Factors Influencing Digital Health Engagement of Older Adults with Multimorbidity during a Longitudinal Trial

Paper Title

Factors Influencing Digital Health Engagement of Older Adults with Multimorbidity during a Longitudinal Trial

Publication Info

• Topic area: Digital health engagement in older adults with multimorbidity
• Keywords: Digital health, multimorbidity, older adults, self-management, longitudinal trial, engagement patterns, triage support, usability, human-centered design

Background and Problem

• Problem / challenge: Older adults with multimorbidity are underrepresented in digital health research, and their engagement with self-management technologies in real-world, longitudinal contexts is poorly understood. Existing tools often fail to account for the complexity of multimorbidity and face high attrition rates.
• Significance: Understanding engagement patterns in older adults with multimorbidity is critical for designing effective digital health interventions that support self-management, improve health outcomes, and reduce healthcare burden.
• Motivation and related work: Prior research has highlighted barriers such as low digital literacy, usability challenges, and lack of human support, but few studies have examined longitudinal engagement in this population. Models like supportive accountability suggest human support can sustain engagement, but scalability remains a challenge.

Solution

• Proposed approach: Mixed-methods analysis of engagement with the ProACT self-management platform during a six-month trial, comparing participants with and without triage nurse support.
• Novelty:
  1. Integration of objective usage data with qualitative insights to explore engagement patterns over time.
  2. Examination of engagement in older adults with complex health conditions, focusing on age, gender, and clinical support.
  3. Identification of multimorbidity-specific engagement mechanisms and barriers.
• Procedure and key techniques:
  • Participants aged 65+ with multimorbidity were randomized into two arms: Arm 1 (ProACT with triage nurse support) and Arm 2 (ProACT without support).
  • Quantitative data on vitals monitoring and app usage were analyzed using Bayesian regression models.
  • Qualitative interviews explored self-management routines, perceptions of the platform, and contextual factors influencing engagement.
  • Mixed-methods analysis triangulated findings to identify patterns, facilitators, and barriers.

Results

• Concrete findings:
  • Participants recorded an average of 19.35 vitals measures per week, with engagement decreasing slightly over time.
  • App usage averaged 3.72 days per week, with patterns of ‘hyper-users’ and ‘non-users.’ Usage declined initially but later stabilized.
  • Younger participants and those with triage nurse support recorded more vitals at the beginning of the trial.
• Advantage over baselines:
  • Triage nurse support provided reassurance, motivation, and guidance, helping sustain engagement and resolve uncertainties.
  • Participants perceived health benefits, such as improved self-management routines, weight loss, and increased activity.
• Experiments / evaluation:
  • Quantitative analysis tracked engagement patterns over 25 weeks using vitals monitoring and app usage data.
  • Qualitative interviews with 43 participants explored lived experiences, facilitators, and barriers to engagement.
• Limitations and future work:
  • Homogeneous sample from Ireland limits generalizability.
  • Attrition reasons and disengagement points need further exploration.
  • Future research will examine agentic AI for scaling triage support and analyze additional factors influencing engagement trajectories.

Summary

This study examined digital health engagement among older adults with multimorbidity using the ProACT platform during a six-month trial. Engagement was found to be non-linear, context-dependent, and influenced by symptom variability, usability, and life disruptions. Triage nurse support played a critical role in sustaining engagement by providing reassurance and guidance. Participants valued the platform for its perceived health benefits and ease of use, but barriers such as technology burden, illness, and external stressors disrupted engagement. The findings highlight the need for flexible, adaptive digital health systems that integrate human support and accommodate episodic use. Future work will explore scalable support models and deeper engagement dynamics.