From Measuring Hospital Noise to Managing It

Hospitals already know they have a noise problem, and hospital noise monitoring has traditionally focused on measuring sound levels.

The more useful question is: what is causing the noise, when does it occur, and did our intervention work?

Research across hospital and ICU environments repeatedly identifies conversations and medical equipment alarms among the major noise sources.[1] The problem is therefore not simply high sound levels. It is that different noise sources require different solutions.

Awareness can change behaviour

At Lancashire Teaching Hospitals in Preston, a SoundEar visual warning display reduced median overnight sound from 57.4 to 53.5 dB. Four months later, most of the reduction remained.[2]

The likely mechanism was behavioural: when noise becomes visible, people react to it.

A small NICU project at Rigshospitalet in Copenhagen showed the same pattern in practice. 11 of 14 staff members reported increased noise awareness and changes in behaviour, including moving noisy activities away from patient rooms.[3]

But the project also highlighted something important:

Not all sound is unwanted.

Parents talking or singing to a premature baby can be valuable. The target is unnecessary noise, not silence.[4]

Noise awareness at the NICU, Rigshospitalet – Copenhagen University Hospital.

Measurement still leaves one major question unanswered

A conventional monitor can tell us:

07:05 — 72 dB.

But it cannot tell us whether that peak was driven by human speech, an alarm, equipment, a rattling trolley or a closing door.

That distinction matters.

Speech may require a behavioural intervention. Alarm noise may require changes in alarm management. Equipment or doors may require investment rather than telling staff to be quieter.

This is the gap the next generation of hospital noise monitoring needs to close.

From sound level to sound source

SoundEar4 is being developed to classify excessive noise into 43 common sound categories, including speech, machinery, traffic, music and impact sounds.

The system analyses sound above the configured alarm threshold and identifies the category that dominated each minute. Louder events carry more weight. If no source clearly dominates, the result is reported as mixed rather than guessed.

Most importantly:

No audio is stored. Only the resulting category is retained.

The purpose is not to record conversations or identify individuals. It is to distinguish between different types of noise so the response can become more specific.

That changes the question from:

Who was making noise?

to:

What is creating the noise, and what can we change?

That difference matters for both staff trust and quality improvement.

Measure → Identify → Act → Verify

The real opportunity is to connect four steps:

Measure → Identify → Act → Verify

Visual feedback creates awareness.

Continuous monitoring shows when problems occur.

Source classification adds context.

Reporting shows whether the intervention worked.

That moves hospital noise management beyond collecting decibel values.

Hospitals do not need more proof that they are noisy.

They need better information about what is driving the noise, what can realistically be changed, and whether that change made a measurable difference.

Interested in how sound-source identification could support noise-reduction work in your ICU or NICU? We are currently developing and testing the next generation of SoundEar hospital monitoring. Contact us to learn more.

References

1. A systematic review of noise pollution in hospitals. Journal of the Acoustical Society of America. Review of 72 studies; reports ICU sound levels of 61–66 dB and identifies patient–staff conversations and medical-equipment alarms as the most frequently reported sources. DOI

2. Plummer NR et al. SoundEar noise warning devices cause a sustained reduction in ambient noise in adult critical care. Journal of the Intensive Care Society, 2019;20(2):106–110. DOI: 10.1177/1751143718767773. PubMed

3. Weis J, Haaber J. Reducing Noise Through Awareness in the NICU. Developmental Observer, 2019. The report documents the 14-person staff evaluation and the practical behavioural changes. IUScholarWorks

4. Rigshospitalet NICU project. The parent information material explains the distinction between meaningful sound and unnecessary noise and confirms that sound pressure is measured without recording audio. View NICU flyer