Care team demo
Privacy & safety
Consent, privacy, responsible AI, and the audit log. Fictional data only. Not for clinical use.
Responsible AI
What the rules use
- The screening result and its interpretation (normal, borderline, abnormal, or critical).
- Whether an abnormal result is a cancer screen (cervical screening, mammogram, or FIT).
- How many abnormal risk findings are still open.
- Age 60 or older, which moves priority up one level.
- Days waiting past the current priority window, which also moves priority up one level.
- The Loop Health Score uses whether verified care happened within 120 days of the result.
What the rules do not use
There is no nationality or gender in the score. Insurance plan is not used either. Priority and the Loop Health Score do not use nationality, gender, or insurance.
The clinician always decides
AI suggestion – clinician decides. Nothing moves forward until a Family Physician accepts, changes, or rejects the suggestion. This app does not diagnose or prescribe.
Override rate
No clinician decisions yet.
Priority 1 among open cases: 23 of 25 (92.0%).
How to report a wrong suggestion
Use Report a safety concern on this page. The report is written to the audit log and cannot be edited or deleted. A Family Physician can also change or reject the suggestion on the record, and must give a reason.
Bias check
Priority level and AI override rate by group. A note appears when a group differs from the average by 10 or more points. This table is for review. It does not change a person’s priority.
Gender
| Group | Open cases | Priority 1 | Priority 1 % | Decisions | Overrides | Override % |
|---|---|---|---|---|---|---|
| Female | 11 | 10 | 90.9% | 0 | 0 | — |
| Male | 14 | 13 | 92.9% | 0 | 0 | — |
Age group
| Group | Open cases | Priority 1 | Priority 1 % | Decisions | Overrides | Override % |
|---|---|---|---|---|---|---|
| 18–39 | 7 | 5 | 71.4% | 0 | 0 | — This group differs a lot from the average (10 or more points). |
| 40–59 | 13 | 13 | 100.0% | 0 | 0 | — |
| 60–75 | 5 | 5 | 100.0% | 0 | 0 | — |
Nationality group
| Group | Open cases | Priority 1 | Priority 1 % | Decisions | Overrides | Override % |
|---|---|---|---|---|---|---|
| Emirati | 8 | 8 | 100.0% | 0 | 0 | — |
| Arab expat | 4 | 4 | 100.0% | 0 | 0 | — |
| South Asian | 5 | 4 | 80.0% | 0 | 0 | — This group differs a lot from the average (10 or more points). |
| African | 3 | 3 | 100.0% | 0 | 0 | — |
| Western | 4 | 3 | 75.0% | 0 | 0 | — This group differs a lot from the average (10 or more points). |
| Other | 1 | 1 | 100.0% | 0 | 0 | — |
Insurance plan
| Group | Open cases | Priority 1 | Priority 1 % | Decisions | Overrides | Override % |
|---|---|---|---|---|---|---|
| Thiqa-style (fictional) | 8 | 8 | 100.0% | 0 | 0 | — |
| Basic (fictional) | 6 | 4 | 66.7% | 0 | 0 | — This group differs a lot from the average (10 or more points). |
| Enhanced (fictional) | 11 | 11 | 100.0% | 0 | 0 | — |
Privacy and data governance
- Fictional data only. These records are not for clinical use and are not real people.
- Data minimisation: the care record keeps the fields needed to close the screening-to-care loop. Programme Manager views use the person ID, not the name.
- Retention plan: demo records stay in this browser until reset. A live service would keep the care record for the period agreed with the care team, then remove or de-identify it. Data is saved only in this browser. Nothing is sent to a server.
- Designed to align with UAE/Abu Dhabi health data rules (to be confirmed with DoH).
- Planned encryption and hosting in-country for a future pilot. This demo does not host clinical data.
Report a safety concern
Use this when a suggestion looks wrong or a safety issue needs review. The entry is appended to the audit log.
Audit log
Read-only. Entries cannot be edited or deleted.
0 of 0 entries. The file is UTF-8 and includes the rows shown.
No audit entries yet.