Wearable Devices and Health Care Use
- Patients with AF who use wearable devices for detection of AF are more than twice as likely to undergo an ablation procedure as comparable patients who do not use a wearable device.
- There is no evidence that use of wearable devices improves heart rate control in patients with AF.
- Younger patients are overall more likely to use wearable devices than older patients.
How do wearable devices impact clinical outcomes and health care utilization in patients with atrial fibrillation (AF)?
This retrospective, observational study aimed to 1) describe patients with AF who used wearable devices and 2) compare pulse rate and health care use between patients with AF who did and did not use wearable devices. Initially, investigators identified a total of 16,320 patients with AF, of whom 348 (2.13%) used a wearable device capable of detecting and monitoring AF. Characteristics of individuals who used and did not use wearables were evaluated using univariate logistic and linear regression. Subsequently, 125 patients who used a wearable device were each propensity matched with 4 controls who did not use a wearable device, creating a cohort of 625 patients with AF. For this cohort, investigators analyzed the 2 co-primary outcomes of 1) pulse rate and 2) a composite health care use score for a 90-day period. The latter measured health care actions that might be prompted by information from a wearable device, like evaluation and management encounters, cardioversion encounters, ablations, telephone encounters, rate-control medication orders, and rhythm-control medication orders. All data were extracted from the Enterprise Data Warehouse of the University of Utah Health from January 1, 2017, through December 31, 2019.
The most commonly used wearable devices were the Apple Watch, Fitbit, and Kardia. The analysis of 16,320 patients showed that, compared to patients who did not use wearables, patients who did use wearables were generally younger (mean age 64.0 vs. 70.0 years; p < 0.001) with lower CHA2DS2-VASc scores (mean 3.6 vs. 4.4; p < 0.001), fewer comorbidities, and higher socioeconomic means. Prior to the first clinic note mentioning use of a wearable device, they had more evaluation and management encounters (mean 7.2 vs. 4.4; p < 0.001) and telephone encounters (median 13.0 vs. 1.0; p < 0.001) than did nonusers. In addition, those who used wearable devices were significantly more likely to have had previous ablation, cardioversion, rate-control medication order, rhythm-control medication order, anticoagulant medication order, or longer time span since diagnosis of AF (755.8 days vs. 481.0 days), all p < 0.001. Among the cohort of 625 propensity-matched patients, overall baseline characteristics were similar. Pulse rates (mean 75.01 vs. 75.79; median 73.68 vs. 75.03) were similar between users and nonusers during the 90-day follow-up period. Mean composite health care use scores were higher in users (3.55 vs. 3.27; p = 0.04), although median interquartile range composite use health scores were not statistically different between the 2 groups. A greater proportion of users underwent ablation procedure during the follow-up period (17.6% vs. 7.4%; p < 0.001). In the 90-day follow-up period, no statistically significant difference was observed in the number of evaluation and management encounters, cardioversion encounters, telephone encounters, rate-control medication orders, or rhythm-control medication orders.
Many patients with AF use wearable devices for self-directed management, which may be in conjunction with a health care team. In this study, mean and median pulse rates did not differ between a propensity-matched cohort of patients who did and did not use wearable devices. Use of these devices is associated with increased overall AF-related health care use; the frequency of ablation procedures more than doubled among those who used wearable devices. Findings support the need for randomized controlled trials to assess the impact of wearable devices on clinical outcomes and health care use among patients with AF.
Several wearable devices have been approved by the US Food and Drug Administration for the detection of AF. However, most of the published studies of these devices focus on their use as screening tools rather than as aids to self-management in patients with a previous diagnosis of AF. For this reason, additional research regarding use of wearable devices in patients with a known diagnosis of AF is warranted.
Keywords: Mobile Applications, Monitoring, Ambulatory, Technology, Wireless Technology, Monitoring, Physiologic, Atrial Fibrillation, Self Care, Delivery of Health Care, Patient Acceptance of Health Care, Patient Care Team, Heart Rate
< Back to Listings