How to calculate HPPD for your unit

Here is how to calculate HPPD: add up the nursing hours worked on your unit, then divide by the patient days for the same period. If your unit worked 192 nursing hours in a day with a census of 24, that day ran at 8.0 hours per patient day. The math is that simple. The hard part is deciding which hours and which patients belong in it, because that choice can move your number by more than an hour per patient day.

This guide walks through the formula, a full month on a 30 bed unit, what the national definition counts and leaves out, and the mistakes that make HPPD lie.

The HPPD formula

Both halves need a little care.

Nursing hours worked are hours actually worked, not scheduled and not budgeted. The CDC's National Healthcare Safety Network uses this definition in its nurse staffing protocol, which builds on the measure endorsed by the National Quality Forum (NQF #0205). It calls them direct patient care hours, sometimes called productive hours, and it counts overtime. Vacation, sick, orientation, education and committee time are nonproductive and stay out.

Patient days are the count of patients on your unit each day, added up over the period. The CDC protocol says the count should happen at the same time every day, and that midnight is preferred. A unit with a midnight census of 24 every night for a 30 day month has 720 patient days.

The same structure works for one role at a time. RN HPPD puts only RN hours on top. The national measure reports total nursing staff, RN, LPN or LVN, and assistive personnel separately, all over the same patient days. The share of hours each group works is your skill mix.

Which hours count, and which do not

This is where most of the arguments start. The CDC protocol uses a simple test: a person's hours count if they have direct patient care responsibilities for more than 50% of their shift, are part of the unit's staffing plan, and have hours charged to the unit. Illinois, in its own state staffing data guidance, uses the same 50% line and adds that someone who passes it reports all their shift hours, not just the direct care part.

Your hospital's internal productivity report may not follow these rules. Many finance departments count every worked hour charged to the cost center, because that is what the budget pays for. Neither version is wrong. They answer different questions: the national measure asks how much care the patients got, and the cost center version asks how many hours the unit spent. The trouble starts when your budget is set one way and your report is run the other.

A worked month on a 30 bed unit

Take a 30 bed unit running 12 hour shifts, with a midnight census of 24 every night for a 30 day month. All of these numbers are invented for the example.

Every day the unit staffs 6 RNs with assignments on days and 6 on nights, 2 techs on days and 1 on nights, and a charge nurse on each shift with no assignment. During the month it also had two new hires on orientation, a sitter for one patient on safety watch, and a couple of education days.

Patient days: 24 × 30 = 720.

Now run it three ways.

1. The national definition. Only the RNs and techs with assignments: 4,320 + 1,080 = 5,400 hours. 5,400 ÷ 720 = 7.5 HPPD.
2. Every worked hour charged to the unit. Add the charge nurses, orientees and sitters: 5,400 + 720 + 288 + 360 = 6,768 hours. 6,768 ÷ 720 = 9.4 HPPD.
3. Paid hours. Add the education time too: 6,768 + 96 = 6,864 hours. 6,864 ÷ 720 = 9.53 HPPD.

Say the unit is budgeted at 8.0 HPPD. On the first number you are under budget and might be told you are short. On the second you are 1.4 HPPD over, which across 720 patient days is about 1,008 hours, or 84 twelve hour shifts. Same unit, same month, same staff. That gap is why the first question in any productivity meeting should be: which hours are in this number?

Using HPPD to plan a shift

Turn the formula around and it tells you how many hours a census earns:

At 8.0 HPPD and a census of 24, the unit earns 192 hours a day, which is 16 twelve hour shifts across days and nights. If census drops to 21, it earns 168 hours, or 14 shifts. That is the arithmetic behind a staffing grid, and it is why a call to flex someone home often comes the morning after a run of discharges.

Two cautions from the floor. First, the target only works if it was built with the same definition your report uses. A grid built on cost center hours that includes the charge nurse will look generous if you are measured on direct care hours alone, and tight the other way around. Second, a midnight census says nothing about the six admissions and five discharges that happened in between. A unit with heavy turnover does far more work per patient day than its census shows, and HPPD alone will not tell you that.

The four mistakes that make HPPD lie

1. Averaging daily HPPD. Say Saturday had a census of 12 and 120 hours (10.0 HPPD) and Sunday had a census of 24 and 192 hours (8.0). The average of the two is 9.0. The real figure is 312 hours ÷ 36 patient days = 8.67. Averaging gives a quiet day the same weight as a full one. Always add up hours and patient days for the whole period, then divide once.

2. Missing short stay patients. Observation, same day surgery and other short stay patients who arrive and leave between midnights never show up in a midnight census, but your staff cared for them. The Illinois guidance calls the midnight census alone "not accurate, and thus not recommended" for units that mix inpatients and short stay patients. It describes better options: add the short stay hours divided by 24 to the midnight total, count the census several times a day, or total every patient's actual hours and divide by 24. In the example above, if short stay patients added 360 patient hours in the month, that is 15 more patient days, and the cost center HPPD falls from 9.4 to 6,768 ÷ 735 = 9.21.

A national study backs this up. Simon and colleagues compared routine patient day reports from 260 units against census counts taken every two hours, and found that two of the five accepted counting methods underestimated patient days, with the effect reaching 7.6% depending on the share of short stay patients.

3. Mixing definitions between budget and actual. If the budget was set on worked hours and the monthly report divides paid hours, the unit will look over budget every month even when nothing changed. Ask finance to show you the numerator line by line once a year.

4. Charging hours to the wrong unit. A nurse who floats to another unit should count there, not at home. A sitter or orientee charged to your cost center by default can push you over without adding a single hour of care your patients received. Check where the hours landed before you explain a variance you did not cause.

What HPPD does not tell you

HPPD is the most common nurse staffing measure in research. A review by Min and Scott looked at 17 studies, found it was the most frequently used, and reported strong evidence that it predicts falls but weak evidence for pressure ulcers. The same review says it needs adjusting for differences between nurses, patients and units, and a consistent way of collecting data.

In plain terms: HPPD is a fair way to compare your unit to itself over time, using one definition. It is a poor way to compare an ICU to a step down unit, or your unit to another hospital, unless you know both were counted the same way. It ignores acuity, turnover and experience mix. Keep it next to those, not in place of them. And remember that hours are not all the same price: overtime hours count the same in HPPD as straight time but cost half again as much. Our guide to the 8/80 overtime rule covers how the hours you schedule turn into overtime.

Once a unit has its schedule and timecards in one place, HPPD can be watched daily instead of read in a report weeks later. If your unit runs on Adelo Workforce, the manager side includes analytics for productivity, HPPD, hours and overtime.

Questions people ask

What is a good HPPD for an ICU?

There is no single right number. It depends on the patients, the unit's turnover and above all on which hours were counted. A useful target is your own budgeted HPPD, built with the same definition as your report, compared month over month.

Does HPPD include the charge nurse?

In the national measure, only if the charge nurse has direct patient care responsibilities for more than half the shift, is part of the staffing plan and is charged to the unit. Many hospital finance reports count every worked hour in the cost center, which includes the charge nurse either way. Ask which version your report uses.

Is HPPD calculated on worked hours or paid hours?

The national measure uses hours actually worked, including overtime, and leaves out vacation, sick, education and committee time. Some internal reports use paid hours instead, which makes HPPD higher. Make sure your budget and your actual are on the same basis.

How does HPPD relate to nurse to patient ratios?

A ratio held for 24 hours converts directly. One RN for every 4 patients on every shift is 24 RN hours ÷ 4 patients = 6.0 RN HPPD. One to 2 is 12.0. Add techs and charge nurses on top, depending on your definition.

How do observation patients affect HPPD?

If they come and go between midnights, a midnight census misses them, so your HPPD looks higher than the care really was. Adding their hours divided by 24 to your patient days, or counting census several times a day, fixes most of it.