How the nursing productivity index really works
The nursing productivity index formula, a worked pay period on a 30 bed unit, why it disagrees with your staffing, and how to walk finance through the gap.
The short version
- The nursing productivity index is target hours divided by actual productive hours worked. Target hours are your census (patient days) times your budgeted HPPD.
- At 100% you used exactly the hours your census earned. Below 100% you used more. Check which way your report divides, because some flip it.
- The index is only as fair as its inputs. Patients who come and go between midnights, orientation coded as productive, sitters, one to one patients and minimum staffing on low census nights all add hours the census never credits.
- Turn a variance into dollars by multiplying the hours over target by the actual average hourly rate, then explain it line by line before finance does it for you.
- A variance that shows up every pay period for the same reason is not a variance. It is how the unit works, and it belongs in the budget.
The nursing productivity index compares the hours your census earned with the hours your staff actually worked. Target hours, which are patient days times your budgeted hours per patient day, divided by actual productive hours worked, gives you a percentage. If your unit earned 2,688 hours in a pay period and worked 2,940, you ran at 91.4%.
That number lands in your inbox every two weeks, and it rarely matches how the unit felt. This guide shows where it comes from, works through a pay period on a 30 bed unit, explains why it disagrees with your daily staffing, and gives you a way to talk about it with your director and your CFO.
The productivity index formula
Two steps.
This is the version Hunt and Hartman lay out in Nursing Management: target hours are the census multiplied by the budgeted HPPD, divided by the productive hours actually worked, usually measured over the 14 day pay period.
A few words carry the whole thing, so it helps to pin them down.
| Term | What it means on an inpatient unit |
|---|---|
| Unit of service | The thing that measures the work. On most inpatient units it is the midnight census, counted as patient days |
| Budgeted HPPD | The hours per patient day the unit is funded for, set in the annual budget |
| Productive hours | Hours worked in the department doing the assigned work for your role |
| Nonproductive hours | Hours paid but not worked on the unit: vacation, sick, holiday, education, bereavement, FMLA leave |
| Target hours | Patient days × budgeted HPPD. What the census earned |
Your report may call the same idea worked hours per unit of service. On an inpatient unit the unit of service is patient days, so that figure is HPPD on worked hours. If you want the HPPD side in more depth, our guide on how to calculate HPPD covers which hours and which patients belong in it.
One warning before you trust any of it. Which way the division runs is not universal. The version above puts target on top, so 100% means right on target, 91% means you used more hours than earned, and 105% means fewer. Some reports divide actual by target instead, and then a number above 100 means over. Look at your report's footnote or ask finance once, and write it on the report.
A worked pay period on a 30 bed unit
Every number here is invented for the example.
A 30 bed unit runs 12 hour shifts with a budgeted HPPD of 8.0. Over a 14 day pay period the midnight census averaged 24, so the unit had 24 × 14 = 336 patient days.
- Target hours: 336 × 8.0 = 2,688 hours.
- Actual productive hours from the timecards charged to the unit: 2,940 hours.
- Productivity index: 2,688 ÷ 2,940 = 91.4%.
The unit used 2,940 − 2,688 = 252 hours more than the census earned. That is 21 twelve hour shifts.
To turn that into money, multiply the hours over target by the actual average hourly rate. At an average of $50 an hour, also invented, that is 252 × $50 = $12,600 for the pay period. With 26 pay periods in a year, the same gap every pay period would run to 26 × $12,600 = $327,600. That annualized figure is the one that ends up on a slide, so it is worth knowing how it was built.
Now the part the report does not show. Here is where those 252 hours actually went.
| Where the extra hours came from | Hours |
|---|---|
| One patient needing a one to one nurse for 3 days, 24 hours a day | 72 |
| New hires on orientation, coded productive, 8 shifts × 12 hours | 96 |
| A sitter for a patient on safety watch, 4 shifts × 12 hours | 48 |
| Three nights when census dropped but minimum staffing held, 3 shifts × 12 hours | 36 |
| Total | 252 |
None of that is waste. Every hour has a reason a nurse manager would defend. The index cannot see any of it, because the only thing it divides by is the midnight census.
Why the index disagrees with your daily staffing
You staffed to the grid every shift, and the report still says 91%. Here are the usual reasons, from the most common on most units.
Patients the midnight census never counts. Admissions, discharges and transfers take nursing time, and a patient who arrives at 9 in the morning and leaves at 5 that afternoon never appears in a midnight count. Hunt points out that the midnight census does not reflect the activity on the unit through the day. Hunt and Hartman describe one fix, equivalent patient days: add up the minutes short stay patients spend on the unit, and every 24 hours of it earns one patient day. They also warn about hours a tool never captures: in their example, adding care hours that had been left out moved a unit from 10.37 to 11.32 HPPD.
Orientation and education coded as productive. There is no industry standard for whether these hours are productive or nonproductive. If your hospital codes orientation as productive time on your cost center, every new hire cohort pulls your index down while the census stays the same. Find out how your hospital codes it, and ask whether orientation can sit in its own line.
Hours that are fixed, not variable. Some positions do not flex with census. Your manager and educator are usually treated as fixed positions. So is the minimum staffing you hold on a unit no matter how low census drops: two RNs on nights with eight patients still earn less than they cost. The lower the census, the bigger the share of hours that cannot flex, and the lower the index.
Patients who need more than the average. A budgeted HPPD is an average. A one to one patient, a sitter, a run of high acuity admissions all add hours that a census count treats the same as a stable patient going home tomorrow.
Hours charged to the wrong cost center. A nurse who floats out and stays charged to your unit, or a sitter billed to your cost center by default, comes off your index for care your patients did not receive. Check the timecard detail before you explain a variance you did not cause.
The other numbers on the same report
The index usually travels with two more, and they answer different questions.
Productive share of paid hours. Productive hours divided by total paid hours. In Hunt and Hartman's example, 4,370 productive hours out of 5,060 paid is 86%, which leaves 14% nonproductive. If that share climbs, the issue is time off, education or leave, not how you staffed the shifts. Our guide to calculating FTEs shows how that nonproductive share goes into the number of people you need to hire.
Overtime percentage. Overtime hours divided by productive hours. Hunt and Hartman call 2% a reasonable benchmark in healthcare. In their example, 125 overtime hours over 4,370 productive hours is 2.9%. Overtime does not change the productivity index, because an overtime hour counts once like any other. It changes the cost, which is why a unit can hit 100% and still be over budget.
| Measure | Formula | What it tells you |
|---|---|---|
| Productivity index | target hours ÷ productive hours | Did the hours match the census? |
| Productive share | productive hours ÷ paid hours | How much paid time was actually worked? |
| Overtime percentage | overtime hours ÷ productive hours | How much of the work cost time and a half? |
| Hours variance in dollars | (target − actual) × average hourly rate | What the gap cost |
How to talk about it with your CFO
The worst version of this conversation is the one where finance explains your variance to you. Get there first.
Bring the numerator line by line. Not "we were busy." Bring the table from the worked example: 72 hours of one to one care, 96 of orientation, 48 of sitter time, 36 of minimum staffing. A CFO can argue with a feeling. A list of hours with a reason next to each one is much harder to argue with.
Ask for an adjusted index, not a pass. Hunt and Hartman show an adjusted productivity percentage that credits known extra work. In the example, if orientation came out of productive hours, actual hours would drop to 2,940 − 96 = 2,844, and the index would rise to 2,688 ÷ 2,844 = 94.5%. Credit the one to one and the sitter too and it reaches 2,688 ÷ 2,724 = 98.7%. That is a number you can manage to.
Name the pattern. If the same reason shows up every pay period, it is not a variance. Hunt and Hartman put it plainly: "A consistent variance is no longer a variance; rather, it's expected activity." A unit that takes short stay patients every day needs that workload counted in its units of service, or its HPPD rebuilt, at the next budget.
Show what flexing actually saved. Flexing staff to census is how a unit stays productive, and it is also where the floor and finance meet. Keep a simple log of who was flexed, called off or sent to another unit, and when. It shows you are managing the hours you can manage, which earns you a hearing on the ones you cannot.
Agree on the definitions once a year. Which roles count as productive, whether orientation and education are productive, whether the charge nurse is in the budgeted HPPD, and whether the report divides target by actual or the reverse. Write the answers down. Most productivity fights are definition fights, and they get settled once instead of every two weeks.
The hard part of all this is that the index arrives two weeks late, and by then nobody remembers why Tuesday night ran heavy. If your unit uses Adelo Workforce, the manager side includes analytics for productivity, HPPD, hours and overtime, so the gap is visible while you can still do something about it. Your staffing grid is the other half: build it on the same HPPD your index is judged on, and the two will finally agree.
Questions people ask
What is a good productivity percentage for a nursing unit?
On the target over actual version, 100% means you used exactly the hours your census earned. Your organization sets the goal, so ask what yours is and whether it is a single number or a range. Running well above 100 for long stretches is not a win either. It usually means the unit was short.
Is the productivity index the same as HPPD?
They use the same inputs but answer different questions. HPPD tells you how many hours each patient day got. The productivity index compares that with what the budget allowed. A unit budgeted at 8.0 HPPD that ran at 8.75 has a productivity index of 8.0 ÷ 8.75 = 91.4%.
Do orientation hours count against productivity?
It depends on how your hospital codes them. There is no industry standard for whether orientation and education are productive or nonproductive. If they are coded productive on your cost center, they lower your index, so ask finance whether they can be tracked on their own line.
What is a unit of service in nursing?
It is the thing that measures the work. On most inpatient units it is the midnight census, counted as patient days. Other departments use visits, OR minutes, treatments or tests.
Why does my productivity drop when census is low?
Because some of your hours do not flex. Fixed positions and minimum staffing stay the same at a census of 15 as at 25, so they make up a larger share of the hours when census drops, and the index falls even if every shift was staffed to the grid.
Sources
- Hunt PS, Hartman D. Meeting the measurements of inpatient staffing productivity, Nursing Management, June 2018, the productivity index formula, productive and paid hours, overtime percentage and the 2% benchmark, equivalent patient days, adjusted productivity and dollar variance.
- Hunt PS. Developing a staffing plan to meet inpatient unit needs, Nursing Management, May 2018, units of service, the midnight census and its limits, productive and nonproductive hours, fixed positions and the lack of a coding standard.
- Fike GC, et al. Hours per patient day: Understanding this key measure of productivity, American Nurse Journal, 2016, why HPPD has no single standard calculation and does not adjust for discharges during the day.
