Dynamic Indicators of Adherence and Retention in Digital Health Studies: A Longitudinal Observational Analysis from the Brighten Study
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918148089839616 |
|---|---|
| author | Hamitouche, Dylan Barkat, Youcef Parekh, Deven Hammer, Eva Benrimoh, David |
| author_facet | Hamitouche, Dylan Barkat, Youcef Parekh, Deven Hammer, Eva Benrimoh, David |
| contents | Background: Effective use of mobile health technologies requires high participant adherence and retention. However, remote digital health studies often face high attrition and low adherence, potentially introducing bias and limiting generalizability. Objective: This study aims to identify longitudinal indicators of participant retention and adherence to develop strategies for improving data collection in digital health studies and understanding how cohorts are shaped by participant withdrawal and non-adherence. Methods: We conducted analyses on the Brighten study, a smartphone-based randomized controlled trial evaluating apps for depression treatment. Participants were asked to complete seven digital questionnaires regularly. Outcomes included adherence (questionnaire completion), engagement (post-baseline participation), and retention (continued participation over time). We analyzed relationships between these outcomes, static factors (e.g., demographics, average questionnaire scores) and dynamic factors (e.g., questionnaire score changes over time). Results: Of 2,201 participants, 1,093 completed at least one non-baseline questionnaire (median completion rate: 37.6%). Adherence was higher among participants with lower average depression severity (P<.001) and those perceiving improvement (P=.001). Demographic factors significantly influenced adherence and engagement. Participants with greater baseline depressive symptoms were more likely to withdraw before completing non-baseline questionnaires (t=-2.53, P=.01). However, symptom improvement was linked to better adherence (U=127,084; P<.001) and retention (HR=0.78, P=.002). Conclusion: Clinical trajectories and perceived improvements in depressive symptoms are key indicators of engagement, adherence, and retention. These findings may enhance data interpretation and inform strategies to boost retention and adherence in future trials. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_00942 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Dynamic Indicators of Adherence and Retention in Digital Health Studies: A Longitudinal Observational Analysis from the Brighten Study Hamitouche, Dylan Barkat, Youcef Parekh, Deven Hammer, Eva Benrimoh, David Quantitative Methods Background: Effective use of mobile health technologies requires high participant adherence and retention. However, remote digital health studies often face high attrition and low adherence, potentially introducing bias and limiting generalizability. Objective: This study aims to identify longitudinal indicators of participant retention and adherence to develop strategies for improving data collection in digital health studies and understanding how cohorts are shaped by participant withdrawal and non-adherence. Methods: We conducted analyses on the Brighten study, a smartphone-based randomized controlled trial evaluating apps for depression treatment. Participants were asked to complete seven digital questionnaires regularly. Outcomes included adherence (questionnaire completion), engagement (post-baseline participation), and retention (continued participation over time). We analyzed relationships between these outcomes, static factors (e.g., demographics, average questionnaire scores) and dynamic factors (e.g., questionnaire score changes over time). Results: Of 2,201 participants, 1,093 completed at least one non-baseline questionnaire (median completion rate: 37.6%). Adherence was higher among participants with lower average depression severity (P<.001) and those perceiving improvement (P=.001). Demographic factors significantly influenced adherence and engagement. Participants with greater baseline depressive symptoms were more likely to withdraw before completing non-baseline questionnaires (t=-2.53, P=.01). However, symptom improvement was linked to better adherence (U=127,084; P<.001) and retention (HR=0.78, P=.002). Conclusion: Clinical trajectories and perceived improvements in depressive symptoms are key indicators of engagement, adherence, and retention. These findings may enhance data interpretation and inform strategies to boost retention and adherence in future trials. |
| title | Dynamic Indicators of Adherence and Retention in Digital Health Studies: A Longitudinal Observational Analysis from the Brighten Study |
| topic | Quantitative Methods |
| url | https://arxiv.org/abs/2412.00942 |