Harnessing mHealth technologies to increase physical activity and prevent cardiovascular disease

Abstract

Research into prevention of cardiovascular disease has increasingly focused on mobile health (mHealth) technologies and their efficacy in helping individuals adhere to heart‐healthy recommendations, including daily physical activity levels. By including the use of mHealth technologies in the discussion of physical activity recommendations, clinicians empower patients to play an active daily role in modifying their cardiovascular risk‐factor profile. In this review, we critically evaluate the mHealth and physical activity literature to determine how these tools may lower cardiovascular risk while providing real‐time tracking, feedback, and motivation on physical activity levels. We analyze the various domains—including user knowledge, social support, behavioral change theory, and self‐motivation—that potentially influence the effectiveness of smartphone applications to impact individual physical activity levels. In doing so, we hope to provide a thorough overview of the mHealth landscape, in addition to highlighting many of the administrative, reimbursement, and patient‐privacy challenges of using these technologies in patient care. Finally, we propose a behavioral change model and checklist for clinicians to assist patients in utilizing mHealth technology to best achieve meaningful changes in daily physical activity levels.

1. INTRODUCTION

Cardiovascular disease (CVD), which is the leading global cause of death, has seen innovations in prevention and care that have led to a 39% reduction in CVD mortality since 2001.1 Despite this progress, additional improvements are needed to further reduce the burden of CVD in our population.

The identification of key risk factors for CVD plays a critical role in our ability to properly inform patients about their CVD risk profile. Physical activity (PA), a lifestyle modification often overlooked by many clinicians, provides ameliorative effects on cardiovascular (CV) risk factors and disease.2 Despite the convincing benefits of PA and the deleterious effects of physical inactivity, most adults do not meet current PA recommendations.3 Current interventions to improve PA rely on in‐person activities, causing numerous constraints on intervention fidelity,4 ultimately requiring new strategies to promote motivation and sustainable behavioral change.

mHealth, the use of mobile and wireless technologies to support the achievement of health objectives, has been proposed as a means to improve PA and ultimately decrease the burden of CVD. mHealth is an emerging tool for disease management, health behavior assessment, and intervention.5 The ubiquity of mobile technology in conjunction with the ability to track real‐time data, such as vital signs and number of steps taken, offers a unique opportunity to intervene in the ambulatory setting. mHealth's potential impact on PA and behavior change is very promising. Unfortunately, the rapid growth of mHealth tools has outpaced the capacity of clinicians to utilize these tools in a cost‐effective, patient‐centered manner that yields meaningful change.

To help address this rapid growth and the subsequent gaps in knowledge, we constructed a narrative review that couples a behavior change model and a checklist of mHealth interventions that help promote increased levels of PA. The result is an essential step toward empowering clinicians to effectively improve patients' adherence to heart‐healthy lifestyle modifications that emphasize PA.

2. SMARTPHONE APP USE TO INCREASE PA

Smartphones, which integrate embedded and external “wearable” sensors and programmable applications, are destined to revolutionize mobile healthcare.6 An estimated 64% of adults in the United States now own a smartphone, and nearly two‐thirds of those users used their phone to access health information in the past year.7 Smartphone ownership is maintained across races, with Hispanics (71%) and blacks (70%) demonstrating the highest rates. Consumer demand has already created an industry for mHealth applications, with ~50% of all mobile subscribers using a fitness application.8

As the list of available products expands, researchers are beginning to get a better picture on how to best harness this technology. The use of smartphone applications to increase PA was recently analyzed by Bort‐Roig et al., who found that in 4 of 5 studies, smartphone applications increased PA by as much as 1100 steps per day (roughly one‐half mile).9 This finding is not surprising, as mHealth technologies have previously been shown to increase PA through various modalities including pedometry, SMS messaging, email, and internet. If sustained, this increase could prove clinically significant. Bravata et al. demonstrated that a step count increase of ~2500 steps during an 18‐week period was associated with significant decreases in body mass index and systolic blood pressure.10 In addition, each 2000 step/d increase in PA, sustained for a mean of 6 years, was associated with a ~10% relative reduction in the incidence of CVD.11 What remains to be understood, however, is how smartphone applications can better leverage these modalities to change behavior and create lasting habits. For this reason, we identified a model from Murray et al.,12 detailing how mHealth interventions lead to clinical change. We then adapted this model to illustrate how mHealth can lead to increased levels of PA (Figure 1).

Figure 1.

The mHealth behavior‐change model, leveraging the critical components of mHealth technology and behavioral‐change theory to achieve long‐term CV protection. Abbreviations: AHA, American Heart Association; CDC, US Centers for Disease Control and Prevention; CV cardiovascular

In the model, major domains through which smartphone applications can impact PA include knowledge, social support, behavioral change support, decision support, and self‐efficacy. The following sections describe how available apps target each domain to implement behavior change. We also review the use of these behavior‐change techniques in relevant clinical studies.

2.1. Public health information

Smartphone applications improve user knowledge through education about PA and substantiate clinician advice. Smartphones employ many behavior‐change techniques to increase user knowledge, which commonly include providing credible/population‐based recommendations for PA, detailing the benefits of a behavior, advising where and when to perform the behavior, and giving normative information about others' behavior. In fact, among PA apps available, behavior‐change techniques targeting knowledge are among the most frequent.13

Computer‐assisted instruction has been considered an effective and versatile tool for behavior change,14 which the user‐friendly interface of mobile applications enables. Researchers have also employed a variety of means to educate participants via mobile apps. Previous studies attempted to alter the PA practices of individuals by providing them various educational and instructional resources; however, due to limitations such as small sample sizes and self‐report, none reported significant increases in PA.15, 16

In summary, the distribution of public health information to increase knowledge is pervasive among smartphone apps; however, the efficacy of significantly increasing PA has not yet been established.

2.2. Social support

Mobile applications employ social, technological, and decision support to influence users. Social support, which is emphasized in patient‐facing applications, can be emotional, instrumental, or informative. It is separate from social influence, which can also have both positive and negative effects on a health behavior.17 Behavior‐change techniques that apps use to enable social support include allowing users to attain encouragement, facilitating approval from others, and providing opportunities to share and compare accomplishments via social media. Social‐support components are present in more than half of all popular electronic activity monitors, such as those created by Fitbit,18 and are nearly ubiquitous to PA apps.19

Social support comes from peers, which may be especially important in adolescent populations.20 Research into the efficacy of mobile‐app interventions has employed a range of techniques to enable social support. In the Mobile Pounds Off Digitally (Mobile POD) study, the formation of anonymous Twitter cohorts attempted to facilitate social feedback and group discussion. In this study, however, Twitter use was sporadic and PA was not increased.15 Another study suggested that the quality of social support is critical to its effectiveness; Rabbi et al. found that automated messages that were individually personalized, contextualized, and actionable increased walking duration significantly more than a similar amount of generic advice.21 Although some studies have shown success from motivational and social support, more research is needed to determine how smartphones can effectively provide support to drive behavior change.

Thus, social‐support components, despite their prevalence, appear to vary in quality, and their efficacy may depend on the extent to which the support is personalized.

2.3. Behavioral‐change support

Mobile apps offer behavioral change and decision support by prompting the user to create goals, specific plans of action, and a behavioral contract. Applications also employ behavior‐change techniques that involve intention formation, environment restructuring, and the provision of rewards to support user behavior. In comparison to social support, Yang et al… found that behavior change and decision support techniques were less common among PA apps.19 Direito et al additionally found that free apps employ behavior‐change and decision‐support techniques such as intention formation and specific goal‐setting less frequently than do paid apps.22

Researchers are only beginning to explore the ways in which smartphones can offer behavioral change and decision support. A recent randomized controlled trial of a smartphone‐based mHealth intervention found that participants receiving clinician‐scripted personalized text‐message support increased PA significantly over PA tracking alone: 2534 more steps (P < 0.001).23 Another study, which rooted its intervention in behavioral change theory (It's LiFe!), had participants keep diaries about the enjoyment and exertion noted during PA, asked participants to list barriers and facilitators to exercise, and helped them devise daily action plans to achieve goals through a mHealth app.24 The It's LiFe! study determined that applications with clinician‐supplemented counseling sessions had a significant effect on daily PA levels (mean difference, 12 min/d; 9%; P < 0.001). Lastly, the TEXT ME trial supported this concept, showing that the use of a lifestyle‐focused text‐messaging service compared with usual care resulted in modest improvement in CVD risk factors, including levels of low‐density lipoprotein cholesterol.

These studies suggest that for a smartphone app to be most effective, it must serve as an extension of the clinician, rather than as a stand‐alone intervention.

2.4. Self‐efficacy and motivation

Self‐efficacy is defined as the belief in an individual's ability to succeed in certain situations or accomplish specific tasks. Users can actively interact with their apps through self‐report, manually logging specific PA and monitoring the achievement of goals. Currently, self‐efficacy measures are also fairly underrepresented on the application market, with paid apps being more likely to prompt self‐monitoring.

Self‐efficacy can add to patient motivation; conversely, a lack of self‐efficacy can deteriorate motivation. Glynn et al… demonstrated that short‐term use of an app that includes PA tracking and goal‐setting helped patients increase PA levels by ~1000 steps per day.25 When passively tracked step counts were manually entered into an application, Kirwan et al found that participant engagement produced behavior change via increased PA for the duration of the 90‐day study.26 Therefore, it is apparent that patients must play an active role in lifestyle modifications, and applications can help.

Based on our expert opinion, as research into applications continues, clinicians can consider making recommendations that patients use specific apps that incorporate multiple behavioral‐change techniques. These apps should work on several levels, not only educating, but enabling social, behavioral, and decision support, all while increasing user self‐efficacy. Patients should actively track PA behaviors in addition to using passive pedometer tracking. Patients would also benefit from the integration of social interaction onto the mHealth platform. Motivating patients to alter their behaviors is not as easy as buying a smartphone and downloading an app. A lasting behavioral change requires collaboration between clinician and patient, where progress toward short‐term activity goals can be monitored over time. One example of this strategy is the mActive trial, in which smart texts using the patient's clinician's name provided positive reinforcement for patients reaching daily goals and words of encouragement for patients infrequently surpassing their goals. By leveraging the patient‐clinician relationship, the mActive trial saw nearly twice as many participants in the text‐receiving arm achieve their goal of 10 000 steps per day.23 Only after a patient demonstrates he or she can consistently achieve short‐term goals can a clinician confidently prescribe the type of long‐term exercise prescriptions that meaningfully impact CVD risk. Although significant CVD risk reduction results from long‐term improvements in PA levels via exercise prescriptions, leveraging clinical supervision and motivational feedback to increase PA levels, even minimally, will benefit at‐risk patients.27

Self‐efficacy, though seemingly underutilized, is key to helping clinicians and patients stay motivated and turn short‐term progress into long‐term behavioral change.

3. mHEALTH TECHNOLOGY INTERPRETATION AND ADOPTION BY CLINICIANS

Telehealth platforms such as mHealth are believed to possess remarkable capacity to enhance the value and availability of healthcare.28 Adler‐Milstein et al. reviewed data from the American Hospital Association's 2012 survey, discovering that 42% of US hospitals have adopted some form of mHealth/telehealth.29 They found significant associations between adoption and population density, perceived strategic benefits of telehealth, and state policy related to reimbursement.

In 2016, the American Medical Association conducted the Digital Health Study, which included a survey of 1300 clinicians, to better understand how clinicians are utilizing digital tools and the requirements necessary for adopting these innovations.30 The study found that 39% of providers are already using such technologies to increase patient engagement, with 99% expecting to incorporate these tools into their practice within 2 to 3 years.

As adoption increases, clinicians will place even more emphasis on data accuracy and the comparability of data across devices and platforms. Similarly, as interactions with patients become increasingly more personalized, clinicians will want to ensure that the accelerometer data is reliably extrapolated to energy‐expenditure data appropriate for a particular patient. Some of this work is already being done. Noah et al. demonstrated the reliability of Fitbit devices for step counting and determining energy expenditures compared with indirect calorimetry.31 Use of digital tools for remote monitoring of patients' risk factors is being implemented by 24% of practicing clinicians, and 98% are looking to adopt these practices within 2 to 3 years.

In summary, clinicians already anticipate nearly ubiquitous adoption of mHealth technology to increase patient engagement. In the meantime, they are working to ensure that the data outputs are both precise and accurate.

3.1. Administrative challenges

Despite encouraging clinician and patient uptake, the landscape of mHealth adoption remains hindered by complex administrative barriers. The legal and regulatory framework surrounding mHealth platforms is under active reconstruction, as the guiding principles of the healthcare system were developed years before the disruptive technology of mobile connectivity.32 In the United States, the Food and Drug Administration currently regulates certain medical devices and mobile applications with a specific eye toward data security.33

The Health Insurance Portability and Accountability Act and its extension, the Health Information Technology for Economic and Clinical Health Act (HIPAA/HITECH), regulate the use of patient data by healthcare providers. When a patient's health information is in the possession of a covered entity, such as a clinician, it is governed by the policy set forth in HIPAA/HITECH. Therefore, patient‐generated health data is simply data if a patient monitors step count with a mobile application; however, that information falls under HIPAA/HITECH when it is transferred to a clinician with the intent of monitoring that patient's care.32, 34

mHealth developers have not actively engaged with this complex regulatory environment as a cohesive community, and thus the population of available mHealth tools is incredibly diverse—not only in terms of clinical utility or patient uptake, but also legal permissibility. This situation creates a profound research‐implementation gap between FDA‐regulated developers and HIPAA‐bound clinicians. To mitigate this issue, the American Medical Association (AMA) has proposed the use of a specialized electronic health record application store, with available programs made to securely integrate into existing electronic health record frameworks.30 This would help address the 3 most common clinician requirements regarding the utilization of mHealth tools: malpractice coverage (81%), data privacy (82%), and workflow integration (81%).

Thus, the challenges inherent to integrating mHealth data into the patient record are apparently being addressed thoughtfully with both patient and clinician concerns in mind.

3.2. Reimbursement

There has been a recent increase in interest surrounding the use of mHealth applications, pushed by the unrolling implementation of the Patient Protection and Affordable Care Act's (PPACA) emphasis on efficient healthcare delivery.35 The increased awareness of mHealth's utility has underscored the current lack of clarity surrounding reimbursement schemes for the care provided on such platforms.36 Currently, ~19 states have “parity” legislation that requires third‐party payers to honor telemedical claims, but such piecemeal advances differ among states and fail to scale up in terms of substantial Medicare or Medicaid billing avenues.35 This stepwise approach is significantly correlated with the degree of mobile health adoption by providers, and thus availability to patients.29 The transition of this technological advancement to the primary‐care arena has yet to be properly facilitated within the healthcare reimbursement structure, which may be the most significant barrier to widespread utilization of mHealth within the current framework.37

3.3. Clinical relevance and benefit

Motivating patients to improve modifiable risk factors as a means for reducing CVD is one of the most difficult problems facing the modern clinician. Behavior‐change theory has been established as an effective framework for confronting this issue. Building on the success of this theory, we adapted a model (Figure 1) for the application of mHealth platforms to enhance PA from a Cochrane Review, which analyzed the use of interactive health communication applications for those with chronic disease.12 This model was then operationalized for use by clinicians through the formation of a checklist (Figure 2), a modality chosen based upon its proven utility for guiding patient care.38 The checklist provides a set of pragmatic suggestions meant to assist the clinician in leading a patient toward achieving the American College of Cardiology/American Heart Association (ACC/AHA) PA recommendations.39 We must acknowledge, however, that our checklist is not evidence based, but a useful expert construct meant to try to move the field forward in a structured way.

Figure 2.

Clinician checklist for assisting patients in achieving the physical activity guideline recommendations. Abbreviations: ACC, American College of Cardiology; AHA, American Heart Association

Throughout the ongoing validation of these methodologies, it is evident that PA intervention success is linked to personalized motivational strategies from clinicians,23 compared with those with passive, self‐motivating PA tracking.40 Ultimately, the most effective use of mHealth technology to address CVD risk profile will require that clinician‐patient teams continuously collect data and maximize motivational messages to increase PA, lower CV risk, and improve outcomes. One initial step for clinicians is to incorporate “exercise prescriptions” into their treatment armamentarium and begin using PA as a vital sign; both actions are supported by the PA literature.41 Metkus et al. proposes an example of a staged “exercise prescription” to institute sustainable lifestyle modifications that are required to yield the long‐term benefits of exercise training on CV risk factors. By utilizing exercise prescriptions, clinicians can take a systematic approach to staging increases in individual PA levels and avoid generalizing recommendations that may intimidate certain patients and preclude initiation of the exercise program.42

4. CONCLUSION

In the prevention of CVD, there is opportunity for major improvements in the promotion of and adherence to PA guideline recommendations. mHealth may be used to impact the course of CVD in our nation, with the ultimate goal of ending its run as the leading cause of death. To accomplish this, clinicians may utilize our proposed checklist and incorporate the concepts of our behavior‐change model to effectively guide patients in adopting lifestyle modifications aimed at increasing levels of PA. The emergence of mHealth devices and applications as convenient and motivating tools for patients to track their PA status and goals provides an opportunity to better promote PA guideline adherence. Although the burden of CVD remains high, improving PA remains an effective strategy for reducing the morbidity and mortality of CVD. Implementation of provider‐guided monitoring of PA could revolutionize the current paradigm of PA for addressing CVD risk factors.

Conflicts of interest

The authors declare no potential conflicts of interest.

Feldman DI, Theodore Robison W, Pacor JM, et al. Harnessing mHealth technologies to increase physical activity and prevent cardiovascular disease. Clin Cardiol. 2018;41:985–991. 10.1002/clc.22968