What we measure
Baseline at enrolment, re-tested at months 3, 6 and 12
- ✓Physical measures: weight & BMI, blood pressure, resting heart rate, waist measurement.
- ✓Quality of life: EQ-5D / SF-36 — the standard instruments health economists use, so results translate directly into cost-benefit language.
- ✓Activity level: IPAQ questionnaire plus device data.
- ✓Mental wellbeing: WHO-5 wellbeing index; PHQ-9 / GAD-7 for mood and anxiety.
- ✓Healthcare dependence: GP visits in the last 3 months, number of medications, sick days.
- ✓Monthly pulse: mood, energy and sleep on a 1–10 scale; any doctor visit or medication change.
The headline chart we are building towards
Illustrative target outcome — the shape of the story the dossier tells. Real lines come from the pilot.
Same people, month 0 versus month 12. When wellbeing rises as healthcare use falls, the fiscal argument writes itself.
Under the hood
Four tables. That’s the whole data model.
Think of four linked spreadsheets. Every analysis the dossier needs comes from joining them on the participant’s anonymous ID.
- ID — anonymous code, never a name (GDPR)
- Age, sex, municipality
- Consent date
- Group: exercise or comparison
- Weight, blood pressure, resting HR, waist
- EQ-5D, IPAQ, WHO-5, PHQ-9 scores
- GP visits, medication count, sick days
- Mood, energy, sleep (1–10)
- Doctor visit this month? Medication change?
- Sick days
- Gym check-ins
- Step count (phone sync)
- Sessions per week
Build small, protect everything
For the pilot, an existing research platform such as REDCap can hold tables 1–3 without writing custom software — a custom app comes later, if the project scales. Health data is the most protected category under GDPR: signed consent, anonymised IDs, a named data-protection officer and early legal advice are budgeted from day one, because they are the first things any government partner will ask about.
The measurement schedule
Exactly what is collected, and when
- Consent & demographics
- All physical measures, taken by trained staff
- Full questionnaire set (EQ-5D, IPAQ, WHO-5, PHQ-9/GAD-7)
- Healthcare-use recall: last 3 months
- Track assignment with the link worker
- Mood, energy, sleep (1–10 sliders)
- Doctor visit this month? Medication change? Sick days?
- One optional free-text line: “anything to tell us?”
- Push reminder; link worker calls after two missed check-ins
- Physical measures re-taken
- Full questionnaire set repeated
- Healthcare-use recall refreshed
- Month 6 doubles as the interim-results cut
- Gym check-ins from partner systems
- Step counts via Apple Health / Google Fit (opt-in)
- Walking-group attendance logged by leaders
Why these instruments: EQ-5D produces the quality-of-life values used in official cost-effectiveness analysis across Europe — it is the bridge from “I feel better” to “this saves the health system money.” IPAQ, WHO-5 and PHQ-9/GAD-7 are free, validated in Spanish, and short enough that people actually finish them.
Privacy by design
GDPR and ethics, done properly from day one
- ✓Legal basis: explicit informed consent for health-data processing (GDPR Art. 9), with scientific research safeguards. Participants can withdraw — and have their data deleted — at any time, in the app, in one tap.
- ✓Anonymisation: names live only in a separate, encrypted enrolment record held by the DPO. Everything researchers touch carries only the participant code.
- ✓DPIA before launch: a formal data-protection impact assessment — mandatory for health data at this scale — reviewed by the DPO and shared with institutional partners.
- ✓Ethics approval: the scientific committee submits the protocol to the Balearic research ethics committee (CEI-IB) before any participant is enrolled — the stamp that makes results publishable.
- ✓Data residency: all personal data on EU-hosted, certified research infrastructure. No transfers outside the EU, ever.
- !Why this section is this long: privacy is the first question every institutional partner asks and the one risk that could end the project. Answering it before being asked is itself a credibility strategy.
Building the app
Three phases, no wasted software
REDCap-based, branded front end
Enrolment, assessments and monthly check-ins on REDCap — the free platform built for exactly this kind of health study, GDPR-ready and trusted by ethics committees. A simple branded web app wraps it so participants see “Active Balears,” not a research tool. Weeks to deploy, not months.
Activity sync & dashboard
Opt-in step-count sync from Apple Health / Google Fit, gym check-in feeds from partner systems, and a live dashboard for the operations team: recruitment, retention and check-in rates per track, per municipality — the early-warning system for drop-out.
Custom native app
If EU4Health/Horizon funding lands and the pilot expands, then — and only then — a custom native app is justified: richer engagement features, multi-region support, integration with receta deportiva referral systems. The pilot’s data model transfers unchanged.
From data to dossier
The analysis plan, pre-registered
- 1Before-and-after: each participant’s month-0 versus month-12 scores — wellbeing, activity, physical measures, healthcare use. The headline chart.
- 2Against the comparison group: the same changes measured against people who didn’t join — the answer to “wouldn’t they have improved anyway?”
- 3Dose-response: do people who attended more improve more? If yes, the causal story strengthens dramatically.
- 4Economic translation: EQ-5D changes converted to quality-adjusted life years and avoided healthcare costs, plugged into the Catalan €1→€50 framework with Balearic numbers.
- 5Pre-registration: the full plan is published before results exist, so no one can claim we fished for favourable findings.
- ✓The output: one peer-reviewable paper, one plain-language dossier for the Govern, one open anonymised dataset — three formats, one body of evidence.