Original
Article
&
&
&
&
&
&
&
&
&
&
&
&
&
&
Reducing Light and Sound in the Neonatal Intensive Care Unit:
An Evaluation of Patient Safety, Staff Satisfaction and Costs
Michele Walsh-Sukys, MD, MS
Ann Reitenbach, RN, BSN
Diane Hudson-Barr, RN, PhD
Patricia DePompei, RN, MSN
OBJECTIVES:
To modify an existing Level III neonatal intensive care unit and to compare
light and sound levels in the renovated nursery with an adjacent
traditionally configured nursery. Further, to assess the impact of this practice
on patient safety, staff perceptions of the nursery environments, and to
document costs of renovation.
STUDY DESIGN:
Prospective comparison of light and sound levels in identical six - bed
patient rooms within an existing intensive care unit. One room was
modified to reduce light and sound, and the other served as a control. Costs
of renovation were documented. Patient characteristics, severity of illness
and safety outcomes were documented following modifications. Physician
and nursing staff were surveyed on their perceptions of the renovations.
RESULTS:
Both light and sound were reduced with modifications that incurred modest
costs. Patient safety was not influenced adversely by reduced light or sound
levels. Staff members were highly satisfied with reductions in sound levels.
Reactions to reduced lighting levels were more mixed and led to
modification of bedside lighting.
CONCLUSIONS:
Cost - effective renovations to an existing NICU are possible, desirable, and
do not impact patient safety. The reductions achieved, however, are less than
those reached with new construction.
Journal of Perinatology 2001; 21:230 – 235.
INTRODUCTION
The environments of most intensive care units (ICUs) have
traditionally been filled with invariant bright lighting, a high level of
This study was supported in part by a grant from the Rainbow Babies and Childrens Hospital
Board of Trustees, Cleveland, OH.
Address correspondence and reprint requests to Michele Walsh - Sukys, MD, MS, Rainbow Babies
and Childrens Hospital, 11100 Euclid Avenue, Cleveland, OH 44106 - 6010.
noise and equipment alarms. In addition, the hard surfaces favored
in these environments because of the ease of disinfection may
augment the impact of sound. It is well documented that adults
cared for in ICUs may suffer a disorientating state that has been
termed ‘‘ICU psychosis,’’ which is in part attributed to sleep
deprivation and an environment of constant light and noise that
lacks night/day cues. A similar phenomenon has been described in
children in Pediatric Intensive Care Units.1 – 5
The impact of light and sound in the ICU environment affects
staff members, as well as patients and their families, and may
contribute to elevated stress levels independent of the stress
produced by caring for critically ill patients.6 Stressful working
conditions may contribute to ‘‘burnout’’ and negatively impact
staff retention.
Recently, attention has focused on the newborn intensive care
unit (NICU), and recommendations have been made to limit
both light and sound levels in the NICU.7 It is possible that the
impact of high light and sound levels may be more detrimental
and more long-lasting in the developing human than what has
been described in the adult. Thus, there has been a great deal of
interest in modifying existing NICUs to implement the newer
recommendations. There is also concern that limiting light and
sound levels could impair the ability of staff to assess critically
ill patients, could alter staff alertness, and ultimately could
reduce patient safety in the NICU. Finally, the fiscal environment
faced by health care providers in the 1990s has constrained
resources that might be used for renovation projects. This has
created a need for information on the costs and relative
effectiveness of different modifications to guide decisions during
renovation.
We thus sought to modify an existing Level III NICU and to
compare light and sound levels in the renovated nursery with
an adjacent traditionally configured nursery. Further, we
assessed the impact of this practice on patient safety, staff
perceptions of the nursery environments, and documented costs
of renovation.
METHODS
Existing NICU Environment
The study was performed in an existing Level III NICU at Rainbow
Babies and Childrens Hospital, University Hospitals of Cleveland, OH,
which was 10 years old at the time of study inception. The ICU is a
Journal of Perinatology 2001; 21:230 – 235
# 2001 Nature Publishing Group All rights reserved. 0743-8346/01 $17
230
www.nature.com/jp
Reducing Light and Sound in the NICU
38-bed unit that admits 1100 patients annually. Patients are
transferred to a 30-bed intermediate care area when physiologically
stable and when they weigh at least 1500 g.
Walsh-Sukys et al.
The NICU is divided into seven separated rooms; five rooms
house six patients each, and two rooms house three patients each
(Figure 1). There are also two single-bed isolation rooms. Each
Figure 1. Continuous sound monitoring over a 12 -hour epoch in the control nursery ( top panel ) and modified nursery (bottom panel ) . Sound was
significantly reduced in the modified nursery.
Journal of Perinatology 2001; 21:230 – 235
231
Walsh-Sukys et al.
of the six-bed rooms are configured identically. Each bed spot has
100 ft2 of care space. Lighting is provided with three fluorescent
units that span each six-bed nursery each unit and is controlled
by a single switch that allow the choice of a combination of
indirect and direct illumination, but no ability to light a bed spot
individually. Thus, all patients in that nursery receive the same
lighting regardless of their needs. Supplemental fluorescent task
lighting exists at the countertops that surround the wall of the
nursery. Two of the six-bed nurseries have no natural light, and
three have natural light through a large expanse of windows.
The floor in each nursery consists of cushioned vinyl sheeting laid
over cement, with scrubbable painted walls constructed of wallboard.
Laminated countertops and cupboards are on the periphery of each
patient room. Babies are admitted directly to a patient room and
remain there for the duration of their stay. The babies are initially
cared for under radiant warmers (Ohmeda Medical, Tewksbury, MA)
and transferred within a few hours to incubators (Ohmeda Medical).
Acoustic ceiling tile (Armstrong World Industries, Lancaster, PA;
noise reduction coefficient of 0.65) was the only sound-modulating
feature present in the NICU at the start of the study.
Nursery Modifications
Modifications to light and sound were made over about a 6-month
period in 1997. We allowed a 6-month period of adjustment before
measuring the impacts of light and sound modifications. As a
component of an ongoing program of developmental care, nursing
staff in both the modified and conventional nurseries had been
educated on the potential impacts of light and sound in the NICU on
neonates. No specific educational interventions were conducted as a
component of this study. Nursing and medical staff were obviously
aware of the environmental modifications, but were masked to the
outcomes measured. Because this study covered renovations to a fully
operational NICU it was not possible to obtain sound and light
measurements in empty nurseries before modification.
Sound modifications. The two six-patient rooms without any
windows or natural light were chosen for study. One six-patient
room was designated control and continued care with existing NICU
routines. The second six-patient room was designated as the
experimental nursery and modifications were made in a stepwise
manner. A series of sound modifications of increasing cost were
made: placement of weather stripping on all doors and drawer fronts,
replacement of all metal trash cans with rubber cans, placement of
covers over incubators, installation of carpet along the center of the
nursery (carpet squares with finished pile, 28 oz/yd2; Shaw
Stratton), and installation of sound-absorbing acoustic material
(Silent Wall Panel, 3/4 in., MPC, Westlake, OH; noise reduction
coefficient 0.75) in all monitor bays and sofitts (approximately 33%
of wall surfaces). The carpet was placed only in the center of the
nursery because of concerns that spills of blood and intravenous
fluids would be problematic. The carpet covered 145 ft2 of 522 ft2 of
floor space in the room or 28% of floor space.
232
Reducing Light and Sound in the NICU
Light modifications. To limit the renovation costs, existing
fluorescent lighting was left in place. The existing fluorescent
lighting was generally turned off, but at the staff’s discretion could be
turned on if an individual patient was critically ill such as in a
resuscitation. Three individual halogen spotlights were installed over
the bed space of each neonate in the modified nursery; these were
recommended by a lighting contractor to best meet the needs of our
unit. Each lighting unit had a variable intensity rheostat permitting
individualized lighting for each patient.
Measurements of light and sound. Light measurements were
made with a hand-held light meter (Extech foot candle/lux meter)
on a single day with measurements in both patient rooms taken
within 30 minutes of each other. Because neither nursery had
windows existing daylight was not a factor. The light meter was
placed at the midpoint of the incubator located in the center of each
patient room. Measurements were made on the top of and inside the
incubator at the level of the patient’s head. In the modified nursery,
measurements were made with and without the existing overhead
fluorescent lights illuminated.
Sound measurements were performed by a firm specializing in
industrial compliance (EA Group, Mentor, OH). In each room
sound measurements were continuously recorded to a computer disk
on two different occasions: one 24-hour period (7 AM to 7 AM )
(Larson-Davis Model 705 meter), and one 12-hour period (7 AM to
7 PM ) (Quest Model 2900 sound meter). The meters and dosimeters
were set to capture data in the ‘‘A’’ weighted scale with a 5-dB
doubling rate and slow response. The devices recorded a single
exposure level for each 60-second epoch in the study interval. These
data were summarized for statistical analysis with a single reading
abstracted at consecutive 10-minute intervals during the 12-hour
monitoring period. The microphones were placed on an
inconspicuous shelf above the center incubator in each six-patient
room. Each nursery was fully occupied at the time of the
Table 1 Content of Staff Satisfaction Questionnaire
1. The sound level in the modified nursery is the same as in the other
nurseries.
2. The light level in the modified nursery is the same as in the other
nurseries
3. The sound modifications to the modified nursery have improved the
care of babies.
4. The light modifications in the modified nursery have improved the
care of babies.
5. The lighting changes in the modified nursery are: ( circle one ) Much
too dark, too dark, about right, too light, much too light.
6. I believe that the light in modified nursery is adequate for doctors and
nurses to properly assess babies and perform procedures.
7. The sound modifications in modified nursery create a better working
environment for staff members.
8. The light modifications in modified nursery create a better working
environment for staff members.
9. I would like to see the following additional changes made:
10. Additional comments:
Journal of Perinatology 2001; 21:230 – 235
Reducing Light and Sound in the NICU
Walsh-Sukys et al.
Table 2 Patient Characteristics in Modified and Control Nurseries
Modified nursery ( n = 62 )
Control nursery ( n = 64 )
Significance ( p value )
2350 ± 1079
34.3 ± 5.4
16 ( 25% )
42 ( 65% )
36 ( 58% )
2390 ± 1101
34.7 ± 5.3
12 ( 19% )
36 ( 56% )
35 ( 55% )
0.84
0.69
0.15
0.07
0.56
Birthweight ( g )
Gestational age ( wk )
Number < 1500 g birthweight ( n, % )
Race ( White, n, % )
Gender ( male, n, % )
measurements (with three patients in each room on conventional
ventilators (Infant Star; Nellcor-Puritan-Bennett, Pleasanton, CA).
We attempted to minimize the obtrusiveness of the recording
devices by using small equipment and placing them in an
inconspicuous location. However, staff may have been aware that
sound recordings were in process. We attempted to minimize the
impact of this in several ways: by conducting ‘‘sham’’ recordings
where the monitoring unit was in place but not turned on before the
day of the actual sound recording, repeating recorded measurements,
and conducting prolonged measurements over several shifts.
Patient safety. Safety was assessed over a 6-month period after the
modifications by monitoring the following parameter within each
study room: medication errors, intravenous infiltrates, unanticipated
extubations, nosocomial infections, and mortality. Medication errors
were tracked using an ongoing monitoring system in place at
Rainbow Babies and Childrens Hospital, which utilized individualized incident review by senior NICU leadership and an independent
review by an experienced Pediatric Nurse. Each event was scored for
severity using a tool developed at Oakland Childrens Hospital that
assessed type of error, route of administration, classification of drug,
number of repeated events, and patient impact.8 The event was
classified into four degrees of severity: no impact, significant, serious,
or patient morbidity. These were adjusted for the number of patient
days per room. Intravenous infiltrates and accidental extubations
were tracked using hospital incident reports. The infection control
team reported nosocomial infections using definitions from the
Neonatal Nosocomial Infection Surveillance program of the Centers
for Disease Control. Mortality was ascertained from records of the
NICU’s mortality and morbidity committee.
Staff perceptions. All neonatologists (n=15), neonatal fellows
(n=6), neonatal nurse practitioners (n=14), and NICU nurses
(n=34) with experience in either the control or experimental
nursery were surveyed anonymously regarding their perceptions of
the light and sound modifications. The survey tool consisted of nine
declarative statements with a five-part Likert response scale ranging
from ‘‘strongly disagree’’ to ‘‘strongly agree’’ (Table 1). Two
additional open-ended questions solicited comments regarding
changes the staff would recommend to the modifications. The
questionnaire was distributed to all staff members, with a reminder 1
month later.
Statistical analyses. Light and safety data are expressed as the
mean±1 SD. Data were analyzed with Student’s paired t test for
nominal data and chi-square test for categorical data. Sound data
are measured on a logarithmic scale and therefore are expressed as
the median±1 SD. Differences in sound data were analyzed using
the Median test.9 The level of significance was set at p<0.05.
RESULTS
Patient Characteristics
Demographics of the study population are summarized in Table 2.
There were no statistically significant differences between the
populations in the modified and control nurseries.
Table 3 Costs of Modifications to Reduce Light and Sound To Nursery for Six Patients
Item
Principal intervention type
Sound
Weatherstrip on cupboard doors and drawers
Change trash cans
Carpeting
Incubator covers
Acoustical tile
Individually controlled lighting
Total
Cost ( $ ) *
Light
B
B
B
B
B
B
200
40
715
550
2500
3200
7205
*Costs were incurred in 1997 and include both labor and materials.
Journal of Perinatology 2001; 21:230 – 235
233
Walsh-Sukys et al.
Reducing Light and Sound in the NICU
Table 4 Comparison of Potential Markers of Impact of Reduced Light and Sound*
Patient daysy
Medication errors — all
Medication errors — serious or greater
Intravenous infiltrates
Accidental extubations
Nosocomial infection
Mortality
Control nursery
Modified nursery
Significance
905
4.47
2.20
5.52
5.52
3.28
4.47
891
3.14
0
1.12
6.73
4.62
4.86
0.63
0.15
0.08
0.71
0.64
0.90
*All values are expressed as events per 1000 patient days.
yData were compared between the two nurseries for a 6 - month period after the modifications were made.
Modification of Sound
The interventions applied led to lower sound levels in the modified
nursery compared to the control nursery in LEQ [64.2±1.70 vs.
71.8±0.43 (median±SD) (p<0.001) ], LMAX [65.7±4.64 vs.
72±2.68 (median±SD) (p<0.001) ], and L10 [64.5±71.8 dB,
(p<0.01) ] (Figure 1). Of interest, there was no statistically
significant variation in sound levels within each room over the 24hour period of monitoring (e.g., day shift did not differ from evening
or night shift). Over the entire 36-hour period of continuous
monitoring, we were startled to find at least one monitor or
equipment alarm in every single 10-minute observation epoch, a
total of 216 monitored epochs.
The modifications made to an existing nursery environment
incurred relatively modest costs (Table 3). The total cost of
acoustical interventions was $4005 per six patient room.
Modification of light. Installation of the track lighting provided
individualized, variable lighting to each of the neonates in the
modified nursery. Light levels measured with the existing fluorescent
lighting in the modified nursery were 193 lx, and with the modified
lighting were 12 lx. The individual patient lighting was able to
provide comparable illumination levels to the traditional fluorescent
lights (202 vs. 193 lx) when placed on high output.
The costs incurred for renovations to lighting were modest. In this
study the costs were limited because existing lighting was left in place
but used only during emergency situations. The lighting
modifications constituted the bulk of the cost of the project.
Nursing and physician perception of altered environment.
Sound- Forty-four surveys from 69 eligible staff members were
returned yielding an overall response rate of 64%. The overall sound
level in the modified nursery was judged by 95% of staff members
responding to be quieter than the control nursery. Sixty-six percent
judged the change to be more conducive to the care of babies, and
86% felt the changes produced a better environment for caretakers. No
staff members felt the sound changes had a negative impact on care.
Staff perceptions of the light in the modified environment
were mixed. Ninety-one percent of respondents felt that the light was
Light-
234
reduced in the modified environment. A majority of staff (59%) felt
that the care of babies had overall been improved and 61% felt that
the environment was better for caretakers. However, some members
(35%) felt that although the overall level of illumination was
appropriate, additional light was needed for placement of intravenous
catheters and for other procedures.
Patient and staff safety. There was no difference in the frequency
of potential markers of a negative impact of a reduced-light
environment: medication errors, infiltrates of intravenous lines,
unanticipated extubations, nosocomial infections, and mortality
(Table 4). In fact, the incidence of intravenous infiltrates was less in
the modified nursery compared to the control nursery, although this
did not achieve statistical significance (5.52 vs. 1.12 per 1000 patient
days; p=0.08).
No staff member reported an instance where they felt that their
recognition of an ill infant was impaired or delayed.
DISCUSSION
This study has demonstrated that reductions in light and sound can
be made in an existing NICU environment for relatively modest costs.
Further, the work indicates that light and sound can be reduced
without impacting patient safety. Both nursing and medical staff
appreciated the quieter environment. Although the reductions in
overall sound levels at first glance appear modest, it must be kept in
mind that the decibel scale is a logarithmic scale. Thus, a 5-Db
change represents a 50% reduction in noise level. When judged by
this standard the noise reductions are substantial. Similarly, the
impact of individual lighting created a substantial change in
environment from an ambient lighting level of nearly 200 lx to a
dim ambient lighting level of 15 lx.
The exact level of lighting that is appropriate for an ICU
environment is controversial. Current standards recognize the
conflicting needs of developing premature infants and working
nurses and physicians.10 In the past it has been recommended that
relatively high light levels of 6 to 10 lx be provided to allow
evaluation of an infant’s skin color and perfusion anywhere in the
ICU. More recently concern has arisen about the impact of light on
Journal of Perinatology 2001; 21:230 – 235
Reducing Light and Sound in the NICU
the developing retina, and thus ambient lighting levels have been
reduced.11 – 13 The most recent Illuminating Engineering Society
recommendations suggest ambient light levels of 1 to 2 lx (10 to 20
fc), with individual rheostat controls providing temporary increases
in illumination for assessments and procedures.14 Additional task
lighting that is directed away from the infants allows staff members
the levels of illumination needed for writing, medication calculation,
and procedural tasks. To our knowledge, the data in this study are the
first that demonstrate that the increasingly common practice of a
reduced-light environment can be provided without compromising
patient safety.
In contrast to the controversy over optimal NICU lighting
levels, there is widespread agreement that the majority of NICUs,
and in fact all ICUs, are too loud.15,16 The modifications made in
this study substantially reduced sound exposure (from 72 to 64
Db), but still failed to achieve recommended levels.7 The
modifications in this study were directed at ambient noise in the
ICU; no specific modifications were made to monitor alarms.
Modification of existing equipment is difficult to do. However,
when new equipment is purchased, it is imperative that the
equipment have alarms that are limited in their intensity. Indeed,
newer equipment has the ability to use visual and vibratory alerts
(flashing lights, or radiofrequency communication with vibrating
pagers worn by the nursing staff) for alarms of less severity, while
switching to auditory alarms only for the most serious events:
sustained bradycardia, asystole.
The loud ICU environment may be detrimental not only to the
infants but also to staff members. Loud environments stimulate the
autonomic nervous system and elevate cortisol levels.17 Exposure to
these levels may contribute to staff ‘‘burnout’’ and staff turnover. We
speculate that reduction of light and sound levels may improve
retention of skilled ICU nurses.
After the staff survey results were tabulated, we adjusted the
individual patient lighting conditions in the modified nursery to
increase the intensity of light available for procedures and to reduce
shadows. In addition, we addressed emerging research on the
importance of cyclical lighting to the emergence of circadian
rhythms by instituting a formal policy requiring a minimum of 8
hours of ambient lighting at a level of 100 to 150 lx with reduced
lighting in the evening and night.18
Our study demonstrates that light and sound can be modified
in an existing NICU at modest cost, without impacting patient
safety. Although the modifications made are a start, additional
improvements are possible. Such improvements include: ongoing
staff training to reduce loud conversation, and to reduce optional
noise inputs such as the use of radios for staff enjoyment. A
largely unexplored opportunity may exist to modify monitor and
equipment alarms in ICU equipment. Such changes can only be
made at the manufacturing level. Purchasers of ICU monitors and
equipment can demand these changes, and reward manufacturers
Journal of Perinatology 2001; 21:230 – 235
Walsh-Sukys et al.
who are responsive in their designs. In this way, ICU environments
can be modified to improve the health of our patients, and
ourselves.
Acknowledgments
We acknowledge the invaluable contributions of the NICU nurses and physicians and
the parents of the babies to this study. In addition we thank Jodi Barnum, RN, BSN,
and Maria Humphreys, RN, MSN, for assistance with data collection, and Sharon
Hall for assistance with identifying material specifications. We also thank Cynthia
Bearer, Rick Rodriguez, and Avroy Fanaroff for thoughtful comments on the
manuscript.
References
1. Sveinsson I. Postoperative psychosis after heart surgery. J Thorac Cardiovasc
Surg 1975;70:757 – 62.
2. Hansell H. The behavioral effects of noise on man: the patient with ‘‘intensive
care unit psychosis.’’ Heart Lung 1984;13:59 – 65.
3. Gelling L. Causes of ICU psychosis: the environmental factors. Nurs Crit Care
1999;4:22 – 6.
4. Hughes J. Hallucinations following cardiac surgery in a pediatric intensive
care unit. Intensive Crit Care 1994;10:209 – 11.
5. Simini B. Patients’ perceptions of intensive care. Lancet 1999;354:571 – 72.
6. Oates R, Oates P. Stress and mental health in neonatal intensive care units.
Arch J Child 1995;72:F107 – 10.
7. Committee on Environmental Health, American Academy of Pediatrics. Noise:
A hazard for the fetus and newborn. Pediatrics 1997;100:724 – 7.
8. Lund C. Medication error tool: Oakland Children’s Hospital, Oakland, CA;
1995.
9. Daniel WW. The Median Test. Biostatistics: A Foundation for Analysis in the
Health Sciences. New York: Wiley; 1995. p. 583 – 6.
10. Anonymous. Recommended standards for newborn ICU design. Report of the
Fourth Consensus Conference on Newborn ICU Design, Clearwater Beach, FL;
1999.
11. Glass P, Avery GB, Subramanian KNS, Keys MP, Sostek AM, Friendly DS.
Effect of light in the hospital nursery on the incidence of retinopathy of
prematurity. N Engl J Med 1985;313:401 – 4.
12. Ackerman B, Sherwonit E, Williams J. Reduced incidental light exposure:
effect on the development of retinopathy of prematurity in low birth weight
infants. Pediatrics 1989;83:958 – 62.
13. Reynolds J, Hardy RJ, Kennedy KA, Spencer R, VanHeuven WAJ, Fielder AR.
Lack of efficacy of light reduction in preventing retinopathy of prematurity. N
Engl J Med 1998;338:1572 – 6.
14. Illuminating Engineering Society of North America. Lighting for Healthcare
Facilities, RP29. New York; 1995.
15. Long J, Lucey JF, Philip AGS. Noise and hypoxemia in the intensive care
nursery. Pediatrics 1980;65:143 – 5.
16. Elander G, Hellstrom G. Reduction of noise levels in intensive care units for
infants: evaluation of an intervention program. Heart Lung 1995;24:376 – 9.
17. Kam P, Kam KA, Thompson JF. Noise pollution in the anesthetic and
intensive care environments. JAMA 1994;263:3185 – 90.
18. Hao H, Rivkees SA. The biological clock of very premature primate infants is
responsive to light. Proc Natl Acad Sci USA 1999;96:2426 – 29.
235