Proof register
Every public number traces to a row here. Only Validated rows ship without qualification.Validated21publishable
Modelled1publish with assumptions stated
Observed0do not ship
To confirm4never publishable
| ID | Metric | Value | Status | Definition / notes | Checked |
|---|---|---|---|---|---|
| recall-success-rate | Recall success rate | 75.6% | Validated | Distinct patients who booked divided by distinct patients reached. Pack implies 76.3% (194,566/255,009); use register value until re-run. | 2026-08 |
| bookings-uplift | Bookings uplift | 3.8x | To confirm | Pending Penrose baseline denominator. Do not publish. | 2026-08 |
| practices-live | Practices live on platform | ~120 | Validated | Live today per pack. | 2026-08-13 |
| practices-contracted | Practices contracted | 280 | Validated | Contracted incl. onboarding, commercial pipeline toward 300. | 2026-08-13 |
| practices-served | "250+ practices" (public phrasing) | 250+ | To confirm | CONFLICT: only defensible as contracted count. Tom to pick approved public phrasing (live vs contracted) and check live site copy. | 2026-08-30 |
| patients-live | Patients on live practices' lists | 1.3m | Validated | Registered lists of live practices, measured. | 2026-08-13 |
| patient-population | "2.5m patients" (public phrasing) | 2.5m | To confirm | CONFLICT: 2.5m is contracted pipeline total; 1.3m live. Same decision as practices-served. | 2026-08-30 |
| recalls-sent | Recalls sent, all-time | 912,992 | Validated | Individual automated messages successfully delivered (SMS, email, WhatsApp, NHS App), first go-live to 2026-08-13. | 2026-08-13 |
| patients-reached | Patients reached, all-time | 255,009 | Validated | Unique patients, deduplicated by NHS number. | 2026-08-13 |
| bookings-made | Bookings made, all-time | 194,566 | Validated | Booked via recall link or at practice within 2 weeks of delivered recall. Window attribution, central estimate: state as "attributed to recall". | 2026-08-13 |
| phone-bookings | Staff call-list bookings | 40k | Validated | Booked by phone from platform-generated call lists, additional to automated recall bookings. | 2026-08-13 |
| indicators-computed | Clinical indicators computed | 237 | Validated | Computed nightly per eligible patient: QOF, screening, imms, medicines safety, custom. | 2026-08-13 |
| data-points-daily | Care-record data points processed daily | 1bn+ | Validated | Nightly ingest and evaluation across live patient population. | 2026-08-13 |
| towns-cities | Towns and cities covered | 41 | Validated | England, all-time, practices on platform. | 2026-08-13 |
| booking-rate-nhsapp | Booking rate, NHS App recalls | 44.1% | Validated | Unique patients booked / recalled via channel, 2-week attribution, all-time. 66,035 patients recalled. | 2026-08-13 |
| booking-rate-email | Booking rate, email recalls | 35.8% | Validated | Same definition. 96,177 recalled. | 2026-08-13 |
| booking-rate-text | Booking rate, SMS recalls | 30.8% | Validated | Same definition. 76,996 recalled. | 2026-08-13 |
| booking-rate-whatsapp | Booking rate, WhatsApp recalls | 26.1% | Validated | Same definition. 13,872 recalled. | 2026-08-13 |
| traditional-recall-benchmark | Letter/phone recall response, context | ~20% | Modelled | External context figure from pack; cite as "typically around 20%" with source needed before public use. | 2026-08-13 |
| qof-uplift | QOF achievement uplift with Hippo | +10.3 pts | Validated | Definition in pack; confirm denominator basis before public use. | 2026-08-13 |
| qof-income-supported | Total QOF income supported | £29m | Validated | MAX annual QOF income available to practices on platform (564 pts x £225.49 scaled by list size). NOT Hippo revenue, NOT attributable uplift: phrasing must say "supported". | 2026-08-13 |
| example-practice-qof | Example practice QOF completion | 98.6% | Validated | 556/564 tracked points, 2025/26, one practice, raw indicator completion. Anonymised unless consent. | 2026-08-13 |
| care-targets-met | Clinical care targets met, 25/26 | 1m+ | Validated | Completed indicators across platform population. NOT all Hippo-attributable: phrasing must reflect. | 2026-08-13 |
| languages | Languages supported | 70+ | Validated | Automatic translation to patient's coded preferred language. | 2026-08-13 |
| response-equity-range | Response rate by ethnic group | 37-50% | Validated | High-level ONS categories vs 44% average; groups under 100 suppressed. Anonymisation review before any equity content. | 2026-08-13 |
| nhs-app-recall-conversion | NHS App recall conversion | 53% | To confirm | ALREADY PUBLISHED in case study slug. Pack shows channel rate 44.1% all-time; the 53% may be campaign-specific. Reconcile or amend page. | 2026-08 |
Source: context/proof/metrics-register.md. To add or change a row, edit the file; a draft cites rows by ID under proof_refs.