Adherence to long-term telemonitoring-supported physical activity in patients with chronic heart failure

Study design

All participants took part in a prospective, randomized, parallel group, multicenter controlled trial examining the effects of telemonitoring-supported exercise training in patients with CHF. The “HITS” study (“Heart failure, Individual exercise training, Telemonitoring, Selfmanagement”) was conducted as a collaborative project between Leipzig University (LU), Hannover Medical School (MHH), Leipzig Heart Center (LHC), AOKplus health insurances, IGES institute, DiaVention GmbH, and the Clinical Centers in Wolfsburg (CCW) and Chemnitz (CCC) in Germany. The study was registered in the German Clinical Trials Register under DRKS00019022 (28.05.2020). The institutional review boards of the MHH (No. 8786) and the medical faculty of the LU (479/19-ek) have given ethical approval for the study, and written informed consent was obtained prior to enrollment. The HITS innovation fund project aimed to establish a new model of care that included early diagnosis of early stages of CHF, the treatment of CHF patients according to current ESC and AHA guidelines^17,18, avoidance of hospitalizations, and improvement of treatment adherence, physical performance, and HrQoL.

Subjects

Participants were recruited through a series of information events, referrals from general practitioners, and newspaper and online advertisements. According to predefined inclusion criteria, patients older than 18 years of age with diagnosed CHF of the stages NYHA I, II, and III (New York Heart Association [NYHA] clinical symptom stages¹⁹, including transplants or implants were included in the study. Patients with both existing and newly diagnosed CHF were included. For the latter, N-terminal natriuretic pro-B peptide (NT-proBNP) level greater than 125 pg/ml was used as the cutoff for inclusion according to the ESC criteria¹⁸. Exclusion criteria were the presence of chronic kidney disease, chronic obstructive pulmonary disease, alcohol abuse or use of illegal drugs, active participation in other studies, and any physical or mental condition that precluded participation in an exercise intervention. Participants who did not show up for the examinations or were unwilling to use the study devices were excluded from the study.

All participants were randomly assigned (at a 1:1 ratio) to the exercise group (EG) or the control group (CG) (see Fig. 1). Classification into NYHA stages I, II, or III was based on existing diagnoses or was determined by the study physicians based on the severity of symptoms during PA¹⁹.

Study design and randomization.

For the present analysis, available data from all included patients from all study centers were evaluated.

Procedures

At each study center, anthropometric data (body weight, height, body mass index [BMI], waist and hip circumference) were assessed according to defined standard operation procedures (SOPs) after a general medical examination by a physician (including electrocardiogram, medical history, and physical examination). Fat-free mass, fat mass, and total body water as markers of body composition were estimated by segmental, multifrequency, bioimpedance analysis (MHH: InBody720; Biospace, South Korea; LU. CCC, CCW: Biacorpus RX 4004M; MediCal HealthCare GmbH, Germany).

Cardiac-specific laboratory parameters (NT-proBNP, creatine kinase [CK], isoenzyme creatine kinase-MB [CK-MB], C-reactive protein [CRP], interleukine-6 [IL-6], and a safety blood profile including electrolytes, hemoglobin, hematocrit, thrombocytes, and leucocytes) were obtained from a venous blood sample.

Exercise capacity (measured as peak power output in watts [W_max]) and maximum heart rate (HR_max) were assessed with stepwise cardiopulmonary exercise testing (CPET) on a bicycle ergometer (Ergoline P150, Bitz, Germany) until objective or subjective maximal exertion, or pathological findings for ergometry occurred. To ensure the accurate assessment of maximal exertion, we measured blood lactate levels and considered levels > 4 mmol/l as exhaustion. During each CPET, we measured the respiratory exchange ratio (RER) and used an RER value > 1.10 to indicate maximal effort. In each CPET we achieved O₂-ventilatory equivalent, which can be used as a valid sign of exhaustion when V̇E/V̇O₂ > 30–35. Unlike healthy individuals and athletes, where VO_2max is determined, cardiac patients often do not reach a VO₂ plateau at the end of a CPET. Therefore, we used the VO_2peak value and the maximal wattage as measures of performance. The FRIEND equation²⁰ and the Wasserman-Hansen algorithm²¹ were used to determine peak predicted relative VO₂, and percent peak predicted VO₂ (ppVO₂) was calculated as 100 × (measured relative peak VO₂/peak predicted relative VO₂). Subjective perceived exertion was assessed by the Borg-Scale²². The CPET started with a workload of 20 W or 50 W and was increased in 10 W or 17 W increments per minute. A spirometric system (Oxycon CPX, CareFusion, Würzburg, Germany) was used to measure breath-by-breath oxygen uptake (VO₂ [ml/min]), carbon dioxide production (VCO₂ [ml/min]), and ventilation. Heart rate and blood pressure were recorded and capillary blood samples were taken from the earlobe at rest, 1 min after the start, and every 3 min during the test to determine blood lactate concentrations (Ebio 6666, Eppendorf, Germany). Because we did not have the appropriate equipment at the Clinical Center Wolfsburg, we only performed a stepwise exercise test on a bicycle ergometer at this location without additional spirometry.

A comprehensive transthoracic echocardiography was performed in accordance with AHA/ESC guidelines (CCW: Vivid™ T9, GE Healthcare GmbH, Solingen, Germany; MHH: Vivid iq, GE Healthcare GmbH, Solingen, Germany). If an echocardiography was performed less than 8 weeks previously by a resident cardiologist, the echocardiographic values from the physician’s report were used.

The description of questionnaires used has been described in detail previously²³. Briefly, a 14-item questionnaire was distributed as a primary measure to estimate adherence to the Mediterranean Diet (MDS questionnaire)²⁴. The 23-item Kansas City Cardiomyopathy Questionnaire (KCCQ) was used to quantify physical limitations, symptoms, self-efficacy, social interference and HrQoL in patients with CHF²⁵. The Sedentary Behavior Questionnaire (SBQ) was used to assess the amount of time spent in nine behaviors (watching television, playing computer/video games, sitting while listening to music, sitting and talking on the phone, doing paperwork or office work, sitting and reading, playing a musical instrument, doing arts and crafts, sitting and driving/riding in a car, bus, or train). The nine items were completed separately for weekdays and weekend days²⁶. The results refer to the total number of hours spent in sedentary behavior during a 7-day week. At the study centers MHH and CCW, the Freiburger PA Questionnaire was distributed to calculate the total and exercise-related PA as metabolic equivalents of task (MET) hours per week²⁷. All questionnaires and examinations were carried out at the 6-month (V6) and the 12-month (V12) visits according to the baseline visit (V0).

Intervention

After randomization at V0, participants in the EG were provided with the following study devices:

• wearable activity tracker with accelerometer and GPS sensors (vivoactive 4, Garmin®, Garching, Germany) to record heart rate and daily step count,

• heart rate monitor (HRM-DUAL, Garmin®, Garching, Germany) to record heart rate during exercises,

• upper arm blood pressure monitor (boso medicus system, Bosch & Sohn GmbH, Jungingen, Germany) to measure resting blood pressure,

• personal scale (Trisa Electronics, Body Analyze 4.0, Triengen, Switzerland) to measure body weight,

• tablet computer (Lenovo® TB-X606X, Lenovo® Germany, Stuttgart, Germany) as an interface between the devices.

After testing the devices with the assistance of the study’s sports scientists, EG participants were introduced to the study application, called the “HITS App”, which was installed on the tablet. The HITS App is a native mobile application for android operating system. The user interface has been optimized for tablets. The App was developed by the DiaVention GmbH for the HITS study and included the following components: Home-based exercise training, information on general health improvement, disease prevention, and lifestyle changes (diet, exercise, and self-awareness). All measurement data transmitted by the devices via Bluetooth was summarized and graphically processed in the App. Both participants and study staff had access to the daily measured data. This feature helped to actively engage patients in the intervention and allowed the sports scientists to adjust the intervention at any time if necessary. Most importantly, the App served as a platform for the home-based exercise training videos, which were selected based on group affiliation (NYHA I, II, or III) and/or individual limitations. The videos averaged 30 min in length and were interval training in nature. The intensity of the exercises bodyweight training should be in the moderate to vigorous range. Based on the results of the CPET, a symptom-limited upper limit of heart rate was set by the physicians. The content included exercises to improve aerobic endurance as well as strengthening exercises for the trunk, upper and lower extremities. Throughout the intervention, the task was to maintain PA at the highest level possible, aiming for three exercise videos per week, taken into account individual limitations and preferences. Participants in the EG were instructed to measure their resting blood pressure and body weight on a daily basis as well as to answer questions about adherence to drug therapy and their sleep quality, and to transmit data from the wearable activity tracker to the App. All EG participants were given a wearable activity tracker at baseline to determine PA via steps per day and moderate-to-vigorous physical activity (MVPA) per week. All indoor and outdoor activities, that were done additionally to the exercise training videos via the App, could be recorded on the activity tracker and were presented in steps per day and MVPA.

To evaluate participants’ adherence to the exercise recommendations, the average number of completed exercise videos per week and the daily step count via the wearable activity tracker at V0 and at V12 was analyzed. For both parameters, a minimum of 13 valid and continuous weeks had to be recorded. We considered a daily step count of at least 500 steps as a valid day. We chose the delta of the last 4 weeks before the final exam and the first 4 weeks after the devices were issued as the dependent variable in a multivariate linear regression model. In order to be able to evaluate the progress and the adherence to the 12-month exercise intervention, we compared the average number of completed exercise videos in the first 4 weeks with the average number in the last 4 weeks. Data of the wearable activity tracker were used to determine the daily step count and the heart rate throughout the day. Since these data are only available for the EG, the Freiburger PA questionnaire was used to assess and compare self-reported PA over time and between EG and CG. To assess the sedentary behavior among study participants, the SBQ was evaluated at baseline and throughout the study.

Statistical analysis

First, the Kolmogorov–Smirnov test was used to test for normal distribution. Differences between the two groups were compared using the Mann–Whitney-U-Test (not normally distributed data), the Student t-test for unpaired samples (normally distributed data), or the chi-square test for frequency distributions, using Pearson’s correlation coefficient r as the effect size for the Mann–Whitney U test, Cohen’s d for the Student t-test, and Cohen’s ω for the chi-square test. An effect size of 0.1 indicates a small effect, 0.3 indicates a medium effect, and 0.5 indicates a large effect. Parametric values were reported as mean and standard deviation (SD); non-parametric values were reported as median values and interquartile range [IQR]. For descriptive analysis, absolute frequencies were calculated for categorical variables, and mean and SD for continuous variables. Univariate correlations between parameters were tested using Pearson’s or Spearman’s correlation coefficient depending on normality of data distribution.

Regression analyses were performed to identify parameters associated with changes in PA (MET-hours per week by questionnaire), everyday activity (step count per day by wearable activity tracker) and number of training videos (recorded by HITS App) after the 12-month intervention.

A one-way analysis of variance (ANOVA) was used to test for group differences between the subscales of the Freiburger PA Questionnaire and the SBQ at baseline, using eta squared η² as the effect size, where 0.01 indicates a small effect, 0.06 indicates a medium effect, and 0.14 indicates a large effect. Post hoc tests were corrected according to Bonferoni. An effect size is only reported if the corresponding test was significant. All statistical analyses were carried out per protocol. The type-I-error was set to 5% (two-sided). All statistical analyses were performed using IBM SPSS 28 Statistics (IBM Corporation, NY, USA).

Ethics approval and consent to participate

The study was performed in accordance with the Declaration of Helsinki and current guidelines of good clinical practice. The study was completed as a cooperation project between Leipzig University, Hannover Medical School, Leipzig Heart Center, AOKplus health insurances, IGES institute, DiaVentions GmbH, and the Clinical Centers in Wolfsburg, Chemnitz and Dresden (Germany). The study is registered under DRKS00019022 in the German Clinical Trials Register. The institutional review boards of Hannover Medical School (No. 8786) and medical faculty of the Leipzig University (479/19-ek) ethical approved the study, and written informed consent was obtained prior to inclusion of study participants.