effects of substance use patterns on behavioral and brain development, recently launched the Adolescent Brain and Cognitive Development (ABCD) Study. This nationwide, longitudinal study is recruiting over 10 000 9-year to 10-year olds and following them for 10 years. The ABCD protocol includes a comprehensive dataset of behavioral assessments, multimodal brain imaging, bioassays, and an array of technology-based assessments (Bagot et al., 2018).

The NIH funds a broad range of pediatric digital health research. This research utilizes various digital platforms and increasingly integrates multiple platforms to address multiple levels of influence. These technologies are being utilized to conduct research more efficiently, to assess behavior and its influences “in the wild” and with greater temporal density, and to deliver interventions with greater reach, scalability, and adaptability.

Challenges and Opportunities in Pediatric Digital Health

Digital technologies hold considerable promise for assessing and changing health-related behaviors of youth and their parents to ultimately improve health outcomes. Smartphones provide a ubiquitous platform for prospective, real-time assessment (Shiffman, Stone, & Hufford, 2008), and recent advances in sensor technologies provide automated remote “direct observation” of behaviors and health indicators (Cornet & Holden, 2018). In-person interventions, or at least some components of these interventions, can be automated, greatly extending reach and scalability while also delivering interventions in the context in which the behaviors occur and adapting to these contexts (Nahum-Shani et al., 2018). Extensive use of digital technologies, such as text messaging and social media, especially by today’s youth who are “technology natives,” provides a rich data repository of thoughts, behaviors, and social systems in the natural environment. While there are many advantages of digital technologies for pediatric behavioral research and for improving the health and well-being of children, there a number of challenges to address.

Usability and Engagement

Although health apps are the most frequently downloaded category of apps (Krebs & Duncan, 2015), use typically dissipates rapidly unless engagement is carefully considered (Wagner et al., 2017). To promote “stickiness,” user-centered design approaches are important for mhealth and other digital technologies. Youth users present unique challenges because health and technology literacy vary across developmental stages, and solutions need to be developed that consider not only the end-user but also their parents and health-care providers. In preliminary studies, researchers are encouraged to test not only the effects of the various digital components on the targeted behavior, but also on engagement with these components, and to consider dose-response studies to understand better the relationship of engagement and engagement strategies to outcome.

Health Disparities

Although use of these technologies greatly extends the reach of behavioral pediatric interventions, that reach is not uniform across subgroups. Cell phones are ubiquitous, and smartphones are becoming increasingly so, but home broadband access via desktop/laptop remains unevenly distributed (Pew Research Center, 2018). It is also essential to address access and health and technology literacy challenges, particularly with underserved populations, to ensure that these digital health technologies reduce, not exacerbate, health disparities.

Privacy and Security

The recent NIMH NAMHC Workgroup deliberations and recommendations (NAMHC Workgroup, 2016) highlighted how the nature, granularity, and volume of data collected via these digital technologies introduce new challenges in maintaining privacy and confidentiality. For example, data collected via mobile phones, sensors, or other approaches often involves high volumes of sensitive information, particularly on mental health and substance abuse. With sufficient data, especially geo-tagged data, de-identified data can become identified. Strategies for securely storing and sharing these often-sensitive data are evolving (Martinez-Perez, de la Torre-Diez, & Lopez-Coronado, 2015), but researchers and Institutional Review Boards need to consider factors such as the terms of research participation, the nature of data collected, the data use plans, and the potential risks to both youth and parents, as well as the unique downstream consequences of breaches of sensitive information on future educational, employment, and other opportunities.

Negative Impacts of Screen Time

The considerable promise of these digital technologies for improving the health of youth needs to be balanced with research showing the impact of overexposure to technologies and excessive screen time. The currently available research and public health concerns regarding excess screen time and compulsive digital engagement led the American Academy of Pediatrics (AAP) to publish guidelines on screen time and digital technology use by children (Reid Chassiakos et al., 2016). The development, evaluation, and implementation of digital health technologies for children and their families must balance the benefits of these technologies with their potential negative effects, and design programs consistent with the AAP guidelines.

Pace of Research Versus Technology

Research timelines are measured in years, but technology development lifecycles are measured in months. As a result, evaluation of innovative assessment and intervention solutions delivered on novel technologies can become outdated or even obsolete by the time the typical longitudinal evaluation study is completed and published (Riley, Glasgow, Etheredge, & Abernethy, 2013). To address the slow pace of research relative to technology, agile science approaches have been proposed (Hekler et al., 2016; Patrick et al., 2016). The field also can benefit from test-bed resources that allow researchers to rapidly evaluate new digital health solutions. To address this need, the NIH supported the development of the Eureka research resource (Eureka Mobile Research, 2018). Eureka provides investigators an accessible, nimble, and sustainable infrastructure to conduct efficient and cost-effective mHealth research using a cloud-based multi-tenant platform and a complementary web interface and iOS application for enrolling, retaining, and conducting studies with large cohorts. Although no pediatric studies have utilized this resource yet, Eureka’s flexible design and its privacy and security standards provide pediatric researchers an easily accessible resource for rapid deployment and evaluation of mobile health studies.

Methodological Advances

The evaluation of pediatric digital health solutions can strain the capabilities of traditional methods. The optimization of Just-in-Time Adaptive Interventions (JITAI) requires being able to test various combinations and sequences of real-time intervention delivery (Klasnja et al., 2015). To foster replication, design specifications and considerations need to be more transparent and specific. Data generated by these digital health technologies require analytic approaches that can characterize the variability of intensive longitudinal data (e.g., Dziak, Li, Tan, Shiffman, & Shiyko, 2015) and better leverage big data analytic approaches.

Digital health technologies hold considerable promise for advancing pediatric research and pediatric health, yet the application and implementation of these technologies for assessing and changing health behaviors in children and adolescents has lagged behind that of adults. This lag is somewhat surprising given that the current cohort of children are technology natives for whom these technologies are fully integrated into their daily lives. Although there are a number of challenges to address, some unique to pediatrics, leveraging these technologies for research on pediatric health behaviors has the potential to produce significant advances in our understanding of pediatric health behaviors and how to change them to improve health. The NIH has made considerable investments, not only in research but also in research resources that can assist the research community in applying these digital technologies to understand and improve child health.

Conflicts of interest: None declared.

References