Abstract
Background:
The field of eHealth is evolving rapidly and can provide personalized support for people with chronic conditions. Despite these developments, the optimal personalization strategy remains uncertain, raising the question what works best for whom? In addition, the personalization strategy used is often poorly described. To improve the effectiveness of personalization, a method is needed to compare existing personalization strategies. Such a method can consist of a framework that describes and categorizes existing interventions. The categorization, using the framework, can be used to analyze which aspect of personalization is linked to more effective interventions. Furthermore, the overview provided by the framework can help developers choose and describe the personalization strategy consistently.
Existing frameworks that describe the personalization lack the depth needed to thoroughly compare interventions and link the strategy to effectiveness. To address this gap, this research aims to develop a framework to fully analyze and categorize the personalization strategy. This detailed description allows for comparison of different eHealth interventions to bring the field of personalized eHealth forward.
Methods:
A mixed-methods approach was used to construct the framework, integrating literature review and expert discussions. The development occurred in two phases. In the first phase, an initial framework was iteratively designed through a synthesis of existing frameworks and insights from five eHealth experts. In the second phase, an interactive panel session will be conducted to test and refine the framework. Panelists will be presented with five real-world cases and apply the framework to each eHealth interventions. The resulted descriptions are compared for inconsistencies. Together with feedback gathered during the panel session, the framework is finalized.
Findings:
The literature review has shown that most papers defining personalization categories had some variation on input, method and output categories. Klooster et. al.’s framework had the most defined subcategories. Only two papers categorized the extent of the personalization. The iterative discussion with the experts in the first phase resulted in the framework with three main categories and 19 subcategories. The first category evaluates the type of input variables (e.g., demographics, behavior, etc.) used for personalization and how it is obtained. The second category examines how this input is processed to generate personalized outputs (e.g., rule base, self-learning, retrospective learning, etc.) The final category assesses the output aspects of an intervention (e.g., content, timing and representation) that are personalized and to what extent. Within this category, the output is further divided and graded across five levels of personalization (e.g., group-based, individual-dynamic-based etc.). The input obtained from the panel session will provide insight into consensus of the developed framework and possible adaptations.
Discussion:
The developed framework serves as a structured checklist to categorize personalization strategies within eHealth interventions. By applying this framework, researchers and developers can systematically compare different interventions and link personalization approaches to intervention outcomes. Additionally, the framework can guide eHealth developers in selecting an appropriate personalization strategy that aligns with their intervention goals. This framework has the potential to advance the field by facilitating more precise and effective personalization in eHealth interventions.
The field of eHealth is evolving rapidly and can provide personalized support for people with chronic conditions. Despite these developments, the optimal personalization strategy remains uncertain, raising the question what works best for whom? In addition, the personalization strategy used is often poorly described. To improve the effectiveness of personalization, a method is needed to compare existing personalization strategies. Such a method can consist of a framework that describes and categorizes existing interventions. The categorization, using the framework, can be used to analyze which aspect of personalization is linked to more effective interventions. Furthermore, the overview provided by the framework can help developers choose and describe the personalization strategy consistently.
Existing frameworks that describe the personalization lack the depth needed to thoroughly compare interventions and link the strategy to effectiveness. To address this gap, this research aims to develop a framework to fully analyze and categorize the personalization strategy. This detailed description allows for comparison of different eHealth interventions to bring the field of personalized eHealth forward.
Methods:
A mixed-methods approach was used to construct the framework, integrating literature review and expert discussions. The development occurred in two phases. In the first phase, an initial framework was iteratively designed through a synthesis of existing frameworks and insights from five eHealth experts. In the second phase, an interactive panel session will be conducted to test and refine the framework. Panelists will be presented with five real-world cases and apply the framework to each eHealth interventions. The resulted descriptions are compared for inconsistencies. Together with feedback gathered during the panel session, the framework is finalized.
Findings:
The literature review has shown that most papers defining personalization categories had some variation on input, method and output categories. Klooster et. al.’s framework had the most defined subcategories. Only two papers categorized the extent of the personalization. The iterative discussion with the experts in the first phase resulted in the framework with three main categories and 19 subcategories. The first category evaluates the type of input variables (e.g., demographics, behavior, etc.) used for personalization and how it is obtained. The second category examines how this input is processed to generate personalized outputs (e.g., rule base, self-learning, retrospective learning, etc.) The final category assesses the output aspects of an intervention (e.g., content, timing and representation) that are personalized and to what extent. Within this category, the output is further divided and graded across five levels of personalization (e.g., group-based, individual-dynamic-based etc.). The input obtained from the panel session will provide insight into consensus of the developed framework and possible adaptations.
Discussion:
The developed framework serves as a structured checklist to categorize personalization strategies within eHealth interventions. By applying this framework, researchers and developers can systematically compare different interventions and link personalization approaches to intervention outcomes. Additionally, the framework can guide eHealth developers in selecting an appropriate personalization strategy that aligns with their intervention goals. This framework has the potential to advance the field by facilitating more precise and effective personalization in eHealth interventions.
Original language | English |
---|---|
Pages | 22-23 |
Number of pages | 2 |
Publication status | Published - 11 Jun 2025 |
Event | 14th Supporting Health by Technology Conference 2025 - U Park Hotel, University of Twente, Enschede, Netherlands Duration: 10 Jun 2025 → 11 Jun 2025 https://www.healthbytech.com/ |
Conference
Conference | 14th Supporting Health by Technology Conference 2025 |
---|---|
Country/Territory | Netherlands |
City | Enschede |
Period | 10/06/25 → 11/06/25 |
Internet address |