Mobile health and cardiac rehabilitation in older adults

Abstract With the ubiquity of mobile devices, the availability of mobile health (mHealth) applications for cardiovascular disease (CVD) has markedly increased in recent years. Older adults represent a population with a high CVD burden and therefore have the potential to benefit considerably from interventions that utilize mHealth. Traditional facility‐based cardiac rehabilitation represents one intervention that is currently underutilized for CVD patients and, because of the unique barriers that older adults face, represents an attractive target for mHealth interventions. Despite potential barriers to mHealth adoption in older populations, there is also evidence that older patients may be willing to adopt these technologies. In this review, we highlight the potential for mHealth uptake for older adults with CVD, with a particular focus on mHealth cardiac rehabilitation (mHealth‐CR) and evidence being generated in this field.

dynamic and quantifiable. 3 To date, mHealth has been used to facilitate recovery after acute myocardial infarction (AMI), 9 monitor arrhythmias, 10 and to track ambulatory blood pressures. 11 Applications have also been created to encourage medication adherence, facilitate social support, 3 [12][13][14][15][16] and hypertension. 11,17,18 Smartphones themselves act as a passive sensor, as most are equipped with 9-axis motion sensor and a 3-axis accelerometer which can be used to track distance traveled, count steps, and even detect falls. 4,19 Such accelerometers have been used with some success in mobile applications that track patient activity in CR programs. 20,21 More recently, the increase in commercially available wearable devices (such as the Apple Watch [Apple, Cupertino, California]) has made arrhythmia detection possible by portable sensors. Wrist sensors, using photoplethysmographic monitoring (optically-obtained measurements that detect changes in light related to blood flow in capillary beds, similar to those used for skin-based pulse oximeters) can measure heart rate, heart rate variability, and use associated algorithms to detect arrhythmias including atrial fibrillation. 22 Portable electrodes, such as the AliveCor Kardia Band (AliveCor, Mountain View, California), can also be connected to a wearable device and quickly generate a one-lead electrocardiogram (ECG) to detect the presence of atrial fibrillation. 10 Similar technology has not only been employed in the United States and Europe, but also lower-resource settings such as rural India to detect atrial fibrillation. 23 Improvement in medication adherence has been a target of mHealth development, and could be of particular use in older adults.
F I G U R E 1 Role of mHealth-CR in older adults Adherence can be assessed remotely via several mechanisms, including "E-blisters," or medication packets that send an electronic signal after being opened, 24 or even pills can be embedded with a miniature sensor that emits a signal after entering the acidic environment of the stomach. 4 With several mobile applications using mobile reminders for medication administration, 13 there is opportunity for development of streamlined programs that potentially can decrease polypharmacy and confusion over home medication regimens while emphasizing medication adherence. Patient education and empowerment with self-monitoring is a potentially powerful tool offered by mHealth. mHealth interventions aimed at self-monitoring of chronic conditions have shown improvement in reducing harmful behaviors such as smoking (albeit in younger populations). 18 Two large ongoing studies will assess mHealth interventions aimed at improving self-management in older populations.
The SMArTVIEW trial (Self-MAnagemenT-VIsion for patient EmpoWerment), currently ongoing in the United Kingdom, will analyze whether Bluetooth enabled monitoring devices and mHealthdelivered educational materials will optimize the ability of older adults following cardiac surgery to recognize when they require medical attention with hopes of improving postoperative outcomes. 25 Also ongoing, the HATICE trial (Healthy Ageing Through Internet Counselling in the Elderly) will provide older adults at high risk for CVD with an interactive internet platform with educational materials and remote support by a coach with the goal of optimizing risk factors. 26 As the population ages and CVD becomes more common, these types of mHealth applications could be of particular utility in older adults. A list of selected studies 11,13-18,20,21,23,24,26-39 regarding mHealth in this domain that may be relevant to older adults is provided in Table 1. The remainder of this narrative review will focus on mHealth cardiac rehabilitation (mHealth-CR) in older patients, which represents a particularly attractive target for intervention.

| CR: BENEFITS AND CURRENT BARRIERS
CR, which is traditionally offered as a comprehensive center-based program, has long been recognized as playing an important role in secondary prevention of events related to CVD, and is recommended by the American College of Cardiology (ACC) and American Heart Association (AHA) for use after acute myocardial infarction (AMI), percutaneous coronary intervention (PCI), and coronary artery bypass revascularization (CABG), as well as for chronic stable angina, and heart failure with reduced ejection fraction. 41,42 Accordingly, referrals to CR for all of the above diagnoses (as well as symptomatic peripheral arterial disease [PAD]) are reimbursed by the Center for Medicare and Medicaid Services. 43,44 CR reduces all-cause mortality, 45 cardiovascular mortality, and hospital readmissions, and improves health-related quality of life and exercise capacity. 46 Traditional CR is generally based at a single ambulatory center and involves a structured, supervised exercise program (usually 3 sessions per week for 36 total sessions) that are supervised by trained physicians, nurses, and exercise therapists.
Despite the known benefits of CR, referral and participation rates have remained stubbornly low. More than 80% of patients who are eligible for CR in the United States do not participate. 47 Fewer than twothirds of patients who are eligible for CR are referred, and even when referred, only about one-half will attend the first prescribed session. 48 Certain populations, including older adults, are particularly underreferred. 47 Once enrolled, a substantial proportion of patients do not complete the prescribed 36 sessions; this is clinically relevant, as many studies suggest a dose-dependent relationship between number of CR sessions attended and improved outcomes. 49,50 Older adults have several unique barriers to sustained participation in CR, including transportation issues (lack of a vehicle or vision/hearing impairment that precludes driving), cognitive impairment, and physical limitations ( Table 2). 51 These barriers and the aforementioned suboptimal referral and participation rates have been noted by the ACC/AHA, which have recommended that CR be "reengineered to include a wide array of service options that meet the needs of individual patients." 47

| NONTRADITIONAL CR
Home-based CR has emerged as an alternative method of delivery to traditional center-based CR programs. Home-based CR involves prescribed exercise that can be carried out in a variety of settings and can be delivered "mostly or entirely outside of the traditional centerbased CR setting." 52 An option for decades in other countries (Australia, Canada), and even in some select health systems in the United States, home-based programs have been evaluated by two recent Cochrane Reviews revealing no difference in outcomes (mortality, exercise capacity, cardiac events, or quality of life) when compared to traditional center-based CR programs. 53,54 While some studies have shown that there is no difference in rates of adherence in home-based vs traditional CR programs, 53 others have suggested that patients may complete home-based CR at higher rates. 55 Patients, when given the choice between a home and center-based CR program, often prefer a home-based approach. 56 There are several important differences between home-based CR and traditional center-based CR. Center-based programs are located in medical facilities with groups of patients under direct in-person supervision from physicians with access to emergency response capabilities.
During their sessions, patients in traditional CR are typically monitored on telemetry. Home-based CR programs, by definition, occur in settings other than healthcare facilities, and while no standardized home-based CR program currently exists, patients typically are not directly supervised by medical staff, they are not monitored on telemetry, and they do not have the experience of exercising in groups. A recent ACC/AHA statement accordingly deemed home-based CR a reasonable "alternative option to recommend for select clinically stable low-to-moderate risk patients who cannot attend traditional center-based CR." 52 Data regarding the safety of home-based CR have shown events are rare (one study estimates 1 event per 50 000 patient hours), 57 but studies may have been underpowered to detect the risk of significant cardiovascular events in higher risk subgroups. 52 to date involving hybrid home and center-based CR, found no increased risk of cardiovascular events in the group prescribed at-home exercise compared to traditional care. 42 The recent HONOR trial, which evaluated an mHealth-augmented home-based exercise program for patients with PAD (in patients with a mean age of 70 years), also found no significant difference in significant adverse events between the intervention and control group. 58 While further research is needed into the safety of home-based CR programs, these results provide indirect support that they can be prescribed safely.

| MHEALTH IN CR
Given that home-based CR has been an option for years, the extension of mHealth to augment CR delivery represents a logical evolution in the broadening of its availability. Several studies, mostly outside of the United States, have used mHealth in creative ways to augment the CR experience. In an early study, Worringham and colleagues utilized a smartphone-based program with a single-lead ECG and GPS to employ a remote, walking-based CR program for 134 monitored sessions in six patients in Australia, with data transmitted to an exercise therapist; patients using the program had similar improvements in 6-minute walk test compared to those using traditional center-based CR. 34 A study of 62 male patients in Poland undergoing CR after acute MI also showed that a home-based program, with remote ECG monitoring and a mobile device with preprogrammed exercise instructions, showed similar benefit in physical capacity compared to traditional CR. 59 Several small studies have also supported the feasibility of various mHealth applications in the setting of CR, with success in monitoring medication adherence, 13,21 estimating levels of physical activity, 60 and providing real-time data to clinicians for feedback. 33 Notably, older adults in these studies did not report difficulty using the mHealth technology. 21,27 Unfortunately, rigorous data regarding the effectiveness of mHealth in CR are still lacking. To our knowledge, one of the largest randomized-controlled trial to date regarding mHealth in CR investigated 120 post-AMI patients in Australia, randomizing patients to traditional CR vs a smartphone-based home CR delivery model, with the smartphone application allowing for exercise monitoring, educational content delivery, and nutritional and psychological counseling via a web-based portal. Results were promising; patients randomized to the intervention arm were more likely to participate in CR (80% vs 62%), had higher completion rates (94% vs 68%), and had similar positive exercise outcomes to the traditional CR group. 20 A more recent trial in Minnesota comparing traditional CR and mHealth-augmented CR also showed benefit; briefly, participants in the mHealth group received a smartphone or web-based application that provided them with educational materials and allowed them to record their exercise Through an mHealth application, patients could record when they were taking medications, and e-blisters (packages that send an electronic signal once opened) recorded medication compliance. Patients randomized to mHealth group had increased adherence to oral diabetes medication (metformin); otherwise, no significant improvement in adherence was found. The results of these trials, however, may not be generalizable. In the Australian study, participants were 87% male and had a mean age of 56 years 20 ; in the US study, most were also male (82%) and relatively young (mean age 63). 40 These trials therefore may have limited external validity when considering more typical "older adult" patients (commonly defined as age ≥75) who have both the highest CVD burden and the lowest familiarity with mHealth. make certain devices more difficult to manipulate. 63 The recent development of personal voice assistants ("Siri" by Apple; "Alexa" by Amazon), which interpret user voice commands, may be particularly useful in older populations, where vocalization is often maintained, and should be a focus in the development of mHealth applications going forward. 64 While literature supports that older adults may accept new technologies, they often do so with less confidence than younger adults. 65 In a qualitative study on the use of mHealth in older patients with heart failure, lack of knowledge, and even "fear" of misusing the technology, were the most commonly cited barriers to mHealth adoption. 66 In addition to implementation barriers, there are other potential barriers to success: most notably, current mHealth-CR platforms gener-

| FUTURE DIRECTIONS
For these reasons, it is imperative that further research into mHealth, and specifically mHealth-CR, be performed in older adult populations.
mHealth represents an opportunity to widen the scope of CR for those patients who would benefit the most from it. The RESILIENT trial (Rehabilitation Using Mobile Health for Older Adults with Ischemic Heart Disease in the Home Setting), now underway, is a prospective, randomized, NIH-funded multicenter trial will employ an mHealth intervention to augment traditional CR in patients ≥70 years old. Participants in the intervention arm will be given mHealth-CR software (through an electronic tablet device) that will allow for remote contact by an exercise therapist, and will be given activity trackers (eFitBit) to document their level of physical activity. Intervention (300 participants) and control groups (100 participants) will also be referred to traditional center-based CR per ACC/AHA guidelines; the trial therefore represents a hybrid approach to CR. After 3 months, investigators will assess functional capacity (via 6-minute walk test), goal attainment, health status, ability to perform instrumental ADLs, and hospital readmissions. 36 The upcoming EU-CaRE trial (European Study on Effectiveness and Sustainability of Current Cardiac Rehabilitation Programs in the Elderly) will evaluate 238 patients age ≥65 who have declined traditional CR and enroll them in an mHealthguided home-based CR program with devices providing advice and coaching throughout the study period. Functional status (peak VO2), CR uptake, adherence, and participation will be recorded and compared to a traditional CR group (1720 patients). 37 These trials, and others which are likely to take place, will be vital to the developing field of mobile CR, providing information regarding the safety and efficacy of mHealth-augmented CR in older adults.

| CONCLUSION
mHealth has enabled numerous avenues for remote management of CVD. Older adults, with the highest burden of disease, may stand to benefit the most. mHealth-CR represents a particularly attractive area given traditional barriers to facility-based CR. Small studies have demonstrated potential benefits to mHealth-CR, but older adults have been under-represented, and further research will help to elucidate engagement and outcomes among older adults who are prescribed this intervention.