Moving eHealth powered medication adherence interventions from trial to real world as part of the SMILe implementation science project

Medication adherence is suboptimal in many chronic condition populations. Medication adherence is the “process by which patients take their medications as prescribed” and consists of three phases: initiation (i.e., the first intake of a newly prescribed medication), implementation (i.e., following the prescribed daily regimen) and discontinuation (i.e., permanent cessation of the prescribed medication regimen). Non-adherence to the prescribed medication correlates strongly with declining chronic disease management, patient health status, quality of life and ability to work, combined with increases in health care costs and mortality rates. Because of the high prevalence and potential consequences of medication non-adherence, much research now focuses on improving methods of defining and measuring medication non-adherence, assessing its determinants, and developing effective multi-level interventions to reduce it.
“The use of information and communication technologies in support of health and health-related fields,” i.e., eHealth, can support delivery of medication adherence interventions. eHealth itself is not a specific intervention, but rather a mode of delivery for educational, behavioral and psychosocial interventions. Evidence suggests positive effects of eHealth-delivered behavioral interventions on medication adherence, health and economic outcomes. However, for reasons attributed mainly to low quality and/or insufficient alignment between the proposed solution and the context, eHealth interventions are rarely sustained in daily clinical practice.
Despite decades of research on the phenomenon of medication adherence and its effective management (i.e., assessment and support) in a number of chronically ill populations, comparatively little has focused on allogeneic stem cell transplantation (alloSCT) recipients. For example, although medication non-adherence has been found to be common in this population and is clearly associated with adverse outcomes, detailed information on adherence patterns and effective interventions is lacking. In addition, as in other chronically ill populations, alloSCT recipients commonly lack successful implementation and sustained use of medication adherence evidence in real-world settings, i.e., in daily clinical practice. This widespread failure to implement empirical evidence is an enormous waste of research resources. More importantly, though, it prevents patients from benefiting from effective interventions.
The implementation of medication adherence evidence from research into real-world settings is complicated by research teams’ tendency to use self-developed (i.e., non-standard, untested) methods to support their implementation efforts. The process of translating medication adherence evidence into clinical practice calls for appropriate methods, such as those available from implementation science. As the "scientific study of methods to promote the systematic uptake of research findings and other evidence-based practices into routine practice," implementation science’s central focus is on bridging the gap between trials and real-world settings. Research teams who recognize the benefits of this focus can include implementation science principles in their projects from the earliest stages, then report their work in peer-reviewed journals to help others implement their evidence into clinical practice.
This eight-chapter thesis has the overall aim of supporting the translation of eHealth powered medication adherence interventions from trial completion to real-world use. It is part of the Integrated Model of Care after Allogeneic SteM Cell TransplantatIon faciLitated by eHealth (SMILe) implementation science project, which aims to develop/adapt, implement and evaluate its integrated care model at the University Hospitals of Freiburg, Germany (UKF), and Basel, Switzerland (USB). Chapters 1 and 2 introduce its topic and describe its objectives. Chapters 3 to 6 present its scientific articles; Chapter 7 is a policy brief; and Chapter 8 is synthesis and discussion.
Chapter 1 provides a general introduction to implementation science, medication adherence, eHealth and alloSCT. Chapter 2 describes the overall document’s aims.
Chapter 3 describes a systematic review to assess the type and extent of information available in published solid organ transplantation and alloSCT medication adherence intervention randomized controlled trials (RCTs) to support real-world implementation of adherence-enhancing interventions. To facilitate the assessment processes, the adapted Peters’ criteria of implementation-relevant information were used (e.g., context, stakeholder involvement, implementation strategies, implementation outcomes reported). Twenty-three studies were included, all of which focused on the implementation phase of medication adherence. Even the most frequent criterion — feasibility study (43%) — was reported in fewer than half of the selected studies, while the two least frequent — implementation strategies (4%) and process evaluation (4%) — were reported in a single study each. The rarity of implementation-relevant information reported in medication adherence RCTs in transplant settings hinders the implementation of adherence-enhancing interventions in real-world transplant settings.
Chapter 4 reports on the theory-driven development of an eHealth powered medication adherence module (implementation and persistence phase) as part of an integrated care model for alloSCT patients. The Behavior Change Wheel (BCW) both guided the development process and allowed its combination with implementation science and computer science principles. To work through the BCW's three steps, we employed findings from the SMILe study’s previously-conducted contextual analysis at the UKF, published evidence and input from stakeholders (i.e., patients, relatives, clinicians). After the third step, we added a fourth to develop the SMILe medication adherence module, including its digital component. To guide other researchers in developing eHealth powered behavior change interventions for other chronically ill populations and settings, this chapter provides a detailed description of this development process.
Chapter 5 describes a qualitative study to explore alloSCT patients’ medication adherence- and self-management-related experiences and strategies, and to learn their perspectives and preferences regarding Electronic Monitoring (EM) devices. In three focus groups, participants discussed their post-alloSCT medication management-related challenges, e.g., frequent schedule changes. They also discussed how, to overcome these challenges, they developed strategies to bolster their adherence, such as linking their medication intake to other habits. During these focus groups, the participants additionally had the opportunity to test three EM systems, i.e., the Medication Event Monitoring System (MEMS) Cap®, Helping Hand® and Button® devices. As participants particularly appreciated the small size of the MEMS Button®, which made it possible to store it in or next to existing pillboxes, this EM device was selected for the SMILe hybrid effectiveness-implementation RCT at the USB.
Chapter 6 reports on a systematic review conducted to provide an overview of published criteria for the description and evaluation of eHealth smartphone applications (apps). This was a first step towards developing a tool to systematize health care professionals’ assessment of eHealth apps for their patients. Overall, we identified 205 distinct criteria in 128 articles. No study included all criteria; and the labeling of criteria was inconsistent. To organize the criteria, based on the original dimensions reported in the included studies, we developed a conceptual framework for eHealth app evaluation. To fill potential gaps in the list, the research team conducted a consensus discussion. As a result, eleven new criteria were added, for a final total of 216. However, while this list is arguably comprehensive, it would be impractical for use in clinical practice. Therefore, our next step will be to use a Delphi consensus-building process to develop a much-shorter version of the instrument, as well as a user manual.
In Chapter 7, we present a recently developed, future-oriented policy brief on priority setting to implement medication adherence interventions in Switzerland. In collaboration with stakeholders (i.e., researchers, health care professionals, patients, health care administrators and policy makers), we agreed on five priorities for creating a more enabling ecosystem for the development, implementation and sustainability of medication adherence interventions: raising public awareness, preparing a multilevel ecosystem, translation from research to real-world settings, strengthening the patient-as-partner paradigm and monitoring medication adherence. Working on these priorities will help bridge the gap between research and clinical practice and reduce the burden of medication non-adherence in Switzerland and beyond.
Chapter 8 discusses how to move forward with medication adherence research in Switzerland and beyond. We position this dissertation’s work in relation to the priorities outlined by the policy brief described in Chapter 7, while also providing guidance to relevant stakeholders regarding possible next steps to strengthen medication adherence management at a national level.
This thesis advances medication adherence in the alloSCT population and beyond. Methodologically, we provide a clear description of approaches that support the implementation of eHealth powered medication adherence evidence into real-world practice. With this as its goal, it begins by applying the adapted Peters’ criteria as a helpful, innovative method of evaluating implementation-relevant information in medication adherence RCTs in transplantation. Second, it provides clear guidance on the development of eHealth powered interventions that use state-of-the-art methodologies to support both behavior change and real-world implementation. The methods used here can also be adapted to other populations, behaviors and settings. Third, we provide guidance on how a specific measurement method, i.e., EM, can be selected for optimal application in real-world settings. Fourth, we propose a new conceptual framework for the evaluation of eHealth apps in clinical practice. And fifth, we provide an overarching policy framework that will guide future efforts to support and tackle medication non-adherence as a major public health policy issue in Switzerland and beyond.
Content-wise, we highlight the enormous gap between the discovery of empirical evidence regarding medication adherence interventions in transplantation and the application of that evidence to real-world clinical practice. Further, we present the preferences of alloSCT patients regarding their use of an EM device, as well as the first medication adherence intervention to be delivered both by an alloSCT transplant team and by an eHealth app and implemented in real-world settings. Finally, we provide a comprehensive overview of eHealth app evaluation criteria published in the scientific literature.
All of our methods can serve as guidance for other researchers. Based on our findings and recommendations, this thesis’ sharp focus on moving eHealth powered medication adherence interventions from the trial environment to real-world practice gives it the potential to achieve real impact in real-world settings.