# CarePulse IoT — Training plan for a hospital go-live

Delivered by the Caresoft implementation lead with the hospital's nursing educator and biomedical head. Use the DEMO hospital
(`tools/demo_seed.php`) on a training server — never real patients. Role manuals: `docs/manuals/`.

## 1. Timeline (pilot of 1 ward + ICU + 20–50 home patients)
| Week | Activity | Who |
|---|---|---|
| −2 | Kick-off; confirm wards, devices, HIS link, alert limits, escalation trees, consent text; nominate super-users (2 nurses per ward, 1 coordinator, 1 biomed) | Implementation lead, medical director, nursing head, IT |
| −1 | Super-user training (full day) + configuration with them | Super-users |
| −1 | Role sessions (below), hands-on on the training server | All roles |
| 0 | Go-live with floor support every shift for 3 days; daily 15-min huddle on alerts, AI feedback, device issues | Implementation lead + super-users |
| +1 | Review: alert volume, response times, missed readings, AI usefulness; tune limits; refresher for night staff | Medical director, nursing head |
| +4 | Pilot review against success measures (§4); plan roll-out | Management |

## 2. Sessions
| Session | Audience | Length | Content | Assessment |
|---|---|---|---|---|
| S1 Platform overview & safety | Everyone | 30 min | What CarePulse does and does not do; bedside alarms first; downtime procedure; privacy and consent | 5 questions |
| S2 Nursing — ward & ICU | Nurses | 90 min | Central station, NEWS2 list, AI early warning, alerts (acknowledge, act, close, feedback), charting vitals to the HIS, waveforms & strips, therapy board, nurse call board, medicines (Given / Skip), photo readings, device binding at bed | Observed: handle 3 simulated alerts end-to-end; chart one set of vitals |
| S3 Doctors | Doctors | 45 min | Patient page, trends, AI risk reasons and summary, NEWS2, reviews and notes to patients, thresholds per patient, teleconsult with live vitals, CGM report, ECG / stethoscope review, ABDM sharing | Review two demo patients and write a note |
| S4 RPM coordinators | Coordinators | 90 min | Enrolment and consent, programmes, issuing devices and app pairing (QR), patient portal, missed-reading and medicine follow-up, WhatsApp messages, reports to patients, adherence list | Enrol a patient, pair a phone, send a report |
| S5 Biomedical | Biomedical engineers | 90 min | Device registry and labels, gateways and edge health, device models, maintenance (breakdowns, PPM, calibration, uptime), floor maps and tags, cold chain and utilities, data-quality anomalies | Register and bind a device; close a work order; place a gateway on a floor map |
| S6 Hospital admin | Client admin, IT | 60 min | Users and roles, 2-step sign-in, settings tabs (alerts, messaging, HIS, telehealth, operations, AI & medicines, privacy), escalation trees, scheduled reports, audit log, subscription | Create a user, change a limit, read the audit trail |
| S7 Patients & caregivers | Patients | 10 min at enrolment | App pairing, taking a reading, medicines, what happens after an alert, emergency notice | Takes one reading in front of the coordinator |
| S8 Caresoft Level 1 | Caresoft support / DevOps | 90 min | Clients, plans, billing, migrations, system health, backups and restore test, device library, leads, support access rules | Run a backup and restore test; issue a demo invoice |

## 3. Materials
* Role manuals (`docs/manuals/*.md`) → print the one-page quick cards at each nurse station.
* Screenshots: `tests/screenshots/`.
* Demo patients to practise on: Sunil Naik (ward, deteriorating — AI warning), Harish Gupta (ICU, ventilated), Ramesh Kulkarni
  (home BP + pill dispenser, missed doses), Mohammed Ansari (COPD, rescue inhaler overuse), Gopal Reddy (heart failure, weight gain),
  Pooja Shinde (CTG), Lakshmi Iyer (CGM).

## 4. Pilot success measures
| Measure | Target |
|---|---|
| Critical alerts acknowledged within 15 min | ≥ 90 % |
| Ward vitals charted to HIS from devices (no manual transcription) | ≥ 80 % of sets |
| RPM readings adherence | ≥ 75 % of expected readings |
| Nurse call answered within target | ≥ 80 % |
| AI alerts rated useful | track; review if < 50 % |
| Staff trained and assessed | 100 % of rostered staff |
| Unplanned downtime | 0 incidents > 15 min |
