Absence Is the Metric Health Systems Track and Never Plan With
Every health system records it. Almost none plans with it.
6 min read
By Rob Howie, Founder & CEO · Senzo · September 2026
Quick Answer
Unplanned absence is recorded by every health system, usually for payroll, and almost never used as a planning input. It is the only one of the four workforce streams that reflects how the workforce is doing rather than how it is sized, which is why it moves first: 60 to 90 days before overtime concentrates and exits begin. Read at unit and provider-type level against the unit's own baseline, it is the earliest warning a planning function has. Used as a unit-level productive-hours deduction, it is the difference between a supply model that holds and one that overstates capacity.
Key Takeaways
- Absence is the only stream staff control, so it is the first to move when a unit is under stress
- The signal is the trend against the unit's own baseline, not a system-wide threshold
- In a supply model, absence belongs at unit and provider-type level; one system-wide factor is the most common way to overstate capacity
Ask a health system for its unplanned absence rate and you will get a number, usually from payroll, usually system-wide, usually for last quarter. Ask what the planning function does with it and the answer is almost always nothing. It is tracked for pay and for attendance management. It is not treated as information about the future.
That is a mistake with a specific cost, because of what absence is.
What counts as absence?
Unplanned hours away from scheduled work: sick time, unplanned leave, no-shows. Planned vacation and approved leave are scheduled in advance and belong to supply, not to absence. The line matters. Vacancy, overtime and staff flows describe how a unit is sized and staffed. Unplanned absence describes how the people in it are doing. It is the only one of the four streams that the workforce itself controls.
Why does it move first?
A nurse who is exhausted, disengaged or job-searching cannot create a vacancy and cannot schedule an overtime shift. She can call in. Absence is the channel available to staff before any other signal registers, which is why a sustained rise in one unit precedes the rest of the sequence: overtime concentrates to cover the absent hours, the people covering it start to leave, and vacancy, the number everyone reports, moves last.
In practice the lead is 60 to 90 days. That is the window in which a planning function can do something other than open a posting.
How do you read it?
Against the unit's own baseline, not a system threshold. There is no absence rate that is a problem everywhere. A unit that has run 5% for two years and rises to 7% for ten weeks is under stress. A unit that has always run 8% is a unit that has always run 8%. The system-wide average, which is what most reports carry, hides both.
By provider type within the unit. Absence concentrates. One provider type in one unit can be in trouble while the unit's total looks unremarkable, because the others are not.
Alongside overtime, not instead of it. Rising absence with flat overtime is a unit absorbing the gap by leaving shifts short. Rising absence with rising overtime is a unit covering the gap with the people who are still there, which is the pattern that produces exits.
What does it change about supply?
The other half of the mistake is in the supply model. Most health workforce plans convert paid hours to bedside hours with a single productive-hours factor, and absence is buried inside it. The factor runs roughly 78 to 85% in Canadian acute care, and it differs by unit more than any other component of supply, almost entirely because absence differs by unit.
A single system-wide factor therefore overstates supply exactly where it is weakest. The unit with 9% absence is planned as if it had 5%, and the gap shows up as chronic overtime that nobody modelled. The correction is simple to state: absence belongs in the supply model at unit and provider-type level, from the unit's own history, refreshed as the data arrives.
Where does the data come from?
The scheduling or attendance system, usually summarised for payroll at the pay period. That summary is enough. It does not need to be daily and it does not need a new system; it needs to be joined to vacancy, overtime and flows for the same unit and the same period, which is the join most health systems have never built. The definitions Senzo uses are in the data dictionary.
What Senzo does with this
Pulse reads absence at unit and provider-type level against each unit's own baseline, alongside the other three streams, and surfaces the sequence above in plain language as it forms. Forecast, in build, carries the unit-level absence into the supply model so the plan reflects the capacity a unit actually has. Both are being built with three design partners; the programme has the details.
Frequently asked questions
What counts as absence in workforce planning?
Unplanned hours away from scheduled work: sick time, unplanned leave, no-shows. Planned vacation and approved leave are scheduled and belong to supply, not absence. The distinction matters because unplanned absence is the only one of the four streams that reflects how the workforce is doing rather than how it is sized.
Why does absence move before vacancy and overtime?
Because it is the channel staff control. A nurse who is exhausted or disengaged cannot create a vacancy or an overtime shift; she can call in. Sustained unplanned absence in one unit is the earliest measurable signal of workforce stress, typically 60 to 90 days before overtime concentrates and exits begin.
How should absence be used in a supply model?
As a productive-hours deduction at unit and provider-type level, not as a system-wide factor. Paid hours minus vacation, education, and unplanned absence gives bedside hours; the absence component varies by unit far more than the others, and a single system-wide rate is the most common way to overstate supply.
What absence rate is a problem?
There is no universal threshold; the signal is the trend against the unit's own baseline. A unit that has run 5% for two years and rises to 7% for ten weeks is a stronger signal than a unit that has always run 8%. Senzo compares each unit and provider type to its own history, not to a system average.
Where does absence data come from?
The scheduling or attendance system, at whatever period it already reports. Most health systems hold it at the shift level and summarise it for payroll; the summary is enough to plan with if it is joined to vacancy, overtime and flows for the same unit and period.