The Ward Clock Says 30 Hours, the Discharge Summary Says Two Days
Admission and discharge are recorded to the minute. Almost nothing that happens afterwards is. Bed occupancy, the discharge letter, the departmental dashboard and the national statistics return all speak in days, so a stay that a nurse experienced as an afternoon, a night and most of the following morning has to be turned into a figure a coder, a bed manager and an auditor will all recognise.
The division is easy; the disagreement is about what a day means. Elapsed hours divided by 24 gives the true duration of the episode. The counts used for occupancy and for many payment rules are not durations at all — they count midnights crossed. The two answers can differ by a full day for the same patient, and knowing which one a form is asking for matters more than the arithmetic.
Why One Stay Produces Two Different Numbers
Occupancy is counted at midnight
Clinical protocols are written in hours
The 24-hour line changes the record
Statistics are published in days
Reading an Admission Record Into the Number a Form Wants
Start from the two timestamps in the notes, work out the elapsed hours, and only then decide which kind of day the destination field is asking for.
Enter the elapsed hours from the episode
Type the difference between admission and discharge, such as 30 or 7.5. A comma is accepted for the decimal point, so 7,5 works too, and spaces typed inside long numbers are ignored.
Read the decimal day and check it against the midnights
A result of 1.25 d is one and a quarter days of care. Compare it with the calendar: if only one midnight fell inside the episode, the occupancy return will still show one day. Report the elapsed figure for clinical audit and the counted figure for the census, and label which is which.
Reverse it when a pathway is specified in days
Press the swap button (↔) for d → h to turn an expected stay written as "2–3 days" into the 48–72 hour window a discharge planner, a transport booking or a bed forecast actually works to.
Copy the plain figure into the audit sheet
The copy control on each field gives you the number alone, with no unit and no spacing, ready for a spreadsheet column that will later be averaged across a cohort. Ctrl + C inside the field behaves the same way.
How Long Each Pathway Runs Once You Divide by 24
Typical episode lengths across common care pathways, shown as the hours a ward record holds and the days a summary or a statistics return will carry. Values are illustrative of the pathway rather than of any one service.
| Care pathway | Elapsed hours | Elapsed days | How the stay is usually counted |
|---|---|---|---|
| Ambulatory day case | 6 h | 0.25 d | In and out on one calendar date; no midnight, no bed-day |
| Emergency short-stay assessment | 12 h | 0.5 d | Recorded as an attendance rather than an admission |
| Extended observation | 23 h | ≈0.958 d | Deliberately kept under the 24-hour line |
| Overnight post-operative stay | 30 h | 1.25 d | One midnight crossed, so one day on the census |
| Stay expected to span two midnights | 36 h | 1.5 d | The point at which inpatient status is normally assumed |
| Elective joint replacement | 72 h | 3 d | Quoted to the patient in days from the outset |
| Critical care episode | 120 h | 5 d | Counted separately as critical-care bed-days |
| Inpatient rehabilitation block | 336 h | 14 d | Reviewed weekly; hours stop being the working unit |
Notice how the fourth column stops mattering as you go down the table. Below about two days the exact hour drives the record, which is why the 23-hour row exists at all: an episode is shaped around a threshold rather than around when the patient happens to be ready. Above a week nobody counts hours, and the same division that felt fussy at the top is simply how a fortnight gets written down.
What This Pair Is Good For on the Ward Round
Answer a length-of-stay question mid-round
Both fields update live, so an hour figure read off the notes becomes the decimal day a board round is discussing before the conversation has moved on.
Turn an expected stay back into hours
One press of the swap arrows converts a pathway written in days into the hour window that discharge planning, transport and pharmacy actually schedule against.
Weeks and minutes share the same screen
Searchable unit lists on both sides mean a 336-hour rehabilitation block reads as two weeks, and a 90-minute recovery window reads in minutes, without leaving the page.
Decimals that survive an average
Results carry up to eight decimals, so a cohort of short stays does not collapse to a column of zeroes and ones before anyone works out a mean.
Questions From the Bed Management Meeting
Why does a 30-hour stay show up as a single bed-day?
Because occupancy is counted midnight to midnight, not by duration. The convention counts the day of admission and excludes the day of discharge, and any part of a day counts as a whole one. A patient admitted Monday afternoon and discharged Tuesday evening was present at one midnight, so the census records one day even though 1.25 days elapsed. Run the same patient in an hour later and out an hour earlier and the census answer does not move at all — which is exactly why elapsed hours, not counted days, are the right input to a flow or capacity analysis.
What makes 24 hours such a hard line in the record?
It is the boundary most short-stay rules are built around. Coding guidance separates episodes of more than eight but fewer than 24 hours from those that run past a day, and observation pathways are frequently designed to conclude before the 24-hour mark so the episode stays outpatient. That is why 23-hour units exist as a named thing. The clinical decision should of course drive the discharge, but the clock is genuinely load-bearing here, so record admission and discharge times precisely rather than rounding them to the nearest hour.
How is a day case counted if no night is spent in a bed?
As an episode with a length of stay of zero days under midnight counting, which is why day-case activity is reported as a separate rate instead of being buried in the bed-day total. In elapsed terms it is nothing of the sort: a six-hour theatre-and-recovery pathway is 0.25 of a day of real capacity, and a unit running six of them through one bay has used 1.5 days of that bay. If your capacity model only reads counted days, high day-case throughput will look free and the bay will keep running out of slots.
Does an expectation of two midnights mean 48 hours?
No, and this is where the two counting styles diverge most sharply. Midnights are boundaries, not durations. Someone admitted at 23:00 crosses their second midnight only 25 hours later; someone admitted at 01:00 needs 47 hours to reach the same point. So a stay described as spanning two midnights can be anything from roughly 25 to 48 elapsed hours, and admission time alone shifts it by almost a full day. Convert the actual timestamps rather than assuming a midnight count implies a duration.
Why is average length of stay reported in days when wards work in hours?
Comparability, mostly. Days are the unit every benchmarking dataset, funding model and international comparison already uses, and a mean expressed in hours would suggest a precision the underlying records rarely support. The cost is that the distribution is heavily skewed: a long-stay patient at 336 hours outweighs a dozen at 6 hours, so a mean in days can rise while the typical patient's experience is unchanged. Keep the hour-level data, convert for reporting, and put a median and a long-stay count next to any average you publish.
No comments yet. Be the first to comment!