Activity trackers: always a good thing? Three considerations when implementing activity trackers in cardiac rehabilitation

This invited commentary refers to ‘Clinician perspectives on activity trackers for cardiac rehabilitation: a multinational study’ by T.R. Pettersen et al., https://doi.org/10.1093/eurjcn/zvag072. Several systematic reviews show that the use of activity trackers can improve physical activity levels and physical fitness. This has been demonstrated in cardiac rehabilitation and in the secondary prevention of cardiovascular disease1–3 as well as in other clinical groups and in the general population.4 The use of activity trackers for increasing physical activity participation is therefore recommended in the 2021 European Society of Cardiology guidelines on cardiovascular disease prevention (class IIa recommendation, i.e. ‘should be considered’), and it can be expected that the class of recommendation will increase in future guideline editions, as additional high-level evidence has become available since. Despite this supporting evidence and guideline recommendations, the implementation of activity trackers in cardiac rehabilitation remains limited. Pettersen et al.5 address this issue through their survey of over 300 cardiac rehabilitation clinicians in Australia, Brazil, Canada, Norway, and the UK. Based on clinicians’ perceptions and experiences of activity trackers, Pettersen et al.5 highlight the importance of clinicians’ motivation for the use of activity trackers in cardiac rehabilitation and derive helpful recommendations for implementation. The premise behind the study by Pettersen et al. is that activity trackers are recommended and should be used more widely to support cardiac rehabilitation patients’ physical activity. While we fully agree with this premise, we would like to use this commentary to highlight three critical points which—in our observation—are perhaps less readily acknowledged in trials and systematic reviews of activity trackers, but which constitute important considerations for implementation in practice. Titles of systematic reviews such as ‘Effectiveness of wearable activity trackers to increase physical activity’4 may lead readers to assume that the devices themselves constitute the entire intervention, i.e. that it suffices to provide activity trackers to patients, or recommend that patients acquire activity trackers themselves. However, when we look up the details of primary studies included in systematic reviews, we find that activity trackers were mostly implemented in combination with other intervention components that were not delivered to the control groups. These other intervention components include personalized exercise prescription and physical activity goal setting with clinicians, and monitoring, motivational messages, and individual feedback on patients’ physical activity levels by clinicians (cf. supplementary data in the systematic review by Kaihara et al.3). Methodologically, this means that some portion of the effect we see in controlled trials can probably be attributed to these inputs and interactions with clinicians. In clinical practice, this means that these inputs and interactions need to be considered, and clinicians should be allocated the necessary time and resources. A common assertion is that activity trackers are not susceptible to recall and social desirability bias and therefore preferable over subjective measures of physical activity such as questionnaires and diaries.6 However, this does not mean that activity trackers always provide accurate measurements. The leisure and sports industries offer a vast array of consumer wearable devices that feature accelerometry for physical activity tracking and photoplethysmography for heart rate measurement. These devices display variable degrees of (in)accuracy depending on which parameter is measured, the measurement range within one single parameter, and the conditions of measurement.6 For example, a device may be reasonably accurate for measuring step count, but less accurate for detecting minutes of moderate-to-vigorous physical activity; a device may measure heart rate accurately in the lower range, but less accurately in higher heart rate zones7; and measurements under free-living conditions tend to be less accurate than laboratory-based measurements.6 Pettersen et al.5 rightly point out that measurement accuracy must be ascertained for each device, based on data provided by the manufacturer as well as independent scientific validation studies. Moreover, measurement inaccuracies also provide opportunities for clinicians to guide patients in ‘discovering’ their devices and detecting and interpreting inaccuracies. This includes reviewing the correct application of the wearable according to manufacturers’ instructions (e.g. a loose wrist-worn device will return less accurate heart rate measurements) and trying out strategies to self-check the accuracy of measurements (e.g. by comparing with a manual pulse measurement or own step count). A growing body of qualitative evidence offers insight into the patient experience of activity trackers,8 including studies from cardiac rehabilitation.9 These studies support an overall positive view of activity trackers but also report several adverse experiences that should be considered when implementing activity trackers in practice. Patients who notice that measurements are inaccurate and unreliable can lose trust in the device and feel frustrated and demotivated to use the device. When physical activity measurements are monitored by clinicians, a mismatch between device-measured and perceived physical activity (e.g. when the activity tracker fails to record physical activity) can result in pressure for patients to justify or explain themselves to clinicians. Of note, the negative impact of inaccurately low physical activity measurements on users’ mindset, self-esteem, and mental health has also been shown in experimental studies.10 When wearing an activity tracker for an extended period, aspects of comfort, compatibility with daily activities and clothing, as well as the aesthetic of the device should not be underestimated. And increased awareness of one’s physical (in)activity, not meeting physical activity recommendations and receiving reminders to be physically active, can elicit feelings of guilt, shame, worry, or a sense of failure, which can result in disengagement rather than an encouraging patient experience.8,9 Clinicians should be sensitive to these possible patient experiences and dynamics, and the implementation of activity trackers in cardiac rehabilitation should be embedded in a clinician–patient relationship based on shared decision-making and accounting for individual patient preferences and needs. In summary, the evidence base for effectiveness of activity trackers in cardiac rehabilitation is robust. As we now turn to implementing activity trackers in routine practice, the nuances of patients’ and clinicians’ perceptions and experiences become increasingly important to inform well-designed implementation strategies and practices that will realize the potential of activity trackers in routine care. Studies such as the survey by Pettersen et al.5 provide valuable guidance for implementers of activity trackers in cardiac rehabilitation. Stefan Tino Kulnik (Conceptualization, Writing—original draft [lead], Writing—review & editing [equal]), Veronika Leitner (Conceptualization, Writing—original draft, Writing—review & editing [equal]), and Dirk Leysen (Conceptualization, Writing—original draft, Writing—review & editing [equal]) No funding was received for this commentary. This commentary does not contain new data.