Finding the people
no one is reaching.
The Missing Million helps social prescribers identify early signs of isolation among older adults — before crisis point — using weak signals already held across care and community systems.
"The people most at risk are often the least visible."
The problem
Isolation rarely arrives loudly. It accumulates quietly, in signals nobody reads together.
A missed appointment. A bereavement quietly noted in one system. A prescription left unfilled. Each looks ordinary on its own. Together they describe a person slipping out of view — months before anyone is alerted.
Older adults in England report often feeling lonely
Age UK, 2024
Higher mortality risk associated with chronic isolation
Holt-Lunstad meta-analysis
Average gap between first weak signal and care system flag
Lambeth pilot data
Of cases where prescription gaps preceded social withdrawal
The Missing Million model
How The Missing Million works
From scattered signals to earlier human contact.
Detect
The Missing Million reads weak signals already held across primary care, community and council systems — missed appointments, bereavement, living alone, reduced prescription pickup, declining engagement.
Prioritise
Transparent, explainable risk scoring surfaces residents most likely to disengage — with every contributing factor visible and traceable to its source.
Reconnect
Outreach guidance helps social prescribers make earlier, more sensitive human contact — informed by context, never replacing judgement.
A glance inside
Quiet, focused, operationally believable.
The Missing Million is built for the rhythm of real outreach work. A single clear priority list. Explainable scores. Sensitivity guidance for first contact. No alarms. No noise.
Launch the appTrust & ethics
Designed for trust, not surveillance.
The Missing Million supports human judgement. It does not replace it. Every score is explainable, every signal traceable, every outreach decision made by a person.
Human-in-the-loop
Every recommendation is reviewed by a social prescriber before any action is taken.
Explainable scoring
Each risk score breaks down into named, weighted factors traceable to their data source.
Data stays within the PCN
No data leaves your network. No external sharing. No third-party model training.
Not a diagnostic tool
The Missing Million prioritises outreach. It makes no clinical judgement and replaces no clinical pathway.
Manual override
Practitioners can dismiss, deprioritise or annotate any flag — and the system learns from it.
Ethical safeguards
Opt-out by default for excluded categories. Bias audits on every model release.
Outcomes