Where People Age in Place with Dignity

DREAMS Ecosystem Platform: Network Structure & Participant Engagement Pathway (Digital Rural-urban Ecosystem for Active Management & Successful Aging)

What the comparison is designed to show

• Health outcomes by intervention intensity — is monitoring alone sufficient, and for whom?
• Cost-effectiveness of each tier, to inform a scalable model for Year 3
• Whether urban and rural participants respond differently to the same programme

The platform being built alongside

• Three tiers: data collection → risk prediction engine → tailored intervention output
• Built during this year's pilot, informed by what the intervention teaches us
• Full operation is the goal for Year 3 (2027)

Three implementation problems identified

• Education bias in cognition. Three-word recall gave a ~50% false-positive rate; specificity fell to 45.7% among those with elementary education or less — 72.8% of the sample.
• Incorporation bias in locomotion. The chair-rise item overlaps SPPB; removing it dropped AUC from 0.768 to 0.670.
• Vitality under-detection. Adding BMI ≤ 20 kg/m² raised sensitivity from 59.0% to 68.2% with no specificity cost.

Prevalence–structure discordance

Prevalence looked similar across urban and rural sites — almost nothing else did. Rural locomotion sensitivity was 85.5% against 53.2% in urban residents, using the same item and the same protocol.

Conversely, urban residents were far more likely to be detected for cognition (adjusted OR 4.82) and vitality (OR 4.10). A single uniform national screening protocol does not perform uniformly.

What a session looks like

• Groups of 8–16, meeting twice weekly at the village senior centre
• Physical activity, nutrition, cognitive training, and social participation
• A 12-week cycle of 24 sessions, with pre-, mid- and post-assessment
• Open to all older adults at the centre, not only those screened as frail

Tailored centre by centre

Baseline assessment is returned to each centre as a profile of its own older adults — depression, cognition, fall risk, nutrition, grip strength, hearing and loneliness — so that every site emphasises what its participants actually need. One centre may be appetite-driven, the next dominated by fall risk.

Delivering at the centre removes the travel barrier a clinic-based programme would impose, and places the intervention inside an existing social structure, so taking part is a continuation of daily life rather than an appointment.

Grounded in the cohort

The intervention it feeds was designed from the team’s study of 464 older adults in Gyeongnam, which found that cognitive function and psychological capacity mediate the path from frailty to dependence in daily living. That is why the programme addresses cognition, nutrition and psychological domains alongside physical activity, rather than exercise alone.

Aligned with national policy

The work sits alongside Korea’s Integrated Care Support Act, which took effect in March 2026 and links medical care, long-term care and community support around the local area. The dashboard is intended as the mechanism that detects health risk early and routes an older adult to the service they need.