Intervene

Ranked by bed-day risk · next 48h
STALE
Likely to slip · 48h
7
miss expected discharge
Open actions · in window
10
to prevent delay
Predicted excess · total
15.0d
bound if all hit expected
Occupancy now
9%
9 / 96 beds
Model abstained
0
needs human review
Intervention boardwork top-down
#PatientDayTime to slipPredicted barrierNext best actionOwner · dueAt riskConf
1
NY0398
6 Tower
D10 / 3 over 7d Review readiness
unassignedescalate
2.0d Low
2
NY0388
Telemetry
D7 / 6 over 1d CLClinical readiness Order PT/OT eval
Rehab Coordinatordue day 1
2.0d Med
3
NY0392
6 Tower
D7 / 6 over 1d AUAuthorization Open prior-auth
Utilization-Review Nursedue day 2
2.0d Low
4
NY0385
5 East
D4 / 4 today CLClinical readiness Order PT/OT eval
Rehab Coordinatordue day 1
2.0d Low
5
NY0396
4 West
D7 / 7 today CLClinical readiness Order PT/OT eval
Rehab Coordinatordue day 1
2.0d Low
6
NY0399
ICU
D1 / 3 in 2d AUAuthorization Open prior-auth
Utilization-Review Nursedue day 2
2.0d Low
7
NY0377
Ortho
D8 / 9 in 1d AUAuthorization Open prior-auth
Utilization-Review Nursedue day 2
1.0d Med
8
NY0391
5 East
D6 / 9 in 3d AUAuthorization Open prior-auth
Utilization-Review Nursedue day 2
1.0d Med
9
NY0397
5 East
D1 / 4 in 3d Review readiness
unassignedescalate
1.0d High
Admin-baseline model (AUROC ≈0.63) · predicted excess is a definitional bound, not causal bed-days saved · abstentions routed to review · calibration gap disclosed · fairness gate →