Short answer
An operating theatre can move out of band and back again entirely between two manual readings, leaving a compliant looking logbook and no record of the excursion. In comparable deployments we work to roughly 27 minutes as the duration at which an out of band humidity or pressure condition starts to matter for infection control, against manual rounds that typically run hours apart. Continuous monitoring closes that gap to about 60 seconds.
The arithmetic of sampling
Consider a theatre checked twice a shift. That is a reading roughly every four hours, so the record contains six data points a day and 23 hours and 50 minutes of nothing.
Now consider an event that develops over twenty minutes, sits out of band for half an hour and recovers over another twenty. It has a decent chance of falling entirely between two readings. Both readings are in band. The logbook is complete and compliant. The excursion is not in the record because the record was never looking.
What actually causes the drift
- Filter loading. Loading changes the pressure drop across the AHU, which changes delivered airflow, which changes both air change rate and the pressure differential the room can hold. It develops over weeks and crosses a threshold on one particular day.
- Humidifier and cooling coil interaction. Humidity is controlled by two subsystems working against each other. When one is slightly out of tune the room hunts, and the hunt is widest at low load, which often means overnight or between cases.
- Door discipline. A propped door during a turnover changes the pressure relationship immediately. It is entirely routine and no one records it.
- Load changes elsewhere on the AHU. A theatre sharing an air handling unit with other spaces is affected when demand elsewhere changes. The cause is in another room.
- Ambient conditions. In hot and humid climates the plant works closer to its limit for more of the year, so the margin absorbing all of the above is thinner.
Every one of these is gradual and none of them announces itself. That is the common thread: these are not faults that trip an alarm, they are conditions that develop.
What continuous monitoring changes
| Manual rounds | Continuous monitoring | |
|---|---|---|
| Resolution | A sample every few hours | A reading every 60 seconds |
| What the record proves | The room was in band when someone looked | What the room did for every minute it was in use |
| When you learn about an excursion | At the next round, or not at all | While it is happening, with time to act |
| Diagnosis | A number with no context | A trend that points at the filter, the coil or the door |
| Preparing for a survey | Assemble and transcribe logbooks | Export the period |
| Tuning HVAC for efficiency | Too risky to attempt | Possible, because any drift is visible immediately |
The point is the decision, not the data
It would be easy to read this as an argument for more measurement. It is not. It is an argument for moving a decision earlier.
The decision that matters is whether to start the next case in this room. Made with a reading from three hours ago, it is a guess dressed as a check. Made with a live readiness verdict and a trend, it is a decision. Everything else, the evidence pack, the accreditation record, the efficiency programme, follows from having built the capability to make that one decision properly.
How long does an environmental excursion have to last to matter?
The working figure we use from comparable deployments is roughly 27 minutes for an out of band theatre humidity or pressure condition. It is an illustrative planning value rather than a clinical threshold, and a site survey replaces it with your own baseline and your own protocol limits.
The useful comparison is not the exact number but the ratio: minutes against manual rounds that run hours apart.
Why do manual logbooks look compliant when there have been excursions?
Because a logbook records the moments someone was looking. An event that develops and recovers between two readings leaves both readings in band and no trace in between.
The record is honest. It is just far lower resolution than the phenomenon it is trying to capture.
What usually causes theatre humidity drift?
Most commonly filter loading changing delivered airflow, humidifier and cooling coil control hunting at low load, door discipline during turnovers, demand changes elsewhere on a shared air handling unit, and ambient conditions in hot humid climates thinning the margin that absorbs the rest.
All of them are gradual, which is why they are caught by trends rather than by threshold alarms.
