This cluster randomized controlled trial pilot study (ISRCTN61225589) assessing the effect of additional distance technology-based rehabilitation on patients with CVD was conducted between September 21, 2015, and November 30, 2017. The study protocol was approved by the Ethics Committee of the Central Finland Health Care District. The 12-month CR program was executed in groups of 10 rehabilitees each, which is standard practice at the Peurunka rehabilitation center. Sample size was defined by the rehabilitation groups that began the rehabilitation during the years 2015-2016. Physical activity outcomes were measured 3 times during the intervention: at baseline, 6 months, and 12 months. Participants or caregivers were not blinded to the intervention. However, the physical activity outcome assessor (author SH) and the researchers (authors TP and AK) who performed the statistical analyses of the results were blinded to the treatment allocations of the groups.

Participants were recruited from among patients with CVD who attended the CR program at a rehabilitation center in Finland. The eligibility criteria of the participants included age (18 years or older), diagnosed cardiovascular risk factors, angina pectoris with physical working capacity limitations, myocardial infarction, coronary artery bypass graft surgery, or coronary angioplasty. Exclusion criteria included musculoskeletal disorders, cognitive or memory impairment, or if the independent use of computer or remote technology application was not possible.

Participants had been allocated into 6 groups by officers of Social Insurance Institution of Finland, who also scheduled the inpatient periods for each group. Officers in charge of the allocation were not informed about the research project. Randomization was designed by three members of the research group (authors HK, TS and AK), and it was completed before interventions or any of the rehabilitation courses started by two members of the research group (authors HK and TS) and one person outside the research group generating the random allocation sequence.

Participants were cluster randomized in cluster pairs (1 and 2, 3 and 4, 5 and 6) by picking up numbered papers (either “A” or “B”) that issued to which treatment allocation the group belonged. Pairwise randomization of clusters bypassed any systematic effect bias due to the season in which the rehabilitation was conducted. The groups of rehabilitees were randomized to control and experimental clusters, one of each starting in autumn, winter, and spring. After randomization, rehabilitation center staff individually enrolled participants who provided written consent in the intervention. Participants were assigned to the interventions by two members of the research group (authors HK and TS).

Cardiac rehabilitation courses were arranged by the Social Insurance Institution of Finland, and in this study, the courses were held in and the data were collected from one rehabilitation center. The group-based courses were driven by a multidisciplinary team and consisted of meetings with a doctor; physiotherapist; nurse; and, optionally, a social worker, psychologist, or dietitian. The CR courses aimed to promote multidimensional self-efficacy by focusing on psychosocial factors related to coping with CVD in daily life. During the course, participants obtained information about CVD, counseling for managing daily activities with heart illness, and group discussions with peers, in addition to physiotherapy and aerobic exercise. During the inpatient periods in the rehabilitation center, participants performed health- and functioning-related tests.

A 12-month rehabilitation course, that both groups attended, consisted of three 5-day inpatient face-to-face rehabilitation periods in the rehabilitation center (at the beginning of the study, at month 6, and at month 12). The rehabilitation program promoted adaptation to life with CVD. Promoting physical activity was one part of the content, which aimed to reduce the barriers CVD rehabilitees have with respect to exercising [11]. Participants were provided information on the health benefits of physical activity based on the American Heart Association (AHA) recommendations [12] and encouraged to exercise in accordance with their condition. The walking goal was 10,000 steps per day. Individual goal setting of total physical activity followed the Specific, Measurable, Achievable, Relevant and Time-bound goal setting framework [13]. Between inpatient periods, participants followed their designated exercise programs. Participants in the conventional cardiac rehabilitation control group received the usual rehabilitation program, which was identical to the experimental group but did not contain technology. In the control group, the instructions on how to perform exercises, goal setting, and self-monitoring were paper-based for the control group. Goals were determined using the Goal Attainment Scale [14].

Participants in the additional distance technology experimental group received an additional distance technology program, which aimed to promote health and functioning-related lifestyle changes, such as physical activity and healthy nutrition. Participants set and monitored their individual goals and received instructions on how to perform exercises via the Movendos mCoach internet software [15]. The experimental group was instructed and motivated to perform physical activity self-monitoring with a wrist-worn Fitbit Charge HR [16] activity monitor. The Fitbit Charge HR displayed steps per day, energy expenditure, heart rate, and the quantity and quality of sleep. At the beginning of the intervention, the experimental group received 1.5-hours face-to-face support on the use of activity monitoring technologies from an information technology specialist, a nurse, and a psychologist. During the second inpatient period, 6 months later, the participants attended another 30-minute counselling session. In addition, the participants received a tutorial video on how to use Fitbit Charge HR and Movendos mCoach. The participants were contacted twice a month through the application. They received one automatic prompt to engage in physical activity, and the second contact was made by a physiotherapist who gave feedback of the activity recorded in the application.

Objective physical activity for all participants, in terms of steps per day, light physical activity (LPA), and MVPA were measured individually from each participant using a hip-worn Fitbit Zip (San Francisco, CA) accelerometer during the outpatient period for one week at 0, 6, and 12 months. Participants were instructed to use the device all the time while awake and not during sleep. The Fitbit Zip device automatically defines thresholds for the outcomes between LPA and MVPA. The days participants walked 1000 to 30,000 steps per day were included in the analysis. Subjective MVPA was measured using the International Physical Activity Questionnaire (IPAQ) [17], in which the participants recalled their activity over the last 7 days. During the inpatient periods, participant clusters were asked about the possible harms and other problems related to the use of technology and the intervention. Subjective adherence to the treatment was measured individually with a separate questionnaire (S. Hakala, unpublished data, October 2020). Adherence on using internet software was measured by analyzing the number of recordings made to the software, and the number of messages sent to the care provider during the 12-month intervention.

Two researchers (authors AK and TP) conduced the statistical analyses using IBM SPSS Statistics for Windows, Version 25.0 [18]. Outcome analysis was blinded, analyzing outcomes from group A and group B. After all the analyses were made, the encrypted treatment allocations of both groups were revealed. Absolute scores of outcomes measured with Fitbit Zip and IPAQ were used in analysis of changes in outcomes from baseline to 6 months, and from 6 to 12 months. Linear mixed models evaluated the effectiveness of the intervention on physical activity outcome. An independent samples t test (two-tailed) or Mann-Whitney U test analyzed differences between groups. Mean difference (MD) values described the differences between groups’ means, whereas positive values favored the experimental group and negative values favored the control group. In-group changes were analyzed based on tests of normality and using a paired samples t test (two-tailed) or Wilcoxon signed-rank test. Confidence intervals were reported with point estimates.

The participants’ mean age was 60 (SD 6.0; range 41-66) years, and 81% (48/59) of the participants were male (Table. 1). The level of physical activity in both groups was between low and moderate at baseline (see Multimedia Appendix 1). Bypass surgery was performed for 9 (15%) out of 59 participants, and angioplasty was performed for 75% (44/59) of participants either once (36/44, 61%), twice (5/44, 8%) or more (3/44, 5%). Cardiovascular operations were performed 3 to 6 months before the intervention (10/50, 20%), 6 to 12 months before the intervention (24/50, 48%), more than 12 months before the intervention (14/50, 28%), or during the intervention (2/50, 4%). The dropout rate during the intervention was 10% (6/59; Figure 1).

The experimental group had a greater increase in the minutes per week of LPA measured with the Fitbit Zip accelerometer compared to the control group (MD 324.2 min, 95% CI 77.4 to 571.0; P=.01; Figure 2), but there were no statistically significant differences between groups in the minutes per week of MVPA (MD 12.6 min, 95% CI –90.5 to 115.7; P=.82) or in steps per day (MD 1084.0 steps, 95% CI –585.0 to 2752.9; P=.20; Figures 3 and 4). Likewise, there were no statistically significant differences between groups with respect to the duration of MVPA measured by the IPAQ (MD 34.0 min, 95% CI –327.3 to 395.3; P=.85).

No statistically significant differences were observed between the experimental and control groups in the accelerometer-measured LPA minutes per week (MD –87.9 min, 95% CI –379.2 to 203.3; P=.54; Figure 2), MVPA minutes per week (MD 70.9 min, 95% CI –75.7 to 217.6; P=.33), or steps per day (MD 867.1 steps, 95% CI –2099.6 to 3833.9; P=.55; Figures 3 and 4). As such, there were no differences between groups in the MVPA minutes per week measured with the IPAQ (MD –292.2 min, 95% CI –755.6 to 171.2; P=.21).

During the 12-month intervention, participants in the experimental group made an average of 98 (SD 167; range 0-716) recordings regarding their physical activity in Movendos mCoach and sent an average of 6.4 (SD 6; range 0-26) messages to the care provider. The results from a subjective questionnaire indicated that on a scale from 1 to 7, in which 1 indicated “totally agree” and 7 indicated “totally disagree,” the participants almost or somewhat agreed they had actively used the Fitbit Charge HR (mean 2.4, SD 2.0). Participants somewhat disagreed that they had actively used Movendos mCoach (mean 4.9, SD 2.2). In total, 14% (4/29) of participants reported developing eczema as a result of the wristband of the Fitbit Charge HR. No other harms were reported.