Region forecast
Harbor Point
49/ 100 · two-week forward risk49%High(47–51)
EMS naloxone uses
22
naloxone uses
ER overdose visits
20
ER visits
test-strip positivity
19
% positivity
recent losses
3
recent losses
What's driving this forecast
Each indicator's weighted contribution. The heaviest names the driver: ER overdose visits.
- ★ ER overdose visitsweight 1.1 · 32% of forecast
- EMS naloxone usesweight 0.9 · 29% of forecast
- test-strip positivityweight 0.8 · 22% of forecast
- recent lossesweight 4 · 17% of forecast
raw = 22×0.9 + 20×1.1 + 19×0.8 + 3×4 → score = min(100, round(raw / 1.4)) = 49
forecast-run.chain
$ no anchored forecast run yetanchor a snapshot from the board to create one
What-if simulator
Adjust aggregate indicators to preview the forecast. Nothing is saved.
49/ 100 forward risk49%High(47–51)
- ★ ER overdose visitsweight 1.1 · 32% of forecast
- EMS naloxone usesweight 0.9 · 29% of forecast
- test-strip positivityweight 0.8 · 22% of forecast
- recent lossesweight 4 · 17% of forecast
Open methodology — every weight in plain sight
A transparent weighted model turns aggregate leading indicators into a 0–100 two-week forward risk score. The heaviest-weighted input names the driver. No person-level data exists anywhere in this pipeline.
| Indicator | Weight | Reference ceiling |
|---|---|---|
| EMS naloxone uses | 0.9 | 60 |
| ER overdose visits | 1.1 | 40 |
| test-strip positivity | 0.8 | 50 |
| recent losses | 4 | 10 |
- score = min(100, round(Σ(value × weight) / 1.4))
- surge flag raised at score ≥ 70
- confidence = 0.6 × data completeness + 0.4 × indicator consistency; sparse or contradictory inputs widen the band (up to ±25) so we never imply false precision