The Impact of Ecological Momentary Assessment (EMA) in Type 2 Diabetes: Examining study engagement, self-efficacy, self-management and emotional well-being

Abstract

Objective: To examine the role of self-efficacy, engagement with a smartphone app for ecological momentary assessment (EMA) and depressive symptoms on self-reported self-management behaviors in Type 2 Diabetes, using both quantitative and qualitative methods. ¶Participants and Methods: This study is a secondary analysis of data from a parent project which used repeated ecological momentary assessment of symptoms, affective states, and self-management behaviors among adults with Type 2 Diabetes, via a phone application known as “MyDay.” The sample was comprised of 62 participants with Type 2 Diabetes from the Albert Einstein College of Medicine/Montefiore Medical Center. Participants first completed self-report questionnaires, including self-ratings of oral diabetes medication adherence and blood glucose monitoring adherence, at a baseline study visit. Then, they used the MyDay app for 14 days (completing 3 surveys per day), before returning to the lab for a follow-up visit to complete self-report questionnaires and engage in a semi-structured, audio-recorded exit interview with study staff. First, bivariate correlations were run to evaluate the cross-sectional relationships between baseline study variables of interest. Pearson’s r correlational analysis was also run between baseline scores of the self-rating adherence measures and the follow-up self-ratings of adherence. A series of linear mixed effects models were run to evaluate the within-subjects changes in self-management across these two timepoints, examining the potential moderating impact of self-efficacy, depressive symptoms and percentage of completed EMA surveys over the monitoring period, as an indicator of engagement. Additionally, semi-structured exit interviews were transcribed and analyzed by a team of qualitative coders for themes related to change in self-management across multiple domains (diet, exercise, medication-taking, blood glucose monitoring, and checking one’s feet) over the course of study participation.¶Results: Baseline self-efficacy was correlated with self-ratings of oral medication adherence and blood glucose monitoring adherence at both the baseline and follow-up timepoints, such that greater self-efficacy was associated with greater adherence. Self-ratings of blood glucose monitoring adherence significantly improved from baseline to follow-up, however oral medication adherence showed no significant change over time. Within the linear mixed effects models, we did not find depressive symptom severity or self-efficacy to be significant moderators of the relationship between baseline and follow-up scores on either of the adherence measures. Survey completion percentage moderated the relationship between baseline and follow-up self-reported blood glucose monitoring adherence, such that participants who completed more surveys showed greater improvements in blood glucose monitoring from baseline to follow-up. Within the qualitative analyses, participants reported change in four of the five hypothesized domains of self-management: taking medications, blood glucose monitoring, adhering to exercise recommendations and adhering to diet recommendations, and participants attributed these changes to their participation in the EMA protocol. We found that these qualitative reports were associated with high self-efficacy, as participants with high self-efficacy scores were more likely to endorse change in self-management behaviors during their exit interview. ¶Conclusions: Participation in this 2-week EMA app study was associated with significant changes in self-reported diabetes self-management, both adherence to recommended blood glucose monitoring procedures (identified in the quantitative analyses) as well as adherence to dietary recommendations, exercise regimens, and medication taking (identified in the qualitative analyses). While it is not possible to determine any interventional effect of the EMA app as we did not have a control group in this study, participants who engaged with the app more over the 14 days (by completing more surveys) reported greater improvements in adherence to blood glucose monitoring (in quantitative analysis), as did participants with high baseline self-efficacy (in qualitative analysis). These findings, while limited due to a small sample size, potentially underpowered analyses, and study design without a control group, may have clinical implications for the development of future interventions to improve adherence to important self-management behaviors in this at-risk medical population.