Tracking the fraction of Hong Kong's frail elderly who receive Care Food-compliant meals.
Q4 2026 baseline: ~5% (~7,500 of 150,000 estimated). Target Q4 2028: 50% (~75,000).
First-of-kind population adoption metric in IDDSI-implementing economies. Methodology co-designed with HKCSS Care Food Working Group and HKU Swallowing Research Laboratory (partnerships proposed, pending HKEX Foundation × HKCSS Impact Funding Scheme 2026 award). Governed as open data for the Hong Kong eldercare ecosystem.
Quarterly target adoption rates (%) for the 24-month programme period. The baseline is set at 5% (Q4 2026), representing the pre-programme floor. Year 1 target is 25%; Year 2 target is 50%.
Chart shows quarterly targets. Actual quarterly reports publish at careez.org/adoption-index/q[n]-[yyyy] from Q1 2027 onwards. Intermediate quarterly values are programme targets, not forecasts; actuals will be published with 95% CI.
Programme targets from Q4 2026 baseline through Q4 2028. Intermediate targets reflect AI Assistant deployment milestones, RCHE rollout cadence, and family caregiver certification scale. All rates are targets; actuals published quarterly with methodology notes and confidence intervals.
| Quarter | Target Rate | Est. Beneficiaries | Key Milestones |
|---|---|---|---|
| Q4 2026 — Baseline | ~5% | ~7,500 | Pre-programme floor. Track A ~3,500 + Track B ~2,500 + Track C ~1,500. |
| Q1 2027 | 10% | ~15,000 | AI Assistant launch + first 10 RCHE deployments |
| Q2 2027 | 15% | ~22,500 | Family caregiver rollout begins |
| Q3 2027 | 20% | ~30,000 | 30,000 family caregiver registrations |
| Q4 2027 — Year 1 End | 25% | ~37,500 | 30 RCHEs + 5,000 certified carers + 300 Champions |
| Q1 2028 | 30% | ~45,000 | Cross-border FDH cert pilot begins |
| Q2 2028 | 35% | ~52,500 | 50 RCHEs cumulative |
| Q3 2028 | 42% | ~63,000 | HKCSS–HKU outcome report published |
| Q4 2028 — Year 2 End | 50% | ~75,000 | 60 RCHEs + 5,500 certified carers + 500 Champions + 150,000 family registrations (subset complies) |
The Hong Kong 照護食 Adoption Rate = (Number of HK frail elderly receiving Care Food-compliant meals in a given quarter) ÷ (Total HK elderly population with clinically indicated dysphagia or chewing difficulty). Reported as a percentage with 95% confidence intervals and disaggregation by setting (RCHE / community / home / hospital discharge).
Population scope: aged 60+ with ICD-10 R13 dysphagia diagnosis, or aged 60+ with EAT-10 score ≥ 3.
Breakdown: ~28,000–42,000 RCHE residents with dysphagia (40–60% of 70,000+ RCHE residents, per Baijens et al. 2016) + ~100,000–140,000 community-dwelling elderly (8–12% prevalence among 1.6M aged 65+).
Denominator refinement is a Year 1 deliverable via Hospital Authority dysphagia screening data integration.
Highest evidence. Meal sourced from a HKCSS Care Food Directory vendor (Sections A or D) or carrying a current HKCSS 照護食標誌 mark.
Baseline estimate: ~3,500 elderly receiving directory-vendor meals.
Operational evidence. Meal prepared by a carer with a current 照護食 Adoption Champion or AI Assistant-issued certificate; or prepared using Care Food Directory Section B products at recommended dosages; or verified by AI Assistant compliance check (image + texture analysis vs IDDSI).
Baseline estimate: ~2,500 elderly at programme-participating RCHEs.
Community-tier evidence. Meal prepared by a family caregiver who completed the AI Assistant cert pathway; self-reported via AI Assistant intake check-in.
Weighted at 0.7 of Track A/B evidence to account for self-report bias. Annual independent audit + HKU cross-validation.
Baseline estimate: ~1,500 community-dwelling elderly.
| Source | Track Contribution | Cadence |
|---|---|---|
| HKCSS Care Food Directory updates | Track A vendor coverage | Quarterly |
| 照護食標誌 certified products in market | Track A coverage | Quarterly |
| SeniorDeli AI Assistant aggregate logs (anonymised) | Track B + Track C verification counts | Real-time → quarterly report |
| HKCSS partner RCHE programme outcomes | Track B RCHE coverage | Quarterly via HKCSS reporting |
| Hospital Authority dysphagia discharge data | Denominator refinement | Quarterly (if HA partnership confirmed) |
| HKU Swallowing Research Laboratory community studies | Methodology validation + denominator triangulation | Annual |
Privacy: All AI Assistant aggregate logs are anonymised at source. Individual carer or elderly data never leaves device-level or partner-operator boundaries. Only aggregate counts and demographic proportions are surfaced to the Index calculation, in alignment with HKMA Privacy Commissioner guidance for health-adjacent data. Methodology is versioned: v1.0 launches Q4 2026; v2.0 planned Q4 2027 with Year 1 outcome data integration.
Four milestones establish the Index's credibility and governance infrastructure in Year 1, ahead of programme-scale adoption measurements from Q2 2027.
The Index is designed to outlive the 24-month programme and any single institution's involvement. Governance is open, reproducible methodology is published, and the data licence ensures any researcher or policy body can independently replicate the metric.
All Index data is published under CC-BY-SA 4.0. Quarterly reports include the adoption rate, 95% confidence intervals, Track A/B/C breakdown, setting disaggregation, denominator basis, and a changelog of any methodology changes. Data files (CSV + JSON) accompany each HTML report.