The effect of telemonitoring interventions on health outcomes in individuals with type 2 diabetes, heart failure and coronary artery disease: Results from three randomized controlled trials

The increasing burden of chronic conditions creates a need to develop more effective approaches to improve management and health outcomes of chronic conditions. Information- and communication technologies provide tools to promote the management of chronic conditions.
This thesis presents results of three randomized controlled trials that assessed the effect of telemonitoring interventions on health outcomes in individuals with type 2 diabetes (T2D), heart failure (HF) and coronary artery diseases (CAD). In all studies, telemonitoring involved self-monitoring of chronic condition related health parameters on a weekly basis, and sharing these data with healthcare professionals using a mobile phone. In addition, each study had a specific patient decision support component linked to the telemonitoring data. Study I, the Mobile Sipoo study, was conducted at the healthcare center of Sipoo, and included 51 patients with T2D who were followed for 10 months. The participants in the intervention arm recorded their weight, blood pressure, steps and/or blood glucose. These data were further linked with an automatic feedback system that provided patients with real-time, behavioral theory-based feedback messages that summarized the telemonitoring data, and motivated patients and provided them with behavioral skills to strengthen their self-management practices. The primary aim of Study I was to improve glycemic control, measured as HbA1c, and decrease blood pressure. Study II, the Heart at Home study, was conducted at the Cardiology Outpatient Clinic of Helsinki University Central Hospital and included 94 patients with systolic heart failure. The intervention participants were instructed to monitor their weight, blood pressure, heart rate and symptoms, and the data were linked with real-time, short feedback messages that summarized the data in relation to pre-specified individual target values. The primary aim of Study II was to reduce HF-related hospitalizations during the 6-month follow-up. Study III, the Renewing Health study, was conducted at healthcare centers in South Karelia and included 519 patients with T2D (n=250), CAD (n=227) or HF (n=42). Participants monitored their weight, blood pressure, steps and/or blood glucose. In addition, each participant received individual health coaching calls every 4 to 6 weeks to empower the patients and to teach them appropriate self-management skills tailored for each condition. The primary aim of Study III was to improve the health-related quality of life in all patients, and to reduce HbA1c in patients with T2D. In all studies the control group consisted of patients receiving standard care. As adherence is a prerequisite for achieving the intervention effects, adherence to the telemonitoring interventions was investigated in detail using the log files, and was defined as a percentage of weeks including at least one health parameter recorded. Analyses were performed according to the intention-to-treat principle.
In all studies, adherence to the weekly telemonitoring was moderate (Study III) to high (Study I and II) with the median percentage of adherent weeks being 93%, 90% and 66% in Study I, II and III, respectively, without major attrition in time. The telemonitoring intervention in Study I demonstrated a statistically significant improvement in glycemic control by reducing HbA1c by 0.44 percentage points. However, the blood pressure levels did not differ between the treatment arms. In study II, the telemonitoring intervention did not significantly reduce HF-related hospital admissions but, in fact, the utilization of the healthcare resources increased with the number of appointments and calls to the HF-nurse being 2-5 times higher in the telemonitoring arm, and more unscheduled visits to the cardiology clinic. A combination of telemonitoring and health coaching in Study III did not improve the health-related quality of life in patients with T2D, HF or CAD. Neither did it reduce HbA1c in patients with T2D.
Sustained, fairly high adherence seen in all studies suggests that weekly telemonitoring of health-related parameters is feasible. Nevertheless, high adherence does not guarantee positive health effects. Two of the studies showed no improvement in health outcomes although participants were actively involved with telemonitoring. This indicates that telemonitoring as such might not be effective in improving chronic disease outcomes. Positive health effects were seen only in study I, where the individuals received real-time, behavioral-theory based feedback messages that summarized the TM data, and motivated and guided patients to take actions to promote self-management. Putting the results together, the findings of this work support earlier research findings on the importance of grounding interventions on behavioral theory and providing timely feedback with enriched content to promote self-management and further improve the health outcomes of individuals with chronic conditions. However, telemonitoring interventions might increase the use of healthcare resources, especially personnel resources by requiring more time of the responsible nurse. Thus, sufficient resources should be ensured and the benefits gained evaluated in the light of other findings. Telemonitoring interventions should be carefully designed to target patients who are likely to adhere to them and likely to benefit from such interventions.