Only 60% of US hospital patients say the area around their room was always quiet at night, the lowest score of the eight measures in the latest national HCAHPS results (CMS HCAHPS National dataset and HCAHPS Summary of Survey Results, July 2026, covering October 2024 to September 2025 discharges). The World Health Organization sets 35 dB LAeq for rooms where patients are treated or observed, yet a survey of Johns Hopkins Hospital found average levels in the 50-60 dB(A) range with no location in compliance (Busch-Vishniac et al., Journal of the Acoustical Society of America 2005). One monitoring study logged 2,558,760 alarms in five adult ICUs over 31 days, and 88.8% of the annotated arrhythmia alarms were false (Drew et al., PLOS ONE 2014). We aggregated data from the World Health Organization, CMS, The Joint Commission, Cochrane, Critical Care Medicine, the American Journal of Critical Care, ECRI and more than a dozen peer-reviewed studies. For how the same WHO noise framework plays out outside hospital walls, see our aircraft noise exposure statistics.
TL;DR
- 60% of US patients say the area around their room was ‘always’ quiet at night, 10% say ‘sometimes’ or ‘never’ (CMS HCAHPS National dataset 2026).
- WHO guidance: 35 dB LAeq where patients are treated or observed, 30 dB LAeq with 40 dB LAmax in ward rooms at night (WHO Guidelines for Community Noise 1999).
- Bedside 24-hour averages in five UK ICUs ran from 54.1 to 59.9 dBA (Darbyshire and Young, Critical Care 2013).
- Daytime hospital sound levels rose by roughly 0.38 dB per year from 1960 to 2005 (Busch-Vishniac and Ryherd, Acoustics Today 2019).
- 2,558,760 alarms were recorded in five ICUs with 77 beds in 31 days, 187 audible alarms per bed per day (Drew et al., PLOS ONE 2014).
- 72% to 99% of clinical alarms are false (Sendelbach and Funk, AACN Advanced Critical Care 2013).
- 98 alarm-related sentinel events from 2009 to mid-2012, 80 of them deaths (The Joint Commission, Sentinel Event Alert 50, 2013).
- Noise explained 11.5% of arousals and 17% of awakenings in ventilated and non-ventilated ICU patients (Freedman et al., AJRCCM 2001).
- Dutch inpatients slept 83 minutes less in hospital than at home (Wesselius et al., JAMA Internal Medicine 2018).
- Earplugs or eye masks: delirium risk ratio 0.55, low-quality evidence (Hu et al., Cochrane Database of Systematic Reviews 2015).
- A night-time noise and light bundle cut ICU delirium incidence from 33% to 14% in a before-and-after study (Patel et al., Anaesthesia 2014).
1. Average Sound Levels vs WHO Guidelines
The WHO hospital targets are close to the noise floor of an empty room. In the UK study, a side room with every device switched off measured 34.1 dBA, while occupied units never dropped below 45 dBA LAeq, and never below 50 dBA next to the patient (Darbyshire and Young, Critical Care 2013). The guideline is less a ceiling that hospitals overshoot occasionally and more a level that a working ICU structurally cannot reach.
The other striking pattern is the lack of a night dip. The Johns Hopkins survey found that many units show little or no night-time reduction, and the Swiss data below show only about 5 dB separating day from night. Measurement method matters a great deal here, which is why our sound level meter and noise monitoring statistics are useful context for reading any single dBA figure.
| Metric | Value | Source |
|---|---|---|
| WHO guideline, rooms where patients are treated or observed | 35 dB LAeq | WHO, Guidelines for Community Noise 1999 |
| WHO guideline, hospital ward rooms at night | 30 dB LAeq with 40 dB LAmax | WHO, Guidelines for Community Noise 1999 |
| 24-hour average LAeq at the central station, five UK ICUs | 51.3 to 59.1 dBA | Darbyshire and Young, Critical Care 2013 |
| 24-hour average LAeq adjacent to the patient | 54.1 to 59.9 dBA | Darbyshire and Young, Critical Care 2013 |
| Average levels across Johns Hopkins Hospital units | 50-60 dB(A) range, no location WHO compliant | Busch-Vishniac et al., JASA 2005 |
| Long-run trend in daytime hospital levels, 1960 to 2005 | Roughly +0.38 dB per year | Busch-Vishniac and Ryherd, Acoustics Today 2019 |
| Swiss mixed ICU patient room LAeq, day / evening / night | 54.12 / 53.37 / 49.05 dBA | Naef et al., Frontiers in Medicine 2022 |
| Spanish surgical ICU, mean night-time LAFeq (22:00 to 08:00) | 57.3 dB | Guisasola-Rabes et al., J Clin Med 2022 |
Context note: sources do not agree on the exact WHO day and night split. The 1999 guideline text pairs 35 dB with treatment and observation rooms and 30 dB with ward rooms at night; the Swiss authors cite 45 dBA by day and 30 dBA by night instead. The trend line from the Acoustics Today review was later extended to 2017, and the authors say reported hospital Leq has kept rising for nearly 70 years. Primary measurement data: Darbyshire and Young, Critical Care 2013 and the WHO Guidelines for Community Noise.
2. Peak Levels and Where the Noise Comes From
Averages hide the events that actually wake people. Beside the patient in UK ICUs, more than 50% of sampled minutes contained a peak between 79.0 and 84.6 dBA, and peaks above 85 dBA occurred for about 25 minutes of every daytime hour (Darbyshire and Young, Critical Care 2013). Against a 40 dB LAmax night target, peaks of that size are not a margin any single fix can close.
The source mix points away from machines alone. In the Swiss room, talking and other human-generated sound filled 39.29% of a 24-hour recording at one bed, more than double the share of monitor alarms. On a US medical ward, patients named staff conversation more often than alarms as the noise that disrupted their sleep.
| Metric | Value | Source |
|---|---|---|
| Sampled minutes with LApeak between 79.0 and 84.6 dBA, beside the patient | More than 50% | Darbyshire and Young, Critical Care 2013 |
| Highest single peak recorded | 127.9 dBA | Darbyshire and Young, Critical Care 2013 |
| Daytime peaks above 85 dBA | About 25 minutes of every hour | Darbyshire and Young, Critical Care 2013 |
| ICU room mean vs peak levels | Mean generally less than 45 dB(A); peaks often greater than 85 dB(C) | Lawson et al., American Journal of Critical Care 2010 |
| Spanish surgical ICU, mean night-time LAFpeak | 75.6 dB | Guisasola-Rabes et al., J Clin Med 2022 |
| Talking and human-generated sound at one bed, 24 hours | 495 min (39.29%) | Naef et al., Frontiers in Medicine 2022 |
| Monitor alarms at the same bed, 24 hours | 245 min (17.6%) | Naef et al., Frontiers in Medicine 2022 |
| Ward patients citing staff conversation / roommates / alarms as sleep-disrupting noise | 65% / 54% / 42% | Yoder et al., Archives of Internal Medicine 2012 |
Outlier note: in the American Journal of Critical Care study, turning an alarm up raised peaks inside the patient room but did not raise them in the adjacent room, and closing the adjacent door did not cut the peaks. Louder alarms add noise for the patient without guaranteeing that staff next door hear them. Source-tagging method: Naef et al., Frontiers in Medicine 2022.
3. Alarm Volume and False Alarm Rates
The alarm problem is one of signal-to-noise in the statistical sense. In 31 days, five adult ICUs with 77 beds produced 2,558,760 alarms, and 88.8% of the 12,671 arrhythmia alarms that were annotated turned out to be false (Drew et al., PLOS ONE 2014). When most alarms carry no clinical information, the rational response for an overloaded clinician is to discount them, which is exactly the behavior alarm fatigue describes.
Across settings, the range of actionable alarms runs from under 1% to 36%, so even the best-performing units in the literature ask staff to respond to roughly two non-actionable alarms for every useful one.
| Metric | Value | Source |
|---|---|---|
| Total alarms, five adult ICUs, 77 beds, 461 patients, 31 days | 2,558,760 | Drew et al., PLOS ONE 2014 |
| Audible alarms and audible burden | 381,560; 187 per bed per day | Drew et al., PLOS ONE 2014 |
| Annotated arrhythmia alarms that were false positives | 88.8% of 12,671 | Drew et al., PLOS ONE 2014 |
| Accelerated ventricular rhythm alarms that were false | 94.8% | Drew et al., PLOS ONE 2014 |
| Clinical alarms that are false, research range | 72% to 99% | Sendelbach and Funk, AACN Advanced Critical Care 2013 |
| Actionable alarm proportion across hospital settings | Less than 1% to 36% | Paine et al., Journal of Hospital Medicine 2016 |
| Average alarm soundings per patient per day | Over 135; over 500 not unusual in ICUs | Busch-Vishniac 2015, cited in Acoustics Today 2019 |
| Alarms that result in no action being taken | Over 90% of the time | Cvach 2012, cited in Acoustics Today 2019 |
Context note: the 2,558,760 total in the PLOS ONE study includes inaudible text alerts; 79.4% of technical alarms were inaudible, which is why the audible count is far lower. The ‘two non-actionable alarms per useful one’ reading uses the 36% upper bound from the Journal of Hospital Medicine review: 64% non-actionable divided by 36% actionable is about 1.8.
4. Alarm Safety Events and Regulatory Response
The safety case rests on small counts with large consequences. Of 98 alarm-related sentinel events reported to The Joint Commission between January 2009 and June 2012, 80 ended in death (Sentinel Event Alert 50, 2013). The alert stresses that reporting is voluntary and that alarm events are underreported, so these are a floor, not an incidence rate.
Regulation followed quickly: the alarm National Patient Safety Goal took effect in 2014, and the same alert notes that ECRI had ranked clinical alarm conditions first or second on its annual hazard list. By ECRI’s 2026 list, no hazard is titled as an alarm issue, although the number four hazard (home diabetes technology recalls) cites alarms that are not communicated reliably. Alarm safety has moved from headline hazard to compliance routine, not to solved.
| Metric | Value | Source |
|---|---|---|
| Alarm-related sentinel events, Jan 2009 to Jun 2012 | 98 | The Joint Commission, Sentinel Event Alert 50, 2013 |
| Outcomes: death / permanent loss of function / extra care or stay | 80 / 13 / 5 | The Joint Commission, Sentinel Event Alert 50, 2013 |
| Respondents agreeing alarm safety is an important issue, Jan-Feb 2013 field review | About 88% of 2,700 | The Joint Commission, R3 Report Issue 5, 2013 |
| Alarm-related patient deaths in FDA MAUDE, Jan 2005 to Jun 2010 | 566 | The Joint Commission, Sentinel Event Alert 50, 2013 |
| Alarm signals that do not require clinical intervention | 85% to 99% | The Joint Commission, Sentinel Event Alert 50, 2013 |
| Hospital respondents calling alarm management a safety issue, 2012 field survey | 90% of almost 1,600 responses | The Joint Commission, R3 Report Issue 5, 2013 |
| Respondents with an organization-wide alarm management process | Fewer than 50% | The Joint Commission, R3 Report Issue 5, 2013 |
| NPSG.06.01.01 effective dates | Jan 1, 2014 (phase 1); Jan 1, 2016 (policies and staff education) | The Joint Commission, R3 Report Issue 5, 2013 |
Primary documents: Sentinel Event Alert 50, the R3 Report on alarm system safety and ECRI’s Top 10 Health Technology Hazards for 2026.
5. Noise, Sleep Arousals and Lost Sleep
Polysomnography studies make an uncomfortable point for anyone who blames ICU sleep loss entirely on noise. Environmental noise was responsible for only 11.5% of arousals and 17% of awakenings in 22 medical ICU patients (Freedman et al., AJRCCM 2001), and Gabor and colleagues concluded that noise and patient-care activities together account for less than 30% of arousals and awakenings. Most fragmentation comes from something other than sound, and the Gabor team pointed to other elements of the ICU environment and treatment.
Outside the ICU, the pattern is sharper. Ward patients in the loudest tertile of night-time noise slept 76 minutes less than those in the quietest tertile, and a 39-hospital Dutch survey found sleep cut by 83 minutes versus home. Self-reported sleep, as our sleep tracker accuracy statistics show for consumer devices, is not the same as measured sleep; the Spanish ICU study found nurses overestimated patients’ sleep quality by more than 8 points.
| Metric | Value | Source |
|---|---|---|
| Arousals / awakenings attributed to environmental noise, 22 MICU patients | 11.5% / 17% | Freedman et al., AJRCCM 2001 |
| Noise arousal index | 1.9 ± 2.1 arousals per hour of sleep | Freedman et al., AJRCCM 2001 |
| Share of total sleep time occurring during the day | 57 ± 18% | Freedman et al., AJRCCM 2001 |
| Sound elevations per hour of sleep; share of arousals and awakenings | 36.5 ± 20.1; 20.9 ± 11.3% | Gabor et al., AJRCCM 2003 |
| Patient-care activities’ share of arousals and awakenings | 7.1 ± 4.4% | Gabor et al., AJRCCM 2003 |
| Medical ward sleep in hospital vs self-reported baseline | 314 vs 382 minutes | Yoder et al., Archives of Internal Medicine 2012 |
| Loudest vs quietest night-noise tertile | -76 minutes of sleep (95% CI, -134 to -18 minutes) | Yoder et al., Archives of Internal Medicine 2012 |
| Dutch inpatients: sleep lost vs home; awakenings per night | 83 min (95% CI, 75-92); 3.3 vs 2.0 | Wesselius et al., JAMA Internal Medicine 2018 |
Context note: the ward study measured a mean night-time Leq of 38.2 dB, and 94% of nights exceeded recommendations for average noise (Yoder et al., Archives of Internal Medicine 2012). Primary ICU sleep studies: Freedman et al. 2001, Gabor et al. 2003 and Wesselius et al. 2018, where 70.4% of patients reported being woken by external causes.
6. Patient Experience: HCAHPS ‘Quiet at Night’ Scores
Patients notice. Quietness scored 60, the lowest of the eight HCAHPS measures nationally, 20 points below Communication with Nurses at 80 (HCAHPS Summary of Survey Results, July 2026 Public Report). Cleanliness, another environment measure staff control directly, scored 74.
Geography matters more than most patient-experience gaps. The state range runs from 48 in California, Connecticut and Massachusetts to 71 in Louisiana, a 23-point spread (71 minus 48). A gap that wide between states suggests that local factors, not a uniform national standard of care, drive how quiet a night in hospital feels.
| Metric | Value | Source |
|---|---|---|
| Patients saying the area around their room was ‘always’ quiet at night | 60% | CMS HCAHPS National dataset, Oct 2024 to Sep 2025 discharges |
| ’Usually’ quiet at night | 30% | CMS HCAHPS National dataset, Oct 2024 to Sep 2025 discharges |
| ’Sometimes’ or ‘never’ quiet at night | 10% | CMS HCAHPS National dataset, Oct 2024 to Sep 2025 discharges |
| Room and bathroom ‘always’ clean (comparison) | 74% | CMS HCAHPS National dataset, Oct 2024 to Sep 2025 discharges |
| National quietness score vs Communication with Nurses | 60 vs 80 | HCAHPS Summary of Survey Results, July 2026 |
| Highest state quietness scores | Louisiana 71; Mississippi and South Dakota 70 | HCAHPS Summary of Survey Results, July 2026 |
| Lowest state quietness scores | California, Connecticut and Massachusetts 48 | HCAHPS Summary of Survey Results, July 2026 |
| Participating hospitals; survey response rate | 3,949; 23% | HCAHPS Summary of Survey Results, July 2026 |
Context note: staff experience mirrors the patient scores. In a survey of 47 nurses, 91% reported that noise negatively affects them in their daily work (Ryherd et al. 2008, cited in Acoustics Today 2019). Data: CMS Provider Data Catalog, Patient survey (HCAHPS) - National and HCAHPS Summary Analyses. The 23% response rate is a reminder that HCAHPS reflects the minority of patients who return the survey.
7. Interventions: Quiet Time, Earplugs and Delirium
Delirium is why ICU noise is a clinical topic and not only a comfort one: a meta-analysis of 48 studies and 27,342 patients put the pooled prevalence of ICU delirium at 31% (95% CI 24-41) (Krewulak et al., Critical Care Medicine 2018). Against that baseline, the simplest intervention, a pair of earplugs, carries the most consistent signal, and the most consistent caveat.
Noise programs can lower sound levels, but the effect is modest and rarely tracked for long. A systematic review of 25 studies found that 14 cut mean levels significantly, by 1 to 16 dBA, with follow-up rarely past one month (Vreman et al., Critical Care Explorations 2023). Over 90% of nurses surveyed felt quiet times were useful (Weber and Ryherd 2016, cited in Acoustics Today 2019). Masking strategies sit next to these programs, and our white noise and sound conditioner statistics cover that adjacent market. None of the figures below is treatment advice; they report what the studies measured.
| Metric | Value | Source |
|---|---|---|
| Studies reporting a significant cut in mean sound level | 14 of 25 (56%); 1 to 16 dBA | Vreman et al., Critical Care Explorations 2023 |
| Single-patient room redesign | 72 to 56 dBA at one month | Kol et al., reported in Vreman et al. 2023 |
| Night noise-and-light bundle: mean sound level; delirium incidence | 68.8 to 61.8 dB; 33% (55/167) to 14% (24/171) | Patel et al., Anaesthesia 2014 |
| Earplugs or eye masks: delirium risk ratio | 0.55 (95% CI 0.38 to 0.80), 2 studies, 177 participants, low-quality evidence | Hu et al., Cochrane Database of Systematic Reviews 2015 |
| Earplugs or eye masks: total sleep time | +2.19 hours (95% CI 0.41 to 3.96), low-quality evidence | Hu et al., Cochrane Database of Systematic Reviews 2015 |
| Earplugs: delirium relative risk; compliance | 0.59 (95% CI 0.44 to 0.78); 86.9% | Litton et al., Critical Care Medicine 2016 |
| 2019-2024 RCTs of earplugs and masks (8 RCTs, 649 participants) | 3 of 6 showed prevention, 3 showed no effect | Yayla and Kaplan Serin, Croatian Medical Journal 2025 |
| Medical ICU sleep quality program: incidence of delirium/coma | OR 0.46 (95% CI 0.23-0.89; p = 0.02) | Kamdar et al., Critical Care Medicine 2013 |
Context note: the strongest delirium effects come from before-and-after designs, which cannot rule out secular trends. A Dutch interrupted time series of nocturnal sound reduction found the delirium rate falling faster after the protocol (slope change of -3.7% per time period, p=0.02) but no change in reported sleep quality (van de Pol et al., Intensive and Critical Care Nursing 2017). Primary reviews: Cochrane, Hu et al. 2015, Vreman et al. 2023, Yayla and Kaplan Serin 2025 and the NIHR Evidence summary of Litton et al., which notes all included studies had a high risk of bias.
Summary: Hospital and ICU Noise by the Numbers
| Metric | Value | Source |
|---|---|---|
| WHO guideline, rooms where patients are treated or observed | 35 dB LAeq | WHO, Guidelines for Community Noise 1999 |
| WHO guideline, ward rooms at night | 30 dB LAeq with 40 dB LAmax | WHO, Guidelines for Community Noise 1999 |
| 24-hour LAeq beside the patient, five UK ICUs | 54.1 to 59.9 dBA | Darbyshire and Young, Critical Care 2013 |
| Johns Hopkins Hospital average levels | 50-60 dB(A), no location WHO compliant | Busch-Vishniac et al., JASA 2005 |
| Swiss ICU room LAeq at night | 49.05 dBA | Naef et al., Frontiers in Medicine 2022 |
| Alarms in five ICUs over 31 days | 2,558,760 | Drew et al., PLOS ONE 2014 |
| False arrhythmia alarms (annotated) | 88.8% | Drew et al., PLOS ONE 2014 |
| Clinical alarms that are false | 72% to 99% | Sendelbach and Funk, AACN Advanced Critical Care 2013 |
| Actionable alarm proportion | Less than 1% to 36% | Paine et al., Journal of Hospital Medicine 2016 |
| Alarm-related sentinel events / deaths, 2009 to mid-2012 | 98 / 80 | The Joint Commission, Sentinel Event Alert 50, 2013 |
| Arousals / awakenings from environmental noise | 11.5% / 17% | Freedman et al., AJRCCM 2001 |
| Arousals and awakenings from sound elevations | 20.9% | Gabor et al., AJRCCM 2003 |
| Sleep lost in hospital vs home | 83 minutes | Wesselius et al., JAMA Internal Medicine 2018 |
| Patients saying ‘always’ quiet at night | 60% | CMS HCAHPS National dataset 2026 |
| State quietness range | 48 to 71 | HCAHPS Summary of Survey Results, July 2026 |
| ICU delirium pooled prevalence | 31% | Krewulak et al., Critical Care Medicine 2018 |
| Earplugs or eye masks, delirium risk ratio | 0.55 | Hu et al., Cochrane Database of Systematic Reviews 2015 |
| Bundle effect on delirium incidence | 33% to 14% | Patel et al., Anaesthesia 2014 |
Methodology and Sources
- World Health Organization, Guidelines for Community Noise, 1999 (hospital guideline values read in the guideline text and as quoted by Darbyshire and Young 2013).
- Darbyshire JL, Young JD, An investigation of sound levels on intensive care units with reference to the WHO guidelines, Critical Care 2013.
- Busch-Vishniac IJ et al., Noise levels in Johns Hopkins Hospital, Journal of the Acoustical Society of America 2005.
- Busch-Vishniac IJ, Ryherd EE, Hospital Soundscapes: Characterization, Impacts, and Interventions, Acoustics Today 2019 (source for the 1960-2005 trend and for the cited Busch-Vishniac 2015, Cvach 2012, Ryherd 2008 and Weber and Ryherd 2016 figures).
- Naef AC et al., Methods for Measuring and Identifying Sounds in the Intensive Care Unit, Frontiers in Medicine 2022.
- Guisasola-Rabes M et al., Noise Levels and Sleep in a Surgical ICU, Journal of Clinical Medicine 2022.
- Lawson N et al., Sound intensity and noise evaluation in a critical care unit, American Journal of Critical Care 2010.
- Yoder JC et al., Noise and sleep among adult medical inpatients: far from a quiet night, Archives of Internal Medicine 2012.
- Drew BJ et al., Insights into the problem of alarm fatigue with physiologic monitor devices, PLOS ONE 2014.
- Sendelbach S, Funk M, Alarm fatigue: a patient safety concern, AACN Advanced Critical Care 2013.
- Paine CW et al., Systematic Review of Physiologic Monitor Alarm Characteristics and Pragmatic Interventions to Reduce Alarm Frequency, Journal of Hospital Medicine 2016.
- The Joint Commission, Sentinel Event Alert 50: Medical device alarm safety in hospitals, 2013, and R3 Report Issue 5: Alarm system safety, 2013.
- ECRI, Top 10 Health Technology Hazards for 2026.
- Freedman NS et al., Abnormal sleep/wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit, American Journal of Respiratory and Critical Care Medicine 2001.
- Gabor JY et al., Contribution of the intensive care unit environment to sleep disruption in mechanically ventilated patients and healthy subjects, American Journal of Respiratory and Critical Care Medicine 2003.
- Wesselius HM et al., Quality and Quantity of Sleep and Factors Associated With Sleep Disturbance in Hospitalized Patients, JAMA Internal Medicine 2018.
- CMS Provider Data Catalog, Patient survey (HCAHPS) - National, modified July 2026, and HCAHPS Project, Summary of HCAHPS Survey Results, July 2026 Public Report.
- Krewulak KD et al., Incidence and Prevalence of Delirium Subtypes in an Adult ICU, Critical Care Medicine 2018.
- Vreman J et al., The Effectiveness of the Interventions to Reduce Sound Levels in the ICU: A Systematic Review, Critical Care Explorations 2023.
- Patel J et al., The effect of a multicomponent multidisciplinary bundle of interventions on sleep and delirium in medical and surgical intensive care patients, Anaesthesia 2014.
- Kamdar BB et al., The effect of a quality improvement intervention on perceived sleep quality and cognition in a medical ICU, Critical Care Medicine 2013.
- van de Pol I et al., Effect of nocturnal sound reduction on the incidence of delirium in intensive care unit patients, Intensive and Critical Care Nursing 2017.
- Hu RF et al., Non-pharmacological interventions for sleep promotion in the intensive care unit, Cochrane Database of Systematic Reviews 2015.
- Litton E et al., The Efficacy of Earplugs as a Sleep Hygiene Strategy for Reducing Delirium in the ICU, Critical Care Medicine 2016, figures read in the NIHR Evidence summary.
- Yayla EM, Kaplan Serin E, The effect of sleep-oriented non-pharmacological interventions in preventing delirium, Croatian Medical Journal 2025.
- Data watch: most measured sound-level data are older than three years. Most recent available data: Naef et al. and Guisasola-Rabes et al., 2022 (Guisasola-Rabes measured in 2017); the UK multi-ICU data date from 2012 and the alarm counts from 2013. Sources disagree on the WHO hospital values: the 1999 guideline pairs 35 dB LAeq with treatment and observation rooms and 30 dB LAeq plus 40 dB LAmax with ward rooms at night, Vreman et al. cite 35 dB by day and 30 dB by night, and Naef et al. cite 45 dBA by day and 30 dBA by night. The UK ICU paper is internally inconsistent on overnight peaks above 85 dBA (up to 16 times per hour in the abstract, at least once every 7 to 16 minutes in the discussion), so we report only the daytime figure. The Cochrane plain-language summary mentions three pooled trials, while the formal delirium analysis pools two; we use the formal analysis. False-alarm ranges differ by definition: 72% to 99% ‘false’ (Sendelbach and Funk) is not the same as 85% to 99% ‘not requiring clinical intervention’ (The Joint Commission). Sentinel event counts are voluntary reports, not incidence. HCAHPS quiet-at-night results refresh quarterly on Care Compare, and ECRI publishes a new hazard list each January.
- Last updated: October 18, 2026. We update this roundup quarterly, and the next refresh is expected when CMS publishes the next HCAHPS Public Report on Care Compare.