Care Capacity Intelligence: Match Capacity to Demand, Safely
- Maturity
- L4
- Domain
- Plan & Cost
- Analytics
- predictive
- Staffing adequacy rate
- ~89%
- Sponsor
- Chief Nursing Officer
- Confidence
- High
The situation
Shift by shift and unit by unit, will NovaCare have the right credentialed capacity for the demand actually arriving — safely and without avoidable premium spend — and where will it fall short before it happens?
The recommendation on the table
Deploy days-ahead unit-shift demand forecasting with a pre-shift safe-staffing flag
Fewer unsafe shifts and earlier, cheaper intervention than same-day agency calls.
Trade-offRequires fine-grained demand, roster and credential data joined reliably at shift grain.
The evidence
NovaCare meets non-deferrable demand with a workforce rostered shift by shift, and when the plan and reality diverge it closes the gap with premium agency and overtime — simultaneously overspending and running unsafe shifts, learning of both only after the fact. The Healthcare flagship, HC-02 turns capacity from a morning-after audit into a days-ahead, shift-level forecast: matching credentialed capacity to forecast demand, flagging safe-staffing risk before the shift, and surfacing where premium spend was avoidable. It found NovaCare is overspending and unsafe at once — a matching failure — with day-shift over-staffing coexisting with night-shift unsafe gaps, and home-care bottlenecks backing up acute discharges.
Care Capacity Intelligence
Match credentialed capacity to forecast, non-deferrable demand at the unit-shift grain — safely and affordably.
| Day | Evening | Night | Weekend | |
|---|---|---|---|---|
| ED | 1 | 2 | 3 | 3 |
| Med-surg | 1 | 2 | 3 | 2 |
| ICU | 1 | 1 | 2 | 2 |
| Specialty | 1 | 1 | 2 | 1 |
| Ambulatory | 1 | 1 | 1 | 1 |
Key takeawayBreach intensity (3 = frequent sub-floor). Night-shift med-surg falls below the safe floor ~1 shift in 9.
- Avoidable (internal float possible)38%38%
- Genuinely needed62%62%
Key takeaway~38% of premium hours filled gaps internal float could have covered with foresight.
Key takeawayDay-shift over-staffing coexists with night-shift unsafe gaps — a distribution problem, not a volume one.
Key findings
About 38% of premium-labour hours filled gaps that internal float could have covered with foresight, while night-shift med-surg falls below the safe-staffing floor roughly one shift in nine. The problem is matching, not only volume.
What we can’t claim
Day-shift over-staffing coexists with night-shift unsafe gaps: the capacity to staff safely often already exists, just on the wrong shift. The inconvenient truth is that efficiency can only be pursued in the band above the safe floor, never through it — so the savings come from redistribution and foresight, not from cutting to the bone.
Recommendations
Deploy days-ahead unit-shift demand forecasting with a pre-shift safe-staffing flag
high priorityFewer unsafe shifts and earlier, cheaper intervention than same-day agency calls.
Trade-off
Requires fine-grained demand, roster and credential data joined reliably at shift grain.
Build dynamic, scope-valid float deployment
high priorityMaterial reduction in avoidable premium spend while holding the safe-staffing floor.
Trade-off
Needs cross-credential scope mapping and a culture shift toward flexible internal deployment.
Analytical framework
How we reached this
Predictive capacity matching — forecast non-deferrable demand at the unit-shift grain and match credentialed capacity under a hard safe-staffing floor.
ConfidenceMedium-High
Analytical framework
How we reached this
Predictive capacity matching — forecast non-deferrable demand at the unit-shift grain and match credentialed capacity under a hard safe-staffing floor.
Methods applied
Statistical techniques
Algorithms
Data sources
Outputs generated
Why this confidence
Patient demand carries strong, forecastable structure (seasonality, calendar) giving good accuracy on high-volume units; bounded below High by acuity/surge variability and by the assumptions in safe-staffing thresholds.
The reasoning
Business context
The flagship and clearest expression of Healthcare's thesis: capacity as a real-time matching problem against non-deferrable demand, bounded by a hard patient-safety floor. It consumes the HC-01 baseline, sponsored by the CNO because nursing is the binding capacity constraint and the front line of safe staffing. It consumes wellbeing signals from HC-03 but does not own burnout.
Expected value
The largest value pool in the Healthcare portfolio — reduced avoidable premium-labour spend, fewer unsafe shifts and the safety/quality cost they carry, and improved patient flow from relieved bottlenecks. Efficiency is pursued only in the band above the safe-staffing floor.
Workforce landscape
Night-shift med-surg falls below the safe floor roughly one shift in nine; ~38% of premium-labour hours filled gaps internal float could have covered with foresight; elective-surgery demand forecasts to ~93% but ED surge only ~82%; day-shift over-staffing coexists with night-shift unsafe gaps — a distribution problem.
The analytics journey
Level 4, predictive. Forecasts unit-shift demand (required care hours, acuity-weighted) with confidence and matches credentialed capacity under a hard safe-staffing constraint. Honest that surge and acuity add variance, shown as widened bands; the safe floor is never optimised away.
Under the hood
Interpretable time-series forecasting projects acuity-weighted required care hours; queuing-theory models assess capacity; linear programming matches staffing to demand under a safe-staffing constraint; coverage gaps are classified internal-fillable / premium / unfillable. The care pathway is modelled as connected, so bottlenecks are visible across settings.
Confidence & evidence
Why you can rely on this
The inconvenient truth
Day-shift over-staffing coexists with night-shift unsafe gaps: the capacity to staff safely often already exists, just on the wrong shift. The inconvenient truth is that efficiency can only be pursued in the band above the safe floor, never through it — so the savings come from redistribution and foresight, not from cutting to the bone.
Method
Confidence is a deterministic read of KPI strength, target and benchmark coverage across this project — shown on an illustrative reference dataset, computed the same way it would be on live data.
Take this further