When the hospital never goes quiet: how noise affects patients’ sleep, stress and wellbeing
Alarms, conversations and sudden sound peaks can fragment rest. Understanding the soundscape helps remove avoidable noise without mistaking comfort for clinical recovery or compromising safety.
A hospital cannot be silent. Alarms must signal risk, staff need to coordinate and some care cannot wait. Yet there is considerable room between impossible silence and a soundscape that never settles.
For someone who is unwell, a door, a corridor conversation or a monitor may fragment sleep, obstruct communication and keep the body on alert. That does not make noise the sole explanation for stress, delirium or recovery. Pain, medication, light, care interruptions and the illness itself also matter.
Hospital noise comes from many places
A systematic review of 72 studies identified patient–staff conversations and medical alarms among the most frequently reported sources. Trolleys, doors, phones, ventilation, cleaning, deliveries, televisions, roommates and clinical procedures all add to the mix.
The review reported levels of 61–66 dB in intensive care and around 63 dB on inpatient wards. These figures combine measurements made with different equipment, periods and locations, so they do not represent every hospital or an individual patient’s experience. They do show why “please speak quietly” cannot solve a system-level problem.
An average cannot describe the whole soundscape
dB(A) weights sound to approximate human hearing. LAeq summarises average sound energy over a set period, but it can hide ten slamming doors. Maximum levels—often reported as LAFmax—also matter, alongside the number of events, their timing and the contrast between background sound and each peak.
The WHO Guidelines for Community Noise proposed 30 dB LAeq,16 h by day and 30 dB LAeq,8 h at night for hospital rooms, with a night-time LAmax of 40 dB. Published in 1999, these are reference values intended to protect rest. They are not a clinical boundary that makes 29 dB safe and 31 dangerous, and they never justify silencing a necessary alarm.
Room acoustics change the experience too. In a reverberant space, sound takes longer to fade and several sources overlap. A 24-hour measurement at the bedside, paired with an event log, reveals more than one reading in the corridor.
Sleep is the most studied outcome, but it has many drivers
A review of 229 studies on regular wards identified noise, light, care interruptions, pain and anxiety as common sleep disruptors. Several non-drug interventions showed promising results, but certainty ranged from moderate to very low and the studies differed substantially.
It is plausible for a sharp sound to cause an awakening or micro-arousal. Finding poorer sleep on a noisy ward, however, does not show that noise accounts for the entire deficit. Lowering sound levels without addressing pain, lighting or overnight observations may leave much of the problem untouched.
Stress, wellbeing and delirium: association is not causation
Unpredictable sounds may sustain alertness, interfere with conversation and reduce a person’s sense of control. A 2025 scoping review found associations between noise and poorer sleep, anxiety, higher heart or respiratory rate, and delirium. However, 21 of the 28 studies came from intensive care, almost 90% used cross-sectional analyses and fewer than half adjusted for confounding factors.
It would therefore be misleading to say that noise causes delirium or delays discharge. New confusion, disorientation, unusual drowsiness, hallucinations or marked inattention require clinical assessment. Reducing stimulation may support care, but it must not replace investigation of pain, infection, hypoxia, medication effects or other causes.
Fewer alarms must not mean less safety
The answer is not to turn volumes down indiscriminately. Organisations need to identify which alarms are actionable, tailor limits within agreed protocols, maintain sensors, define escalation and check that the alert reaches the right clinician.
In one pre-post ICU quality-improvement study, a bundle reduced monthly alarms from 600,452 to 115,927 without reported alarm-related adverse events. It did not reduce overall sound levels or improve perceived sleep, and the design cannot establish safety in other settings. Changing one source may not transform the wider soundscape.
Six steps towards a more supportive sound environment
- Measure before intervening. Combine LAeq, peaks, times, locations and source mapping rather than relying on one average.
- Remove avoidable sound. Check trolley wheels, doors, bins, phones, televisions, deliveries and conversations near bedrooms.
- Improve the acoustics. Use sound-absorbing materials compatible with infection and fire requirements, reduce reverberation and separate work from rest areas.
- Coordinate care. Cluster overnight interruptions where clinically appropriate, without delaying necessary observation or treatment.
- Govern alarms. Follow a formal safety process with clinical engineering and care teams; never make informal changes to safety systems.
- Personalise. Ask what is disturbing and offer earplugs, headphones or masking sounds only when tolerated and compatible with communication and observation.
A pilot should combine acoustic exposure, perceived sleep or comfort and balancing measures: missed alarms, response times, incidents and communication. If rest improves after a bundle, that is an observed change; attributing it to the intervention requires an appropriate comparative design.
Treating the hospital environment as a care variable means listening, measuring and adjusting without losing safety. NeumaCare is developing this approach. If your organisation wants to explore a sensory pilot, get in touch.
Frequently asked questions
How quiet should a hospital room be?
WHO proposed 30 dB LAeq,16 h by day, 30 dB LAeq,8 h at night and a night-time LAmax of 40 dB as reference values. They must be interpreted with the measurement position, peaks and clinical activity; they are not an individual threshold of harm.
Can hospital noise cause delirium?
An association has been observed, particularly in intensive care, but that does not establish direct causation. Delirium is multifactorial, and any acute change in attention or awareness needs clinical assessment.
Do earplugs or white noise work for everyone?
No. They may help some people, but they can reduce environmental awareness, obstruct communication or feel unpleasant. Offer rather than impose them, and check tolerance and clinical safety.
Can alarms simply be turned down to reduce noise?
No. Changes require a governed alarm-management process, including patient-specific limits, maintenance, escalation and confirmation that critical signals remain detectable.
Sources and further scientific reading
- World Health Organization. (1999). Guidelines for Community Noise.
- Pazani-Kohaki M, Yazdanfar SA, Lohrasbi L. (2026). A systematic review of noise pollution in hospitals. The Journal of the Acoustical Society of America.
- Razali A, Mazeli N. (2026). Hospital noise exposure: a systematic review of levels, sources, and effects on patients and staff. Medical Journal of Malaysia.
- Burger P et al. (2023). How to improve inpatient sleep in regular care wards. Acute Medicine.
- Sonoda N, Furuki H, Morimoto A. (2025). Impact of environmental noise on inpatient outcomes. Worldviews on Evidence-Based Nursing.
- Tronstad O et al. (2024). The introduction of a sound reduction bundle in the intensive care unit and its impact on sound levels and patients. Australian Critical Care.
- Fang CS et al. (2024). Effectiveness of sound and darkness interventions for critically ill patients’ sleep quality. Nursing in Critical Care.
- NHS England. Health Technical Memorandum 08-01: Acoustics.
- The Joint Commission. Sentinel Event Alert 50: Medical device alarm safety in hospitals.
- National Institute for Health and Care Excellence. Delirium: prevention, diagnosis and management. NICE CG103.
