British Journal of Anaesthesia, 120 (3): 443e452 (2018)

doi: 10.1016/j.bja.2017.09.006
Advance Access Publication Date: 23 November 2017
Review Article

REVIEW ARTICLES

Systematic review of the effects of

intensive-care-unit noise on sleep of healthy
subjects and the critically ill
S. Horsten1, L. Reinke1,*, A. R. Absalom2 and J. E. Tulleken1
1
Department of Critical Care, University of Groningen, University Medical Center Groningen, NL-9713AV
Groningen, The Netherlands and 2Department of Anaesthesiology, University of Groningen, University
Medical Center Groningen, NL-9713AV Groningen, The Netherlands

*Corresponding author. E-mail: l.reinke@umcg.nl

Abstract
Intensive-care-unit (ICU) patients exhibit disturbed sleeping patterns, often attributed to environmental noise, although
the relative contribution of noise compared to other potentially disrupting factors is often debated. We therefore sys-
tematically reviewed studies of the effects of ICU noise on the quality of sleep to determine to what extent noise explains
the observed sleep disruption, using the Cochrane Collaboration method for non-randomized studies. Searches in
Scopus, PubMed, Embase, CINAHL, Web of Science, and the Cochrane Library were conducted until May 2017. Twenty
papers from 18 studies assessing sleep of adult patients and healthy volunteers in the ICU environment, whilst recording
sound levels, were included and independently reviewed by two reviewers. We found that the numbers of arousals
between the baseline and the ICU noise condition in healthy subjects differed significantly (mean difference 9.59; 95%
confidence interval 2.48e16.70). However, there was considerable heterogeneity between studies (I2 94%, P < 0.00001), and
all studies suffered from a considerable risk of bias. The meta-analysis of results was hampered by widely varying
definitions of sound parameters between studies and a general lack of detailed description of methods used. It is,
therefore, currently impossible to quantify the extent to which noise contributes to sleep disruption among ICU patients,
and thus, the potential benefit from noise reduction remains unclear. Regardless, the majority of the observed sleep
disturbances remain unexplained. Future studies should, therefore, also focus on more intrinsic sleep-disrupting factors
in the ICU environment.

Keywords: Arousal; Intensive care units; Noise; Sleep

Editorial decision: October 26, 2017; Accepted: December 1, 2017

© 2017 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved.
For Permissions, please email: permissions@elsevier.com

443
 444 - Horsten et al.
Key points

The authors performed a systematic review and meta-
analysis considering the effect of ICU noise levels on
patients’ quality of sleep.

They found wide variation in studies, preventing them
from making generalisable conclusions.

However, sleep disturbance remains a clear problem,
and the authors recommend further studies examining
the issue.
Sleep is an important process that is essential for repair and
survival.1 Disrupted sleep is associated with impaired immune
function and increased susceptibility to infections,2e4 alter-
ations in nitrogen balance and wound healing,2,4 and dimin-
ished neurophysiological organisation and memory
consolidation.3 In the intensive care unit (ICU), this may lead
to delirium, prolonged admission, and increased mortality.3
Unfortunately, most patients in the ICU exhibit disturbed
sleeping patterns1,2 characterized by severe fragmentation of
sleep.5 As part of a pilot study, we too found severely frag-
mented sleep and EEG activity that suggest heightened arousal
and signs of sleep deprivation.6
Patients admitted to an ICU are exposed to several intrinsic
and extrinsic sleep-disrupting factors, which were described
previously in more detail by Le Guen and colleagues.7 A
multitude of these factors, most of them interdependent,
likely causes the disrupted sleep observed in the ICU. The most
important environmental factors are assumed to be temper-
ature, light exposure, and noise, the latter of which is most
often associated with disturbed sleep.8,9 Although the exact
mechanism and the significance of sleep disruption by ICU
noise amongst patients are still debated, workplace noise is
known to have a negative effect on ICU staff causing irritation,
fatigue, concentration problems, headaches, and even
burnout.10e13
The 1999 World Health Organization guidelines for com-
munity noise recommend a maximum of 35 decibels, adjusted
for the range of normal hearing [dB(A)] overnight and 40 dB(A)
during the day for hospital environments.14 However, this is
not achievable in a modern ICU unless all equipment is
switched off.15 As a result, sound levels in ICUs far exceed the
recommended levels15e20 with average noise levels between
55 and 70 dB(A), accompanied by peak noise levels of more
than 80 dB(A).21 The Society of Critical Care Medicine’s
guideline for ICU design even states that increased noise levels
can disrupt sleep, although the cited sources do not provide
data on ICU patients’ sleep.22
Consequently, an increasing number of studies focus solely
on sleep disturbance by ICU noise specifically, disregarding
other environmental and illness-related changes that accom-
pany ICU admission. In order to know how to optimize ICU
architecture, improve technology, and guide staff behaviour to
promote sleep, it is crucial to know with a sufficient level of
evidence how large the impact of ICU noise on the quality of
sleep really is.12,22 The aim of our study was to systematically
review the available evidence on the effects of ICU noise on the
quality of sleep in healthy volunteers and ICU patients.
Methods
The Cochrane Collaboration method for non-randomized
studies was used for this systematic review.23
Eligibility criteria
We searched for studies assessing the sleep of adult patients
and healthy volunteers in the ICU environment objectively,
using methods, such as polysomnography (PSG), actigraphy, or
patient self-reports whilst the patient was in the ICU, with
simultaneous registration and recording of sound levels.
Studies were excluded if they met at least one of the following
criteria: included only neonates or children, and assessed sleep
or sound levels using subjective observation only. Although a
very informative method, the assessment of sleep by obser-
vation is known to significantly overestimate the total sleep
time and sleep continuity, and is generally considered to pro-
vide an inaccurate estimation of the quality of sleep.24 Finally,
it is vital that sound levels are objectively measured using
standard units to ensure that results from various studies can
be compared and data can be pooled for meta-analysis.
Outcome
The primary outcome was the number of arousals per hour of
sleep for different sound conditions. This outcome was chosen
because it best represents sleep quality in a single measure and
was, therefore, most commonly used in the reviewed articles.
Search strategy
A literature search was conducted using the following elec-
tronic databases: Scopus, PubMed, Embase, CINAHL, Web of
Science, and the Cochrane Library. The search terms used in
all of the databases were ‘sleep and (noise or sound) and (ICU,
intensive care, or critical care)’. The search was conducted
without any article format, data, or language restrictions, and
included studies published until May 2017.
Study selection
The titles for the articles retrieved from the search were
manually reviewed by two authors. After the removal of letters
to the editor, reviews, abstracts only, and non-article formats,
the remaining abstracts were assessed for eligibility. Only
abstracts of original investigations were included. The refer-
ences of all included articles and those from selected reviews
were checked for relevancy. The following data were extrac-
ted: year of publication, country in which the study was con-
ducted, period of conduct of the study, inclusion and exclusion
criteria, all outcomes, details on interventions, and charac-
teristics of the studies.
Bias risk assessment
Two authors independently assessed the risks of bias of the
studies following the domains from the Cochrane risk of bias
assessment tool: for non-randomized studies of in-
terventions.25 The domains are bias attributable to con-
founding, bias in the selection of participants into the study,
bias in measurement of interventions, bias attributable to
departures from intended interventions, bias attributable to
missing data, bias in measurement of outcomes, and bias in
the selection of the reported results.
Statistical analysis
A meta-analysis on data from studies that measured the
number of arousals per hour of sleep for multiple settings was

performed using the software package Review Manager 5.3
(The Nordic Cochrane Centre, Copenhagen, Denmark). Results
are presented as mean difference with 95% confidence interval
(CI). We calculated a random-effects model. Heterogeneity
was explored by the c2 test with significance set at a P value of
0.05. The quantity was measured with I2.
Results
The search initially returned 1397 hits. After the removal of du-
plicates, 854 citations remained. After the screening of titles and
abstracts, a total of 40 full-text articles were retrieved. Of these, a
total of 20 papers from 18 studies met the eligibility criteria. A
manual search of the references of the included articles and of 34
relevant reviews resulted in the inclusion of four more relevant
reviews whose reference lists were also searched. A flow chart of
the study inclusion is presented in Fig 1.
Study characteristics
Patients
Eleven papers on outcomes from 10 studies concerning ICU
patients were retrieved with a total number of 599 included
Figure 1. Flow chart of study inclusion.
Sleep and noise in the ICU: a systematic review - 445
patients. However, outcomes were only reported on data from
295 subjects; 304 subjects did not complete the study they
were in, of which 279 dropped out of a single study.26 Four
studies were observational,8,27e29 three were crossover stud-
ies,30e32 two used a before-and-after intervention design,26,33
and one was a randomized controlled trial.34 Further charac-
teristics on the studies can be found in Table 1.
Healthy volunteers
Ten papers on outcomes from nine studies concerning healthy
volunteers were found with data on 263 subjects from a total of
268 included; five had repeated-measures designs,31,35e38 two
were crossover studies,27,39 and two used a post-test-only
control group design.40,41 Further characteristics on the
studies can be found in Table 2.
Bias risk assessment
Patients
All studies involving patients were judged to have some risk of
bias for confounding. No study had a low risk of bias for con-
founding (0%), five had a low risk of selection bias (50%), none
Table 1 Characteristics of included studies: patients. ICU, intensive care unit; IQR, inter-quartile range; PSG, polysomnography; RCSQ, RichardseCampbell Sleep Questionnaire; RCT,

446
randomized controlled trial; VAS, visual analogue scale

-
Study (year published) Design Method Groups No patients Mean Mean ICU stay Mean noise level Studied intervention Quality of sleep related to

Horsten et al.
(included) age (SD) (yr) before study sound environment?
± SD (DAYS)

Bosma and colleagues Crossover Overnight d 13 (16) 63 (13) 22 (11) Night: z60 dB Ventilator modes Patienteventilator
(2007)30 PSG discordance causes sleep
disruption.
Boyko and colleagues Crossover RCT 24 h PSG Control 17 (19) 67 (median, 3 (median, 48 dB(A) ‘Quiet routine’ Unable to further reduce
(2017)32 Intervention range 26e85) range 0e17) 47 dB(A) protocol already low noise levels;
no association between
the intervention and the
presence of normal PSG
sleep characteristics
Elbaz and colleagues Observational 24 h PSG d 11 64 (14) d Day: 74 dB(C) d Sleep continuity is
(2017)29 Night: 70 dB(C) disturbed by alarms.
Alarms and emergency
signalling need to be
evaluated.
Elliott and colleagues Observational 24 h PSG d 53 (57) 60 (20) 5 (IQR 2.5e11) Day: 54 dB(A) d Results suggest that a
(2013 and 2014)28,48 RCSQ Night: 50 dB(A) sound reduction
programme is required.
Freedman and Observational 24 h PSG d 22 (24) 61 (16) 18 (20) Day: 59 dB(A) d The impact of
colleagues (2001)8 Night: 57 dB(A) environmental noise
appears much less
important than
previously described.
Gabor and colleagues Observational 24 h PSG d 7 57 (19) 48 (40) Day: 56 dB d Only 30% of observed sleep
(2003)27 Night: 54 dB disruption was
accounted for by sound
and patient-care
activities.
Hu and colleagues RCT RCSQ Control 25 (25) 57 (11) 0 (0) Night: 70 dB(A) Earplugs, eye masks, The studied intervention is
(2015)34 Intervention 20 (25) 57 (11) 0 (0) Night: 70 dB(A) and relaxing useful for promoting
background sleep.
music
Li and colleagues Before and after RCSQ Before 27 (30) 51 (2) d 58 dB Reduce noise, light Environmental noise and
(2011)33 intervention After 28 (30) 49 (6) d 50 dB exposure, and care night-time care activities
activities at night play a role in disrupting
sleep.
Patel and colleagues Before and after RCSQ Before 30 (167) 62 (13) d Night: 69 dB Reduce noise, light Significant improvement
(2014)26 intervention After 29 (171) 61 (16) d Night: 62 dB exposure, and care in patients’ perception of
activities at night sleep
Wallace (1998)49 Crossover Overnight d 13 (17) 57 (20) 13 (8) Night: 58 dB(A) Earplugs All subjects experienced
PSG severely disturbed sleep,
with or without earplugs.
Table 2 Characteristics of included studies: healthy volunteers. ICU, intensive care unit; PSG, polysomnography; REM, rapid eye movement; VAS, visual analogue scale

Study (year published) Design Method Groups No healthy Mean Mean noise level Studied intervention Quality of sleep related to
volunteers age (SD) (yr) sound environment?
(included)

Gabor and colleagues Crossover 24 h PSG 6 Range 23e65 Day: single-patient Single-patient room vs Subjects slept relatively well.
(2003)27 Questionnaire room, 44 dB open-plan ICU bed Noise accounted for a
Night: single-patient significant proportion of
room, 43 dB sleep disruption, but not to
Day: open-plan ICU, pathological extent.
56 dB
Night: open-plan ICU,
51 dB
Hu and colleagues (2010)35 Repeated Overnight PSG 14 (15) 31 (16) Night: baseline, Quiet vs recorded ICU With noise and light
measures VAS 34 dB(A) noise vs recorded ICU conditions, subjects had
Night: noise, 66 dB(A) noise with earplugs poorer perceived sleep
and eye masks quality and suffered from
sleep disruption.
Huang and colleagues Crossover Overnight PSG 40 (40) 41 (12) Night: baseline, not Quiet vs recorded ICU Nocturnal sleep is disturbed in
(2015)39 VAS reported noise and light healthy subjects with
Night: noise, 67 dB(A) conditions exposure to simulated ICU
noise and light.
Persson Waye and Repeated Overnight PSG 17 (18) 23 (18e30) Night: baseline, Quiet vs recorded ICU The ICU sound condition
colleagues (2013)36 measures Questionnaire 20.0 dB(A) noise e 7 dB vs significantly impaired the
Night: ICU, 47 dB(A) recorded ICU noise restorative functions of
peak reduced sleep. Subjective data
supported PSG findings.

Sleep and noise in the ICU: a systematic review

Snyder-Halpern (1985)38 Repeated Questionnaire 10 Range 20e24 Night: baseline < 65 dB Quiet vs recorded ICU Physical and physiological
measures Night: noise, 76 dB noise alternations can occur when
noise interferes with sleep.
Stanchina and colleagues Repeated Overnight PSG 5 (8) 27 (2) Night: baseline, not Quiet vs recorded ICU Peak noise was not the main
(2005)37 measures Subjective reported noise vs recorded ICU determinant of sleep
sleep quality Night: noise, 58 dB noise with white disruption. The percentage
Night: noise þ white noise of arousals associated with
noise, 61 dB noise is substantially greater
compared to recent reports.
Topf (1992)40 and Topf and Post-test only Questionnaire Baseline 105 total 36 Night: not reported Quiet vs recorded ICU Convincing support for causal
Davis (1993)50 control group ICU noise 36 Night: 56 dB noise vs recorded ICU relationship
noise with personal
control over noise
Topf and colleagues Post-test only Questionnaire Baseline 33 36 Night: not reported Quiet vs recorded ICU Support for the hypothesis
(1996)41 control group ICU noise 27 36 Night: 56 dB noise that subjects exposed to ICU
sounds exhibit poorer sleep
Wallace and colleagues Repeated Overnight PSG 6 25 (3) Night: baseline, Quiet vs noise vs quiet Results suggest that sleep is
(1999)31 measures 38 dB(A) with earplugs vs disrupted by exposure to
Night: noise, 62 dB(A) noise with earplugs simulated ICU noise, and
use of earplugs results in
more REM sleep.

-
447
 448 - Horsten et al.
had a low risk of measurement bias (0%), six had a low risk of
bias attributable to departures from intended interventions
(60%), seven had a low risk of bias caused by missing data
(70%), four had a low risk of outcome bias (40%), and all studies
had a low risk of reporting bias (100%). These results are
summarized in Fig 2a.
Healthy subjects
Fig. 2b gives an overview of the bias assessment of the studies
involving healthy subjects on seven domains. Only four
studies were judged to have a low risk of bias for confounding
(44%), and seven had a low risk of selection bias (78%). There
were no studies with a low risk of measurement bias (0%), but
eight were found to have a low risk of bias attributable to de-
partures from intended interventions (89%). Four studies had a
low risk of bias caused by missing data (44%), and the same
amount of studies had a low risk of outcome bias (44%). Two-
thirds of the studies were judged to have a low risk of reporting
bias (67%).
Outcomes
The mean and 95% CI of the difference between the number of
arousals per hour of sleep during the baseline setting and the
Figure 2. (A) Risk of bias assessment for patient studies and (B) studies involving healthy volunteers. Green is low risk, red is high risk, and
yellow is unknown risk.
ICU noise setting, for six studies with healthy volunteers that
reported this outcome, are presented in Fig 3. For the study by
Gabor and colleagues,27 the baseline condition was a single
room and the ICU noise condition an open ICU. For all other
studies, the baseline condition was a quiet environment in a
sleep laboratory and the ICU noise condition consisted of ICU
noises played back in the same sleep laboratory. Persson Waye
and colleagues36 reported the total number of arousals for the
study night, whilst in the other studies the arousal index
(number of arousals per hour of sleep) was reported. There
was a significant difference in the number of arousals between
the baseline and the ICU noise condition (mean difference
9.59; 95% CI 2.48e16.70). There was, however, also consider-
able heterogeneity (I2 94%, P < 0.00001).
Discussion
Our review of the effect of noise on sleep in the ICU shows that
ICU noise seems to have a significant effect on the occurrence
of arousals in six studies performed with healthy volunteers,
in which the effect of the noise level was investigated. How-
ever, the majority of the observed arousals remain unex-
plained because they did not occur within 3 s of a sound peak.
The considerable heterogeneity that was found may be caused
by the large differences in study protocols. Twenty papers
 Sleep and noise in the ICU: a systematic review - 449

fulfilled our inclusion criteria, of which 11 contained data on

Figure 3. Forest plot of studies conducted in healthy volunteers, comparing the arousal index during the baseline and ICU noise conditions. Except for the study by Gabor and colleagues,27 all
studies were conducted in a laboratory setting. The size of squares for mean difference reflects the relative weight of the study in the pooled analyses. Horizontal bars span the 95% CIs. CI,
patients and 10 on healthy volunteers. There were no studies
that reported objective sleep measurements under different
noise conditions in patients. We have summarized some
methodological issues and potential solutions of the reviewed
papers in Table 3, and will discuss the individual risks for bias
in more detail.
All currently reviewed evidence of the effects of noise on
the quality of sleep of ICU patients is subject to considerable
risks of bias. Firstly, because of the multifactorial nature of ICU
sleep disruption, it is difficult to correct for most confounders.
This has led us to conclude that all research studying ICU
patients is inherently sensitive to bias attributable to con-
founding. In healthy subjects, this is less of a limitation
because they are not affected by any underlying illness.
Secondly, obtaining consent from ICU patients or their
families during an inherently stressful ICU admission may
cause selection bias, especially if a small number of patients
were included over a relatively long period of time. However,
as most studies reviewed here used a repeated-measures or
crossover design, they were assessed as having a low risk of
selection bias.
Thirdly, sound levels were not always measured for all
groups, leading to a high risk of bias for the measurement of
the intervention. Furthermore, the outcomes of sound mea-
surements are known to often be computed incorrectly,18
although we were not able to determine the exact method of
sound data analysis in most papers. Some studies required
nurses to keep a record of each patient’s care activities, whilst
others placed dedicated observers in the ICU. This poses a risk
of the Hawthorne or observer effect (i.e. that environmental
conditions are unintentionally altered by the presence of an
observer). Indeed, in the report of a study that focused on
identifying noise in the ICU, it was mentioned that the hospital
staff suggested that the noise levels during the period when
observers were present were not as high as normally experi-
enced.17 Preventing this effect is especially important in
studies assessing the effectiveness of an implemented inter-
vention, such as noise reduction. If personnel, even uncon-
sciously, alter their behaviour because they have been made
confidence interval; ICU, intensive care unit; IV, independent variable.

aware of the topic of noise and interruptions, the effects

cannot be measured reliably and representatively. Further-
more, not all papers mentioned if or which data were missing.
Fourthly, the risk of bias in the measurement of outcomes
was considered high when subjective methods, such as
questionnaires, were used. The intuitive relation between
noise and sleep disruption in healthy subjects is common
knowledge, and thus, subjects can be expected to have pre-
conceptions, further increasing the risk of bias when instruc-
ted on the goals of the study. Another methodological aspect
that we looked at is whether the PSG recordings were scored
blindly or not. This is important in studies with multiple
groups, but it was not applied in all studies with such a design.
Finally, very few indications of bias in the selection of reported
results were found.
Because of these concerns, it is currently difficult to
determine the true effect of noise in the ICU environment on
sleep in patients, or the relative importance in a plethora of
potentially disturbing influences. Although a significant effect
was found in healthy volunteers, all, but one, of these studies
were performed in a sleep laboratory and not in the actual ICU.
In recordings of healthy volunteers’ sleep, around 60%27 of
arousals were immediately preceded by noise events, whilst
several studies in ICU patients have reported that only 11e30%
 450 - Horsten et al.

Table 3 Main methodological issues with past studies and potential solutions for future studies

Methodological issue Possible solution or preventive measure

Confounding Study healthy volunteers or a larger sample of ICU patients; collect data on
confounding factors for correction or multivariate analyses.
Selection bias Use repeated-measures or crossover design.
Observer effect Use telemetry, local automated measurements, long habituation periods, or blinded
measurement intervals.
Incorrect calculation of sound Involve acoustician or equivalent expert in study design and sound data analysis.
parameters
Small sample sizes Develop or validate affordable automated sound- and sleep-recording devices and
analysis techniques.
Measurement bias Use objective measurement and scoring methods, and apply blinded scoring.
Low repeatability of sound Detailed reporting of materials and methods, and calculation of sound parameters.
recording

of sleep disruptions observed in the EEG could be attributed to
environmental noise.8,42 This suggests that other factors pre-
sent in patients might be more significant in disturbing sleep.
The importance of other ICU-related factors on the
observed disturbance of sleep is also suggested by the results
of a recently published Cochrane Review43 on the efficacy of
non-pharmacological interventions for promoting sleep in
critically ill adults. They found some evidence that these in-
terventions can provide small improvements in subjective
measures of sleep quality and quantity, but the quality of the
evidence again was low. The effects on objective sleep out-
comes were inconsistent across 16 studies. Four of the studies
investigated the use of earplugs or eye masks, or both in a total
of 141 subjects. In the majority of these studies, no benefit was
found. The cause of non-response to these interventions re-
mains unclear, although the high risk of bias probably
contributed.
For future investigation of the relationship between sound
and sleep in a clinical setting, we recommend sufficiently
powered (large) sample sizes. Half of the studies included in
this review had a sample size of no more than 20 subjects,
which precludes a detailed analysis. Because there are so
many difficulties in measuring and correcting for confounders
present in the ICU patient population, studies focusing on
healthy volunteers in the real ICU environment, or a combi-
nation of healthy volunteers and patients in the same study,
are perhaps best suited to study to what extent noise is a
sleep-disrupting factor in the ICU.
Additionally, it is also important to pay special attention to
complete and correct execution and description of sound
measurements to facilitate pooling of data and meta-analysis.
Measurement procedures were often unclear with limited
specification of parameters, time constants used, frequency
weighting used, and averaging method. Furthermore, most
studies only focused on noise amplitude, but not on other
relevant acoustic parameters, such as the acoustic spectrum,
reverberation time, perceived loudness, and entropy. The
sound spectrum, for instance, which shows the relationship
between sound level and frequency, is known to be important
for sound perception, but was nevertheless not reported.
Reverberation time, defined as the time taken for sound to decay
by 60 dB once the source has stopped, is similarly important
but underreported.44 Reducing the reverberation time im-
proves speech intelligibility and room acoustics by making
noises sound less harsh, which may play an important role in
reducing the impact of environmental noise on sleep.45
Perhaps more importantly, the information content and den-
sity of sound may also play a part in the degree of sleep
disruption.8,46 Sounds that have a specific meaning, such as
spoken language, are more likely to evoke an EEG potential.47
Generally, it is important for future studies to focus on using
objective measurement methods and ensure that PSG scoring
is performed blinded as much as possible. Although PSG is an
objective measuring method, the scoring of sleep stages is still
a manual process whereby bias can be introduced if data sets
are not presented randomly.
The current evidence on the effects of noise on the quality
of sleep is subject to considerable risks of bias. The limited
meta-analysis that was possible showed a significant in-
crease in arousals during the ICU noise condition in healthy
subjects, but there were no studies that reported on objec-
tively measured quality of sleep of ICU patients under mul-
tiple objectively measured noise conditions. Although this
meta-analysis of results obtained with healthy volunteers
suggests a potential benefit from noise reduction for healthy
individuals, the results obtained in this small combined
sample do not warrant the current narrow focus on noise as
the main sleep-disrupting factor in the ICU population. Future
studies should include sufficiently large sample sizes, and
pay special attention to complete and correct execution and
documentation of sound measurements, to facilitate pooling
of data and meta-analysis. This will enable us to determine
whether the current focus on noise reduction in the design of
new ICUs to improve our patients’ sleep is evidence based.
Because of the highly complex nature of acoustics and its
mechanisms to influence sleep, it is not possible at this
moment to indicate the extent to which noise reduction will
benefit patients, although the well-being of ICU staff favours
noise reduction regardless. Thus, it seems crucial to widen
the scope of ICU sleep research to include other potentially
sleep-disruptive factors, both environmental and related to
critical illness.
Authors’ contributions
Study selection and analysis, writing up of the first draft of the
paper, critical revision, and final approval of the article: S.H.
Study selection and analysis, critical revision, and final
approval of the article: L.R.
Data interpretation, critical revisions, and final approval of the
article: A.A.

Conception of the work, critical revision of the article, and final
approval of the article: J.T.
Declaration of interest
S.H., A.A., and J.T. have no interests to declare. L.R. has
received partial funding by Philips Research Eindhoven for a
PhD position at the University Medical Center Groningen.
References
1. Kamdar BB, Needham DM, Collop NA. Sleep deprivation in
critical illness. J Intensive Care Med 2012; 27: 97e111
2. Friese RS, Diaz-Arrastia R, McBride D, Frankel H,
Gentilello LM. Quantity and quality of sleep in the surgical
intensive care unit: are our patients sleeping? J Trauma
2007; 63: 1210e4
3. Boyko Y, Ording H, Jennum P. Sleep disturbances in crit-
ically ill patients in ICU: how much do we know? Acta
Anaesthesiol Scand 2012; 56: 950e8
4. Cooper AB, Thornley KS, Young GB, Slutsky AS,
Stewart TE, Hanly PJ. Sleep in critically ill patients
requiring mechanical ventilation. Chest 2000; 117: 809e18
5. Drouot X, Roche-Campo F, Thille AW, et al. A new clas-
sification for sleep analysis in critically ill patients. Sleep
Med 2012; 13: 7e14
6. Reinke L, van der Hoeven JH, van Putten MJ, Dieperink W,
Tulleken JE. Intensive care unit depth of sleep: proof of
concept of a simple electroencephalography index in the
non-sedated. Crit Care 2014; 18: R66
7. Le Guen M, Nicolas-Robin A, Lebard C, Arnulf I,
Langeron O. Earplugs and eye masks vs routine care pre-
vent sleep impairment in post-anaesthesia care unit: a
randomized study. Br J Anaesth 2014; 112: 89e95
8. Freedman NS, Gazendam J, Levan L, Pack AI, Schwab RJ.
Abnormal sleep/wake cycles and the effect of environ-
mental noise on sleep disruption in the intensive care
unit. Am J Respir Crit Care Med 2001; 163: 451e7
9. Tembo AC, Parker V. Factors that impact on sleep in
intensive care patients. Intensive Crit Care Nurs 2009; 25:
314e22
10. Ryherd EE, Waye KP, Ljungkvist L. Characterizing noise
and perceived work environment in a neurological
intensive care unit. J Acoust Soc Am 2008; 123: 747e56
11. Fredriksson S, Hammar O, Toren K, Tenenbaum A,
Waye KP. The effect of occupational noise exposure on
tinnitus and sound-induced auditory fatigue among ob-
stetrics personnel: a cross-sectional study. BMJ Open 2015;
5, e005793
12. Ferri M, Zygun DA, Harrison A, Stelfox HT. Evidence-based
design in an intensive care unit: end-user perceptions.
BMC Anesthesiol 2015; 15: 57
13. Raffier-Peres D, Estryn-Behar M. Audit ergonomique en
animation polyvalente. Bruit ambiant, alarmes et
re
impact sur le travail. Arch des Mal Prof l’Environnement 2015;
76: 115e30
14. World Health Organization. Guidelines for community noise.
Ecotoxicol Environ Saf. 1999. p. 55e65
15. Darbyshire JL, Young JD. An investigation of sound levels
on intensive care units with reference to the WHO
guidelines. Crit Care 2013; 17: R187
16. Christensen M. Noise levels in a general intensive care
unit: a descriptive study. Nurs Crit Care 2007; 12: 188e97
Sleep and noise in the ICU: a systematic review - 451
17. MacKenzie DJ, Galbrun L. Noise levels and noise sources in
acute care hospital wards. Build Serv Eng Res Technol 2007;
28: 117e31
18. Busch-Vishniac IJ, West JE, Barnhill C, Hunter T,
Orellana D, Chivukula R. Noise levels in Johns Hopkins
Hospital. J Acoust Soc Am 2005; 118: 3629e45
19. Tegnestedt C, Günther A, Reichard A, et al. Levels and
sources of sound in the intensive care unitdan observa-
tional study of three room types. Acta Anaesthesiol Scand
2013; 57: 1041e50
20. Pulak LM, Jensen L. Sleep in the intensive care unit: a re-
view. J Intensive Care Med 2016; 31: 14e23
21. Pugh R, Jones C, Griffiths R. The impact of noise in the
intensive care unit. Intensive Care Med 2007; 2007: 942e9.
https://doi.org/10.1007/978-0-387-49518-7_85
22. Thompson DR, Hamilton DK, Cadenhead CD, et al.
Guidelines for intensive care unit design. Crit Care Med
2012; 40: 1586e600
23. Higgins JPT, Green S. Cochrane handbook for systematic re-
views of interventions. Wiley; 2011. Available from: http://
handbook.cochrane.org
24. Beecroft JM, Ward M, Younes M, Crombach S, Smith O,
Hanly PJ. Sleep monitoring in the intensive care unit:
comparison of nurse assessment, actigraphy and poly-
somnography. Intensive Care Med 2008; 34: 2076e83
25. Sterne J, Higgins J. Reeves BC on behalf of the development
group for ACROBAT-NRSI. A Cochrane risk of bias assessment
tool: for non-randomized studies of interventions (ACROBAT-
NRSI). 2014. Available from: http://www.riskofbias.info.
[Accessed 1 March 2017]
26. Patel J, Baldwin J, Bunting P, Laha S. The effect of a
multicomponent multidisciplinary bundle of in-
terventions on sleep and delirium in medical and surgical
intensive care patients. Anaesthesia 2014; 69: 540e9
27. Gabor JY, Cooper AB, Crombach SA, et al. Contribution of
the intensive care unit environment to sleep disruption in
mechanically ventilated patients and healthy subjects.
Am J Respir Crit Care Med 2003; 167: 708e15
28. Elliott R, McKinley S, Cistulli P, Fien M. Characterisation of
sleep in intensive care using 24-hour polysomnography:
an observational study. Crit Care 2013; 17
29. Elbaz M, Le  ger D, Sauvet F, et al. Sound level intensity
severely disrupts sleep in ventilated ICU patients throughout
a 24-h period: a preliminary 24-h study of sleep stages and
associated sound levels. Ann Intensive Care 2017; 7: 25
30. Bosma K, Ferreyra G, Ambrogio C, et al. Patienteventilator
interaction and sleep in mechanically ventilated patients:
pressure support versus proportional assist ventilation.
Crit Care Med 2007; 35: 1048e54
31. Wallace CJ, Robins J, Alvord LS, Walker JM. The effect of
earplugs on sleep measures during exposure to simulated
intensive care unit noise. Am J Crit Care 1999; 8: 210e9
32. Boyko Y, Jennum P, Nikolic M, Holst R, Oerding H, Toft P.
Sleep in intensive care unit: the role of environment. J Crit
Care 2017; 37: 99e105
33. Li SY, Wang TJ, Vivienne Wu SF, Liang SY, Tung HH. Ef-
ficacy of controlling night-time noise and activities to
improve patients’ sleep quality in a surgical intensive care
unit. J Clin Nurs 2011; 20: 396e407
34. Hu RF, Jiang XY, Hegadoren KM, Zhang YH. Effects of
earplugs and eye masks combined with relaxing music on
sleep, melatonin and cortisol levels in ICU patients: a
randomized controlled trial. Crit Care 2015; 19: 1e9
 452 - Horsten et al.
35. Hu RF, Jiang XY, Zeng YM, Chen XY, Zhang YH. Effects of
earplugs and eye masks on nocturnal sleep, melatonin
and cortisol in a simulated intensive care unit environ-
ment. Crit Care 2010; 14: R66
36. Persson Waye K, Elmenhorst E-M, Croy I, Pedersen E.
Improvement of intensive care unit sound environment
and analyses of consequences on sleep: an experimental
study. Sleep Med 2013; 14: 1334e40
37. Stanchina ML, Abu-Hijleh M, Chaudhry BK, Carlisle CC,
Millman RP. The influence of white noise on sleep in
subjects exposed to ICU noise. Sleep Med 2005; 6: 423e8
38. Snyder-Halpern R. The effect of critical care unit noise on
patient sleep cycles. Crit Care Q 1985; 7: 41e51
39. Huang HW, Zheng BL, Jiang L, et al. Effect of oral mela-
tonin and wearing earplugs and eye masks on nocturnal
sleep in healthy subjects in a simulated intensive care
unit environment: which might be a more promising
strategy for ICU sleep deprivation? Crit Care 2015; 19: 124
40. Topf M. Effects of personal control over hospital noise on
sleep. Res Nurs Health 1992; 15: 19e28
41. Topf M, Bookman M, Arand D. Effects of critical care unit
noise on the subjective quality of sleep. J Adv Nurs 1996;
24: 545e51
42. Gabor JY, Cooper AB, Hanly PJ. Sleep disruption in the
intensive care unit. Curr Opin Crit Care 2001; 7: 21e7
43. Hu R-F, Jiang X-Y, Chen J, et al. Non-pharmacological in-
terventions for sleep promotion in the intensive care unit.
Cochrane Database Syst Rev 2015; 10: CD008808
44. Xie H, Kang J, Mills GH. Clinical review: the impact of
noise on patients’ sleep and the effectiveness of noise
reduction strategies in intensive care units. Crit Care 2009;
13: 208
45. Berg S. Impact of reduced reverberation time on sound-
induced arousals during sleep. Sleep 2001; 24: 289e92
46. Johansson L, Bergbom I, Waye KP, Ryherd E, Lindahl B.
The sound environment in an ICU patient roomda con-
tent analysis of sound levels and patient experiences.
Intensive Crit Care Nurs 2012; 28: 269e79
47. Buxton OM, Ellenbogen JM, Wang W, et al. Sleep disrup-
tion due to hospital noises: a prospective evaluation. Ann
Intern Med 2012; 157: 170e9
48. Elliott R, Rai T, McKinley S. Factors affecting sleep in the
critically ill: an observational study. J Crit Care 2014; 29:
859e63
49. Wallace CJ. The effect of earplugs on polysomnographic sleep
and description of sleep-disturbing factors in critically ill sub-
jects. PhD Dissertation. Salt Lake City, UT: University of
Utah; 1998
50. Topf M, Davis JE. Critical care unit noise and rapid eye
movement (REM) sleep. Heart Lung 1993; 22: 252e8
Handling editor: J.G. Hardman