Public Defense: Pernille Lunde

Cardiovascular disease (CVD) is the most common non-communicable disease (NCD) and accounts for a burden of morbidity and disability worldwide. Exercise capacity measured as peak oxygen uptake (VO2peak) is an independent predictor of both cardiovascular risk and cardiovascular death as well as all-cause mortality.

Cardiac rehabilitation (CR) is an established part of secondary prevention for patients with cardiac diseases, and its beneficial effects have been well demonstrated. However, adherence to treatment is the most important modifiable factor that compromises treatment outcome, and adherence to healthy behaviour adopted in CR is challenging for many patients.

Aims

The overall aim of this research project was to develop new evidence-based knowledge on the use of a smartphone application (app) to improve patients´ adherence to healthy behaviour post-CR. More specifically, the first aim of this project was to review and assess the effectiveness of the use of apps for lifestyle improvements in patients with NCDs.

The second aim was to prepare for a randomized controlled trial (RCT) by assessing the feasibility of using an app to promote adherence to healthy behaviour post-CR and by providing a detailed description of the protocol used. Finally, the third aim was to evaluate the effects of conducting individualized follow-up with patients via an app for one year post-CR.

Methods

This thesis consists of four papers, and different designs are employed. The first paper is a systematic review and meta-analyses. A systematic search of five medical databases was conducted to evaluate the evidence base regarding the effectiveness of app-based interventions to promote lifestyle changes in patients with NCDs. Outcomes of interest included physical activity, physical fitness, modification of dietary habits and quality of life (QoL).

Risk of bias was assessed using the Cochrane risk of bias tool. The included studies allowed for pooled analyses of glycated haemoglobin (HbA1c). Heterogeneity was tested using the I2, and the meta-analyses were graded. The second paper presents a feasibility study with an experimental pre- and post-test single-arm design with a duration of 12 weeks. Included patients received weekly individualized monitoring and feedback via an app based on their own goals post-CR.

The feasibility outcomes assessed included recruitment rate, adherence to the app, resource requirements and efficacy regarding the ability to detect a change in QoL, health status and perceived goal achievement. The ceiling and floor effects in these outcomes were also evaluated. Criteria for success were pre-set and were based on adherence to the app and satisfaction with the technology.

In the third paper, a study protocol for the RCT is presented to enhance the transparency of the project and further reduce publication bias. The fourth paper presents a single-blinded two-armed RCT. Patients were recruited from three CR programmes at two CR centres in eastern Norway. The intervention group received individualized monitoring and feedback via an app based on their own goals. The control group received care as usual.

Assessments were carried out at baseline (post-CR) and at a one-year follow-up. The primary outcome for examining the effect of the intervention was VO2peak. Secondary outcomes included exercise performance evaluated as time to exhaustion, peak incline and peak velocity, bodyweight, blood pressure, lipid profile, triglycerides, exercise habits, QoL, health status and self-perceived goal achievement.

Results

The systematic review resulted in nine eligible papers, of which seven were included in the meta-analyses. Seven studies included diabetes patients only, one study included cardiac patients only and one study included both diabetes and cardiac patients. The meta-analyses demonstrated a significant overall improvement in HbA1c for the intervention, with low heterogeneity in the short-term and without heterogeneity in the long-term, compared to controls.

The quality of evidence was graded as low and moderate for the short- and long-term, respectively. The feasibility study showed that 71 percent of patients completing CR were eligible for the study. All 14 patients included used the app throughout the study, and satisfaction with the technology was high. The ceiling effect was present in more than 20 percent of the patients in several domains of questionnaires evaluating QoL.

Several points of guidance from the feasibility study were adopted and incorporated into the final design of the RCT. The RCT showed a statistically significant improvement in VO2peak, exercise performance, exercise habits and self-perceived goal achievement in the intervention group compared to the control group.

Conclusions

Individualized follow-up via an app is a feasible and effective method of promoting adherence to healthy behaviour post-CR. Further, our review demonstrated limited research of the use of apps for lifestyle improvement in NCDs other than diabetes. For diabetes, the use of apps seems to improve lifestyle factors, resulting especially in decreased HbA1c.

More research in this field is needed, particularly in terms of app-based interventions that promote adherence to healthy behaviour with consideration for the complexity of the process of adherence, long-term adherence and implementation.