dashboard-01
Clinical caseload dashboard
Four tiles that each carry a denominator and a window, a response panel that compares every patient against their own predicted curve, and a caseload ordered by who needs attention rather than alphabetically.
Treatment response
PHQ-9 against the band predicted by intake severity · 68 patients
Risk queue
Ordered by time remaining, not by severity label
A red count with no clock is wallpaper. Each row carries how much of its window has already gone, because 19h left means one thing on a 24-hour window and another on a 60-day one.
Engagement funnel
Referral to session three · last 90 days
The wall is between session one and three — where behavioral health loses people, and where a clinic-wide no-show rate hides it.
Caseload
Ordered by attention needed, not by name
| Patient | Instrument | Trend | Sessions | Status |
|---|---|---|---|---|
| R. Okonkwo | PHQ-9 16 | flat | 8 | not on track |
| T. Almeida | GAD-7 15 | ↑ 2 | 5 | not on track |
| J. Whitfield | PHQ-9 11 | ↓ 3 | 12 | slow response |
| M. Delacroix | PHQ-9 6 | ↓ 9 | 9 | responding |
| S. Ferreira | GAD-7 4 | ↓ 11 | 14 | remission |
| A. Nakamura | — | — | 1 | awaiting baseline |
What to look at: The tile grid a SaaS dashboard uses, with clinical numbers in it. Every figure renders the quantity it is part of, so nothing reads as a bare count — and there is no count-up animation, because for the second it runs the dashboard would be showing numbers that are wrong.