Key takeaways
- At a health system in Illinois, an open shift posted seven or more days ahead was filled by the hospital’s own staff 38% of the time. Posted inside 24 hours, about 11%. Same units, same staff, and the late shifts carried the higher incentive.
- The data doesn’t tell us why each urgent shift was posted with so little notice. Some coverage needs, though, like approved time off or a standing vacancy, are known well in advance. When those shifts aren’t posted until the final 24 hours, the team loses time it could have used to find coverage.
- Posting a gap the day it is known costs nothing, and lead time is the single largest differentiator in the data. The question for each unit is: how many of those urgent shifts could have been posted sooner?
There is a belief inside most hospitals that nobody wants the last-minute shift. It posts, the phone tree runs, nothing comes back, and one of two things happens: the conclusion is that the staff were not interested, or the incentive goes up at the last minute, which teaches everyone to wait. The 90-day data says something different. When an urgent shift actually reached the people who could take it, the answer came fast.
The lever is time
Group every shift the hospital posted for its own staff by how far ahead it went up, and the fill rate climbs with every step of lead time. Seven or more days ahead, 38.1% filled. Three to seven days, 30.3%. One to three days, 29.5%. Inside 24 hours, about 11%. Inside four hours, under 2%.
Same units, same staff. The incentives were not identical, and the difference ran the wrong way: the week-out shifts carried the lowest average incentive in the dataset, $9.67 an hour, and the shifts inside four hours the highest, $15.00. Even so, a shift posted a week out was nearly three and a half times more likely to fill internally than one posted the day before. Time often did the work that money could not.
That is the whole finding. The rest of this piece is about what to do with it.
Which urgent shifts could have been posted earlier?
The data can’t tell us. We know 349 shifts were posted less than 24 hours before they started, but we don’t know what caused each opening. Some may have resulted from a 0540 a.m. call-out for a 0700 a.m. shift or an overnight census spike. Posting those shifts earlier wouldn’t have been possible.
But not every open shift starts as a surprise. Approved time off is known the day it is approved. A standing vacancy produces the same open shifts every schedule period. Known procedure volumes and seasonal census are in the forecast. Even unplanned absence has a rate a unit can plan against: the Bureau of Labor Statistics puts absence among healthcare practitioners at 3.7% of full-time workers in 2025, against 3.2% across all occupations.
Whether any of those reached the urgent window at this system is a question for the next dataset. Whether they reach it at yours is a question you can answer this month, and if the answer is more than zero, those are shifts that could have been posted at 38% odds instead of 11%.
What changes when you post early
Three things follow for any gap that is known ahead of time.
The shift fills. A week-out posting at this system filled at 38%. That is the top of the curve, and it was reached with a $9.67 an hour incentive, the lowest of any group. A gap that would have gone to overtime or agency at the last minute instead went to an employee who had time to plan for it.
The incentive comes down. The premium on urgent shifts ran to $15 an hour yet had the lowest fill rate. Posting early moves the fill rate into the part of the curve where a modest premium works, so the same coverage costs less per hour and, since it is covered internally, avoids the overtime or outside rate entirely. The incentive data covers this in detail.
Post the gaps you know about before they become urgent. Giving those shifts more time to find coverage leaves the last-minute response for needs that really do arise at the last minute, like call-outs and census spikes. That’s the separate challenge we explore in Hospitals Have Two Open Shift Problems. They Manage Them Like One. It also means the charge nurse spends less time working a phone list for a shift that could have been on the board a week ago.
How to run it
This is a posting policy, not a purchase, and it can start on a single unit next schedule period.
Post every known gap the day it becomes known. Approved time off goes up as an open shift the day it is approved, not the day the schedule is built. Standing vacancies are posted for the full schedule period at once. Then track three numbers weekly by lead-time group: how many shifts were posted, how many filled, and what the incentive paid. Thirty days is enough to see the curve on your own unit.
Track one more thing while you do it: for each shift that reaches the 24-hour window, note when the gap was first known. That is the number the Illinois data could not give you. If a share of your urgent shifts were known days earlier, the policy has a target. If none were, your urgent problem is entirely a reach problem, and that is useful to know too. Either way, the measure of success is how many shifts reach the 24-hour window at all, because that is where the hospital pays the most and fills the least.
Where we come in
Gigly is how the Illinois system put open shifts in front of its own staff, at any lead time. Employees see open shifts on their phones and pick them up, before those shifts turn into overtime or agency. Posting early is the part that costs nothing; reaching everyone at once is what turns an early posting into a filled one. By 90 days, staff had picked up 361 shifts and about 3,700 hours of care that did not go outside.
If you want to see your own curve, pull the last 90 days of open shifts and sort them by how far ahead each was posted. Then, for the ones inside 24 hours, ask when each gap was first known. We are glad to walk through it with you.
Frequently Asked Questions
How far ahead should an open shift be posted
As early as the gap is known. In this data, shifts posted seven or more days ahead filled at 38%, against about 11% inside 24 hours. Every day of lead time moves a shift up that curve.
Does posting early cost anything?
No. It is a change to when gaps are posted, not what they pay. In this data the week-out shifts also carried the lowest incentive, $9.67 an hour, so earlier posting tends to reduce the premium rather than add to it.
What about the shifts that really are last-minute?
They still happen, and posting early does not solve them. They need a different fix: reaching every qualified employee at once, fast. Posting early handles the gaps you knew about; reach handles the ones you did not. How many of each you have is something only your own data can tell you.