The effectiveness and mechanism of digital health in patients’ medication adherence intervention

Abstract

Poor medication adherence has long been a worldwide issue, and digital health have great potential in addressing this issue. However, whether digital health interventions can improve medication adherence effectively has not yet been reached in consensus and the mechanism by which they can improve medication adherence has not yet been fully explored in the literature. These gaps prohibit the stakeholders (e.g., healthcare providers and patients) in making the most of digital health in patient intervention. In this thesis, three interrelated sub-studies are conducted to systematically evaluate the effectiveness of digital health interventions and reveal the mechanism by which they can improve medication adherence. In study 1, network meta-analyses are conducted, in which 81 primary randomized controlled trial studies are identified from 6 mainstream databases, and both direct and indirect evidence from these studies were collected and synthesized. The results show that interventions with multi-component of digital health are generally more effective than those with single component; the optimal digital health interventions vary among different disease groups; consultation, communication, education and medication reminders are critical factors that can improve patients' medication adherence. In study 2, we investigate, from patient experience perspective, the mechanism by which digital health interventions can improve medication adherence. This study defines patients' two kinds of unique experience in the digital health context (i.e., social presence and user engagement) and examines the effect of these experiences on medication adherence. An empirical study is conducted, with both primary and secondary data collected from 376 valid respondents, on an online pharmacy service platform that contain all the critical factors that are identified in study 1. The findings show that both social presence and user engagement can be generated from emotional support and information support, but the path coefficients are significantly different; the two kinds of patient experience are both positively correlated with medication adherence, and these relationships are both contingent upon patients' education levels. Study 3 further explores the mechanism from patient empowerment perspective. Drawing upon the Conscious-Competence Model, health information competence and health consciousness are identified to characterize patient empowerment in the digital health services. Another empirical study is conducted on the same platform to verify the model and hypotheses, with quantitative data from the same sample in study 2 and 18 respondents' qualitative data collected from semi-structured telephone interviews. The findings show that health information competence and health consciousness are both positively correlated with medication adherence, and the effect of health information competence is stronger than that from health consciousness. We further observe that this relationship can be positively moderated by patients' duration of illness. Moreover, patient empowerment can partially mediate the relationship between patient experience and patients' medication adherence. This thesis has implications for both academic research and practice. Theoretically, it helps to clarify the inconsistency in the literature on the effectiveness of digital health interventions and reveals their mechanism in medication adherence intervention. Practically, this thesis provides valuable insights for healthcare professionals, service designers, managers, and policy makers to design, plan, and implement actions for making the most of digital health.