Wearable Devices and Psychological Wellbeing—Are We Overthinking It?

In this issue of the Journal of the American Heart Association (JAHA), Rosman et al presented their findings from a retrospective, propensity‐matched study of patients with atrial fibrillation (AF) with and without wearable devices. They reported a significantly higher rate of symptom monitoring, preoccupation, and anxiety among wearable device users compared with nonusers, and a significantly higher rate of health care usage among wearable users. ¹ Specifically, the authors reported greater formal use of health care resources (ie, AF‐specific outpatient visits, diagnostic tests, ablation) and informal use of health care resources (ie, messages via patient portal, telephone calls). These findings provide greater insight on how wearable devices may impact our patients living with AF, and the authors are congratulated in completing this important study.

Wearable devices may provide substantial benefits for patients and users. Currently, the majority of clinical‐grade cardiac monitors provide a limited duration of arrhythmia monitoring, often up to 2 to 4 weeks, and may be insufficient for those with paroxysmal symptoms. For those with intermittent symptoms who are seeking symptom–rhythm correlation, wearable and handheld devices provide greater user‐centric access to cardiac monitoring, and can potentially provide users with actionable clinical data. Moreover, mobile devices as a means for intermittent cardiac monitoring has been evaluated before as a feasible means for arrhythmia detection. In the eBRAVE‐AF (eHealth‐Based Bavarian Alternative Detection of Atrial Fibrillation) trial, 5501 patients without AF were randomized to digital screening using a smartphone application versus usual care. At 6 months, patients in the screening group were twice as likely to be diagnosed with new‐onset AF compared with the usual care. ² Similarly, mega‐studies using smartwatches from consumer technology companies have also demonstrated a small but real incidence of AF detection. ³ , ⁴ , ⁵

Findings From Rosman et al

The findings from Rosman et al highlight the potential risks of wearable device use. Across 172 patients who responded to their survey, 83 patients in their cohort used a wearable device (Apple Watch and FitBit, most commonly). Wearable device users were younger than nonusers but more likely to have an implantable cardiac device. The authors reported higher scores in cardiovascular symptom monitoring and preoccupation among wearable users, along with greater AF treatment concern according to the AF Effect on Quality‐of‐Life Questionnaire. Wearable users frequently checked their heart rate on a daily basis, perhaps due to accessibility and ease, and always contacted their physician if they received a wearable device notification for possible AF. Furthermore, 15% to 20% of patients felt anxiety or fear with automated wearable alerts or irregular rhythm notifications.

The authors also reported significantly higher health care usage among wearable users. Over a 9‐month period following index wearable purchase compared with matched controls, the authors reported a higher number of AF‐specific outpatient visits and rhythm‐related diagnostic tests and procedures, including a greater number of ECGs, echocardiograms, and ablation procedures. There was also a higher rate of informal health care usage, including messages sent via patient portals to cardiology and primary care clinics.

Limitations of These Findings

It is important to highlight, while provocative, these findings remain primarily retrospective and subject to multiple biases. Creation of the wearable user cohort itself was performed through an electronic invitation, among which 756 patients were invited and only 178 patients responded. Wearable device users who responded may not reflect the overall population of wearable device users (including those who did not respond). For example, wearable device users who recently underwent a diagnostic test or procedure may be more likely to check their patient portal and respond to an electronic invitation, thereby selecting those who used more health care resources.

Second, wearable device users themselves may be self‐selected as those with a higher symptom burden or differing arrhythmic phenotype. Patients living with AF can have substantially differing degrees of symptoms and impacts on the daily quality of life—with the availability of wearable devices, those with greater symptom burden may be driven to seek wearable devices and confirmatory testing, and likewise, have a greater use of health care resources (irrespective of wearable device use). This is something that is difficult to answer without a randomized trial comparing those with and without access to a wearable device. Most suggestive of this selection bias is the finding that wearable users had a greater number of ablation procedures compared with those without a wearable. Do we think the wearable device truly resulted in physicians and electrophysiologists more likely to perform an ablation procedure? Or is it that these patients were more symptomatic and required more interventions, and the wearable device was a manifestation of such?

One major challenge is identifying a control group that can be best matched with the wearable device population. While the authors performed a propensity match for age, sex, AF pattern, socioeconomic status, and comorbidities, whether the 2 populations (wearable users and nonusers) remained slightly different is unclear (age, race, and implantable cardiac device status remained slightly different between groups despite matching). Similarly, while the “wearable purchase date” was defined as the index date for calculating health care resource usage, it is unclear how this date was assigned for nonusers. One might consider a “before‐and‐after” analysis to evaluate the impact of wearable devices on health care resource usage as has been performed in other cardiovascular intervention studies. ⁶ , ⁷ Does health care usage truly increase after an individual purchases a wearable device, or was that individual more likely to seek care irrespective of the wearable device?

Need for Educational Resources and Responsibility of Consumer Technology Companies

Despite these limitations, some wearable users experience significant anxiety, or even fear, when they receive an alert from their wearable device. It is critical that our health care community seeks to improve the understanding of our patients and wearable users alike. We need to ensure that our patients are educated on the potential findings of a wearable device, and the meaning of false‐positive or false‐negative results. While the users in this study were primarily using the Apple Watch or FitBit, patients (and the general population) may also use a variety of different wearable devices, with varying degrees of specificity and sensitivity, and it is critical that we ensure that educational resources are available for these individuals.

Despite their widespread availability, many wearable devices marketed for arrhythmia detection have not been thoroughly evaluated for accuracy and, in many countries, lack regulatory oversight. Consumer technology companies also play a critical role. Devices must be designed with sensitivity and specificity in mind, minimizing the likelihood of false‐positive results that may cause unnecessary anxiety. Educational resources must be built into devices (and alerts), so that users are not left wondering and worrying about a device alert. With the increasing incidence of AF, it is critical that these companies are committed to ensuring their products do not generate excessive anxiety or unnecessary health care usage among their users.

Future Directions for Wearable Devices

Wearable devices have substantially changed our delivery of arrhythmia care. Wearable devices provide users with ad hoc arrhythmia detection and prolonged noninvasive monitoring that is not available through conventional clinical‐grade monitoring devices. ⁸ However, as highlighted by Rosman et al, through the same mechanism, wearable devices may also increase anxiety among our patients, potentially leading to increased formal and informal health care usage. ¹ These findings underscore the need for further investigation into the potential effects of wearable devices on patient's health care use and psychological well‐being, as well as the downstream effects on providers, clinic workflow, and health care expenditures.

The study by Rosman and colleagues also highlights an important gap in arrhythmia care. Are patients driven to use a wearable device because our current health system does not provide patients with sufficient understanding or insight into their arrhythmia? In their simplest form, wearable devices empower our patients to take a more proactive role in their health care by detecting and monitoring their arrhythmias. Future developments in wearable devices may provide direct feedback and drive management of various conditions; for example, the ongoing REACT‐AF (The Rhythm Evaluation for Anticoagulation With Continuous Monitoring of Atrial Fibrillation) trial explores the use of AF detection with a smartwatch as a means to guide pill‐in‐the‐pocket anticoagulation. ⁹ In diabetes care, personal, real‐time glucose monitors have similarly provided patients with greater understanding and control in managing their condition. ¹⁰ This type of patient‐driven monitoring and management can be extended to rhythm control of AF, to documentation, and to avoid AF triggers and beyond. ¹¹

It is incumbent that we participate in developing these tools that will harness the opportunity of wearable devices. Health care providers and researchers should work with patients and consumer technology companies to ensure that algorithms are validated and accurate. Algorithms that seek to guide management of conditions (ie, anticoagulation, antiarrhythmic use) should be vigorously evaluated and regulated accordingly. Our health system itself must develop new mechanisms and infrastructure to manage this health data and “alerts,” so that there are effective channels for patients to share their results and communicate with their providers. We must demand and seek studies that show that these devices result in tangible benefits for their users and not unnecessary anxiety. Finally, effective integration of wearable technology into health care requires careful consideration of data privacy and confidentiality.

Wearable devices are becoming increasingly accessible for our patients and the general public, adding a new dimension in the detection and management of heart rhythm disorders. If used improperly, these devices may lead to anxiety, fear, and potentially increased health care resource usage. Health care providers, consumer technology companies, and users alike must be aligned in ensuring that these devices are accurate and rigorously validated, that patients and users are well educated in their use, and that these devices are thoughtfully integrated into our health system to improve care for those living with arrhythmia.

Disclosures

The authors have no relevant disclosures, funding, or conflicts of interest.