Evaluating the Effectiveness and Sustainability of a Digital Prevention Measure: Protocol for a Mixed-Methods Study

BACKGROUND
Work ability decline among working-age adults is a major public health concern, requiring effective prevention strategies. While conventional in-person prevention programs show benefits, fully digital formats offer scalability and accessibility advantages. However, evidence on the effectiveness, sustainability, and behavioral mechanisms of digital-only prevention programs remains limited. OBJECTIVE This study aims to evaluate the effectiveness and sustainability of a fully digital multicomponent prevention program on work ability, health, and health behaviors, and to examine behavioral mechanisms through ecological momentary assessment (EMA).

METHODS
This quasi-experimental mixed-methods evaluation will compare three intervention formats: fully digital prevention program, digitally supported program (N=147), and conventional in-person program (N=98). Participants are working-age adults (18-65 years) eligible for German pension insurance prevention programs. The digital intervention consists of a 6-months app-based program including coaching sessions, webinars, and health modules targeting physical activity, stress management, and nutrition. ... mehrAssessments occur at five timepoints: baseline (T0), mid-intervention (T1, 3 weeks), post-intervention (T2, 6 weeks), and follow-ups at 3 months (T3) and 6 months (T4). Primary outcomes include the Work Ability Index and SF-12 Physical and Mental Component Summaries. Secondary outcomes include physical activity, motivation, stress, and eating behavior. Daily EMA captures intention, habit, affect, and perceived behavioral control. Data will be analyzed using repeated measures ANOVA/ANCOVA for effectiveness, multilevel modeling for EMA data, and thematic analysis for qualitative interviews.

RESULTS
Data collection for the fully digital prevention program is ongoing, while comparative baseline data from the digitally supported (N=147) and conventional in-person (N=98) programs are already available. Analyses will include repeated-measures ANOVA and ANCOVA to assess changes in work ability, physical and mental health, and related behaviors. EMA data will be analyzed using multilevel modeling to examine daily fluctuations in intention, affect, perceived behavioral control, and habit strength, providing insights into behavioral mechanisms, while qualitative interviews will be analyzed using thematic analysis to identify recurring themes related to user experiences, perceived effectiveness, and contextual factors influencing engagement and sustained behavior change.

CONCLUSIONS
This study will provide evidence on the sustainability of digital prevention effects through 6-month follow-up, address gaps in understanding mechanisms of behavior change through dual process theory tested with real-time EMA data, and contribute comparative effectiveness evidence for digital versus blended versus conventional intervention formats.