Content uploaded by R. Zadeh
Author content
All content in this area was uploaded by R. Zadeh on Dec 22, 2014
Content may be subject to copyright.
36 WWW.HERDJOURNAL.COM © 2014 VENDOME G ROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
The Impact of Windows and Daylight
on Acute-Care Nurses’ Physiological,
Psychological, and Behavioral Health
Rana Sagha Zadeh, MArch, PhD; Mardelle McCuskey Shepley, BA, MArch, MA, DArch;
Gary Williams, MSN, RN; and Susan Sung Eun Chung, PhD(c)
OBJECTIVE: To investigate the physiological and psychological effects of
windows and daylight on registered nurses.
BACKGROUND: To date, evidence has indicated that appropriate envi-
ronmental lighting with characteristics similar to natural light can improve
mood, alertness, and per formance. The restorative effects of windows
also have been documented. Hospital workspaces generally lack win-
dows and daylight, and the impact of the lack of windows and daylight on
healthcare employees’ well being has not been thoroughly investigated.
METHODS: Data were collected using multiple methods with a quasi-ex-
perimental approach (i.e., biological measurements, behavioral map-
ping, and analysis of archival data) in an acute-care nursing unit with two
wards that have similar environmental and organizational conditions, and
similar patient populations and acuity, but different availability of windows
in the nursing stations.
RESU LTS : Findings indicated that blood pressure (p < 0.0001)
decreased and body temperature increased (p = 0.03). Blood oxygen
saturation increased (p = 0.02), but the dif ference was clinically
insignicant. Communication (p < 0.0001) and laughter (p = 0.03)
both increased, and the subsidiary behavior indicators of sleepiness and
deteriorated mood (p = 0.02) decreased. Heart rate (p = 0.07 ), caffeine
intake (p = 0.3), self-reported sleepiness (p = 0.09), and the frequency
of medication errors (p = 0.14) also decreased, but insignicantly.
CONCLUSIONS: The ndings support evidence from laboratory and eld
settings of the benets of windows and daylight. A possible micro-restor-
ative effect of windows and daylight may result in lowered blood pres-
sure and increased oxygen saturation and a positive effect on circadian
rhythms (as suggested by body temperature) and morning sleepiness.
KEYWORDS: Critical care/intensive care, lighting, nursing, quality care,
work environment
ABSTRACT
AUTHOR AFFILIATIONS: Rana Sagha Zadeh is an Assistant Professor in the
Department of Design & Environmental Analysis at Cornell Universit y in Itha-
ca, New York. Mardelle McCuskey Shepley is a Professor and Director in the
Center for Health Systems & Design, College of Architecture, at Texas A&M
University in College Station, Texas. Gar y Williams is the Director of Telemetry
Services at St. Joseph Regional Health Center in Bryan, Texas. Susan Sung
Eun Chung is a PhD Candidate in the Department of Design & Environmental
Analysis, at C ornell University in Ithaca, New York.
CORRESPONDING AUTHOR: Rana Sagha Zadeh, Department of Design &
Environmental Analysis, Cornell University, 2425 Martha Van Rensselaer Hall,
Ithaca, N Y, 14853, USA ; rzadeh@cornell.edu; (6 07) 255-1946.
ACKNOWLEDGMENTS: We would like to acknowledge the valuable con-
tributions of the following members of the thesis advisory committee and
department faculty: Dr. James Varni, Dr. Glen Mills, Dr. Charles Culp, Dr. Zoa
Rybkowski, Kirk Hamilton, Dr. Roger Ulrich, Dr. Susan Rodiek, and Dr. Xue-
mei Zhu at Texas A&M U niversity; Dr. Elisabeth Klerman at Harvard Medi-
cal School; and Frank Becker at Cornell University. We sincerely thank Laurie
Waggener and WH R, Architects for her contribution to the research method-
ology and Yilin S ong, research assistant. We are thankful for the priceless
support provided by the healthcare facilit y leaders and staff: Doris Redman,
Kathleen Krusie, Dr. Mark Montgomery, Patt y Waldron, Kirby Nichols, Linda
Kenyon, Theda A nderson, Renee Briles, Gary J. Rizzato, and David Hall. This
study was supported by the Center for Health Design Research Coalition’s New
Investigator Award 2011–2012.
PREFERRED CITATION: Zadeh, R. S., Shepley, M. M., Williams, G., & Chung,
S. S. E. (2014). The impact of windows and daylight on acute-care nurs-
es’ physiological, psychological, and behavioral health. Health Environments
Research & Design Journal, 7(4), 36 –62.
© 2014 VENDOME G ROUP L LC HEA LTH ENVI RONME NTS RE SEA RCH & DESIGN JO URNA L 37
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
The optimization of the physical environmental conditions for health-
care staff, especially nursing staff, may offer an opportunity to create
high-performing work environments by helping staff to stay alert and
productive. Environmental design that has restorative qualities and is support-
ive of sensitive tasks that demand focus helps caregivers work more effectively.
Conversely, in working environments that are dark, monotone, and institution-
al, with inadequate external stimuli to help caregivers achieve their natural state
of peak alertness and performance, caregivers have to struggle to stay wakeful
and productive.
In healthcare settings, research has documented the positive benefits of healing
environmental elements. Four major components—presence of nature, reduced
noise and reduced crowding, soft and cyclical lighting, and availability of
music—show benefits in healing environments, according to a literature review
on adults and children by Sherman, Varni, Ulrich, and Malcarne (2005). One
hospital study showed that certain physical design features (including natural
lighting, live music, sufficient airflow, optimized layout, and homelike interiors)
improved staff perception of the quality of their work life (Mroczek, Mikitarian,
Vieira, & Rotarius, 2005). Among all the design features studied, the availabili-
ty of apertures that bring in natural light triggered the highest positive response
among hospital employees.
Windows that provide daylight and views of natural surroundings are a salient
feature of the physical environment that promote occupant satisfaction and well
being, as evidenced in corporate office, manufacturing, and healthcare settings.
A study in a manufacturing company in southern Europe on 100 workers inves-
tigated the impact of windows, focusing on three components: sunlight, illumi-
nation, and views (Beale, Lawrence, Leather, & Pyrgas, 1998). e study found
that the employees’ intention to quit was significantly lowered if either sun-
light or nature views were available. In addition, sunlight had a significant effect
on job satisfaction, and nature views significantly lowered occupational stress.
Another study on 333 Dutch office workers indicated that window views had a
direct effect on reducing reported physical and psychological discomfort (Aries,
Veitch, & Newsham, 2010).
Several studies have pointed out the quantitative benefits of presence versus
absence of windows and daylight in healthcare settings, mainly on patients,
and studies have addressed overall benefits on healthcare staff. e document-
ed effects include decreased pain medication consumption and decreased pain
medication costs (Walch, Rabin, Day, Williams, Choi, & Kang, 2005), increased
antidepressant effects of prescribed medication (Benedetti, Colombo, Pontiggia,
Bernasconi, Florita, & Smeraldi, 2003), and reduced duration of hospitalization
in bipolar patients (Benedetti, Colombo, Barbini, Campori, & Smeraldi, 2001).
Windows that provide views of nature are known for their restorative effects.
For example, surgical patients assigned to rooms with a nature view had shorter
lengths of post-operative time in the hospital compared with patients with a win-
dow view of a brick wall and had fewer complaints about their care as recorded
by nurses (Ulrich, 1984).
38 WWW.HERDJOURNAL.COM © 2014 VENDOME G ROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Studies on medical staff indicate an improved perceived quality of the work
environment in association with windows, sunlight, or views. Nurses exposed
to exterior nature views have reported improved perceived alertness and reduced
acute stress, whereas nurses with no view or non-nature views have reported
deteriorated perceived alertness and increased acute stress (Pati, Harvey Jr., &
Barach, 2008). A study on 141 nurses in Turkey reported higher job satisfaction
and less occupational stress when exposed to daylight for more than 3 hours per
day (Alimoglu & Donmez, 2005)
Lighting, natural or artificial, not only helps with visual tasks (Aries, Veitch, &
Newsham, 1998) such as reading medication labels (Joseph, 2006), but it also
affects physiological, psychological, and behavioral functions, according to lab-
oratory research studies (e.g., Cajochen, 2007; Joseph, 2006).
In terms of biological factors, the intensity and the timing of light exposure
can alter circadian rhythms and potentially improve the synchronization of the
body clock, peak cognitive performance, and work activities in a process known
as circadian realignment (Roberts, 2010). By altering circadian rhythms, light
exposure can result in increased body temperature (Badia, Myers, Boecker, &
Culpepper, 1991), decreased blood pressure (Myers & Badia, 1993; Boyce, 1997),
and decreased heart rate (Smolders, de Kort, & Cluitmans, 2012).
In terms of psychologica l and behavioral factors, appropriate lighting can improve
alertness, performance, mood, and social interaction through neuro-hormonal
changes. Bright light exposure suppresses the secretion of melatonin—a hor-
mone that governs alertness and sleepiness. Lighting is actually the most import-
ant environmental stimulus for humans with regard to alertness and sleepiness
(Postolache & Oren, 2005; Crepeau, Bullough, Figueiro, Porter, & Rea, 2006;
Cajochen, 2007).
Field studies on office workers have demonstrated the positive effects of lighting
on alertness, performance, and mood. In a 2012 study, higher alertness and vital-
ity were reported in highly illuminated conditions (1000 lx, compared with 200
lx), and higher physiological arousal (heart rate) and faster response times were
found in a psychomotor vigilance task toward the end of the 1-hour exposure
experiment (Smolders, de Kort, & Cluitmans, 2012). Individual performance is
regulated by time awake, quality of light, prior sleep, and one’s biological clock
(Klerman, 2010, cited by National Space Biomedical Research Institute, 2010).
In addition to timing, good lighting quality (spectrum and intensity) has been
correlated with improved task performance (e.g., van Bommel & van den Beld,
2004; Joseph, 2006). Specifically, adequate lighting has been recognized as an
essential factor for preventing errors by healthcare staff in medication rooms
(Chaudhury, Mahmood, & Valente, 2009) and for potentially enhancing nurs-
ing care (Kamali & Abbas, 2012). Buchanan, Barker, Gibson, Jiang, and Pearson
(1991) empirically tested three illumination levels and found significantly fewer
errors in dispensing prescriptions in high-illumination environments (1500 lx)
compared with low-illumination environments (450 lx).
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONME NTS RE SEA RCH & DES IGN JO URNA L 39
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
Mood also can be improved by lighting (Scott, 2000). Mood and stimulation
have been linked to the intensity level and color temperature of daylight (Bege-
mann, van den Beld, & Tenner, 1997; van Bommel & van den Beld, 2004).
Social interaction and frequency of communication, which negatively correlate
with sleepiness (Kim et al., 2009), have been studied with respect to lighting.
Exposure to light therapy has also been associated with reduced social withdraw-
al and reduced consumption of caffeinated drinks among patients with seasonal
affective disorder (Kräuchi, Wirz-Justice, & Graw 1990). A study using real-time
recording of daylight exposure, mood, and social interactions over a 20-day peri-
od found less conflict, more agreeableness, and improved mood among mildly
seasonally depressed subjects compared with those exposed to bright light (aan
het Rot, Moskowitz, & Young, 2008).
Studies that have found effects of lighting on interaction and interpersonal con-
flicts, as measured by the level of communication, amount of communication,
number of conflicts, and performance appraisals in the workplace. A study on 72
female university students aged 18 to 25 showed that bright light increased the
amount of general and intimate communication among friends (Gifford, 1988).
In general, daylight has also been found to be a more effective form of environ-
mental lighting than electric lighting in boosting alertness and cognitive perfor-
mance (Münch, Linhart, Borisuit, Jaeggi, & Scartezzini, 2012). Natural light
provides high-intensity blue light (400–500 nm) in the morn-
ing. e wavelength smoothly transitions to orange-red light
(600–700 nm) at sundown with minimal blue light intensity
(Roberts, 2010). Human physiology has evolved by respond-
ing to this phenomenon. An effective light wavelength for a
circadian response is mainly between 460 and 500 nm (Gaddy
et al., 1993, as cited by Roberts, 2010). When indoor electric
environmental lighting is not appropriate for our body’s nat-
ural state, occupants will be trapped in “biological darkness”
(Stevens & Rea, 2001).
All of the above qualitative and quantitative evidence is derived from the fields
of human biology and environmental psychology. e research was conduct-
ed in the laboratory or field, mainly not on healthcare workers or hospitalized
patients, but it provides the background to objectively explore the restorative
physiological and psychological effects of availability of windows and daylight
on healthcare employees. Maximizing access to windows with views of nature
and daylight may be a low-cost and easy way to increase the health and perfor-
mance of nursing staff. We know that the performance and well being of health-
care providers is linked to patient satisfaction and perception of care (Rossberg,
Melle, Opjordsmoen, & Friis, 2004). is study addresses the gap in the liter-
ature on frontline caregivers’ work environment by objectively and subjectively
measuring the physiological and psychological impacts of windows and daylight,
as well as the effect of light on work performance (as measured by probability of
errors). Based on the literature, the following hypotheses were tested:
Maximizing access to windows
with views of nature and daylight
may be a low-cost and easy
way to increase the health and
performance of nursing staff.
40 WWW.HERDJOURNAL.COM © 2014 VENDOME G ROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Hypothe sis 1: The presence of windows and daylight will improve phys-
iological responses (e.g., decrease blood pressure and heart rate
and increase oxygen saturation and body temperature) by realigning
circadian rhythms.
Hypothe sis 2: The presence of windows and daylight will reduce sleep-
iness and improve mood, as evidenced by subsidiary behaviors that
represent deteriorated mood and sleepiness.
Hypothesis 3: The presence of windows and daylight will increase the
frequency of communication and social interaction.
Hypothesis 4: The presence of windows and daylight will improve per-
formance, as evidenced by reduced frequency of human-related
medication errors.
Methods
Data were collected using multiple methods with a quasi-experimental approach
(i.e., biological measurements, behavioral mapping, and analysis of archival data)
in an acute-care nursing unit with two wards that have similar environmental
and organizational conditions, and similar patient populations and acuity, but
different availability of windows in the nursing stations.
Setting
e study was carried out in two clusters of nurses’ stations, located in the north
and south wings of an 86-bed acute-nursing unit in a community hospital in
Texas. e nurses’ stations in the north ward have no access to daylight, where-
Figure 1. Windows and views in the north (lef t) and south (right) nurses’ stations.
© 2014 VENDOME G ROUP L LC HEA LTH ENV IRONM ENTS R ESEA RCH & DESIGN JO URNA L 41
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
as the nurses’ stations in the south ward have windows that face north and look
out on portions of the hospital building, the sky, and a courtyard (see Figure 1).
In terms of lighting quality, the windowed condition provides indirect daylight,
which is not available in the windowless condition. In the windowless condition,
T8 fluorescent ceiling-mounted lamps are the only sources of light available. In
the windowed condition, the same electric lighting is available, in addition to
daylight from a north-facing window.
e wards are mirrored in layout, with the sa me colors, finish materials, furniture,
and equipment. Both wards are operated under the same management: one unit
manager, one team leader, and charge nurses who were responsible for different
shifts. Patient rooms are located on both sides of the corridors all along these
two wings (double-loaded corridors, see Figure 2). Although abundant daylight
was available through windows, no direct sunlight was available during the study
period (summer months) due to the orientation of the windows, site obstructions
surrounding the courtyard, and sun angle. Patients in this cardiac unit are
continuously monitored using telemetry technology, and they require nursing
care with technical skills beyond those of a basic medical/surgical nurse but do
not require critical-care nursing. e patient population is similar across the
units, with comparable acuity levels (Meanwindowless = 3.2 ± 0.06, Meanwindowed
= 3.3 ± 0.07, p = 0.16) according to the standardized Patient Classification
System for the hospital. Nurses rotate between the two stations.
Participants
Participants were selected from among day-shift registered nurses (RNs) in the
hospital who rotate weekly between the north and south in-patient acute-care
wards. Initially, 20 RNs were found to be potentially eligible for the study, all of
whom agreed to participate; however, after changes in patient census and work
schedules, 12 nurses remained eligible and were enrolled, all of whom completed
the study. e participants, summarized in Table 1, did not know the variable
Figure 2. Facility oor plan.
Dark gray indicates nurses’ stations; dotted line, the patient rooms accessed by the shadowed RNs
(as adjusted by charge nurses); light gray, the patient rooms.
42 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROUP LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
of interest. is avoided any placebo effects as suggested by Ott and Longnecker
(2008). e study was approved by the Institutional Review Board and Office of
Research Compliance Staff of the Division of Research at the researchers’ uni-
versity, as well as the Institutional Review Board at the hospital.
Procedure
e physiological and psychological effects of the presence of windows and day-
light on RNs were assessed in a quasi-experimental study using a non-equiva-
lent-groups crossover design that caref ully controlled for personal, environmental,
and organizational variables. More specifically, the same participants were stud-
ied under both conditions, with consistent work patterns and protocols. To con-
trol for work fatigue, only participants scheduled to work 2 days in a row each
week were eligible, and observations were only on the second day of their shift.
Environmental settings were controlled by conducting the study in areas with
similar unit layouts, equipment, furniture, and finishes. Work protocol man-
agement, computer and charting systems, patient type and acuity, and nurse/
patient ratios were similar in both conditions. Each participant was studied for
a total of 16 hours through a shadowing procedure (8 hours per day in each
treatment condition for 2 days). is procedure was introduced to the partici-
pants in the morning prior to the study, and participants maintained their daily
activities while being shadowed by the researcher. ey were given the option
to ask the researcher to stop the observation process at any time for any reason,
including the need for privacy or other needs that might benefit the participants
or patients.
e order of windowed and windowless locations was randomly assigned by the
charge nurses. Each observation day was dedicated to one nurse participant.
Two observers (the principal investigator and a research assistant) conducted
the onsite data collection after a set of pilot studies and inter-rater reliability
assessments. Shadowing resulted in a total of 144 hours of behavioral mapping
(across the 12 RNs) on work-related and subsidiary behavior related to sleepiness
and mood. In addition, between the two observers, approximately 16 hours of
Table 1. Participant Demographics
Frequency Percentage
Gender Female 10 83%
Male 2 17%
ToTal 12 100%
Age 20–29 4 33%
30–39 1 8%
40– 49 4 33%
50– 59 3 25%
ToTal 12 100%
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONME NTS RE SEA RCH & DES IGN JOURNA L 43
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
pilot testing and reliability analysis were completed for light measurement and
behavior observation. A total of 120 biological measurements were recorded.
During the data recording, biological assessments (blood pressure, heart rate,
temperature, and oxygen saturation) and subjective momentary sleepiness/alert-
ness assessment were conducted at discrete hourly intervals. Behavioral cues were
recorded in real time, and illumination levels were recorded every 5 minutes and
averaged bi-hourly for analysis purposes.
In addition, archived data on human-related medication errors over a 3-year peri-
od (with more than 25,200 admissions) were collected for both nurse stations.
Study Design
e methods used to test the hypotheses in this study included physiological
assessments (vital signs), behavioral mapping, momentary assessments, and
records analysis. In addition, a digital light meter was used to measure horizon-
tal illumination levels every 5 minutes during the observation and behavioral
mapping period for each participant. e illumination measurement plane was
set to remain at a constant distance from the participants’ eyes.
Because each participant was repeatedly measured over time for the physiolog-
ical assessments (vital signs), behavioral mapping, and momentary assessments,
a mixed model was used to analyze the data. e dependent variables of inter-
est were repeatedly measured for each subject. Bi-hourly measurements for all
the response variables were paired for each individual in both conditions, and
the patterns of the average daily measurements were analyzed in a mixed model
to test the hypothesis of whether physiological psychological and behavioral
responses were improved in the windowed condition compared with the win-
dowless condition (one-directional hypothesis). e model was consistently test-
ed for the treatment (i.e., the presence or absence of windows and daylight), time,
gender, age, sleep duration, number of patients, average total light levels (lx), and
daylight factor (indicating distance from window) as main effects, as well as for
order of data collection and the interaction terms between time with treatment,
average light, daylight factor, and average patient acuity as fixed variables. e
final model included only the significant effects.
To account for the non-independence of the measurements, a subject ID was
entered in the model as a random effect. e main independent variable of inter-
est (presence or absence of window) was measured for each subject as each sub-
ject experienced both conditions. Other independent variables for individual
subjects were time, sleep duration, number of patients, average total light levels,
and daylight factor. Independent variables between subjects were gender and age.
A non-parametric paired test was used to compare the frequency of human-re-
lated intravenous (IV) and non-IV medication errors in the windowed ward
compared with the similar windowless ward while RNs were rotating frequently
between the two.
44 WWW.HERDJOURNAL.COM © 2014 VEND OME GROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Measures
Data were collected using multiple methods with a quasi-experimental approach
(i.e., biological measurements, behavioral mapping, and analysis of archival data)
in an acute-care nursing unit with two wards that have similar environmental
and organizational conditions, and similar patient populations and acuity, but
different availability of windows in the nursing stations. Measures represent-
ing the dependent variables include physiological, psychological, behavioral,
and momentary ecological assessments. In addition to the documentation of
the presence and absence of windows, the participants’ total light exposure was
measured.
Physiological Assessments
To measure physiological responses, the RNs’ vital signs—blood pressure, heart
rate, oxygen saturation level, and body temperature—were measured bi-hourly
for every participant under both conditions using a Carescape™ V100 monitor
from GE Healthcare (Fairfield, CT). e participants were seated during the
measurements. To represent changes in blood pressure, systolic and diastolic
blood pressure were combined to calculate mean arterial pressure (MAP) (Rog-
ers et al., 2001).
Psychological and Behavioral Assessments
Frequency of communication and positive social interaction (measured by
laughter), caffeine intake, illumination levels, and subsidiary behaviors related
to coping with sleepiness were recorded every 5 minutes. Because the study was
conducted in a similar fashion under the controlled and treated conditions, par-
ticipants did not know that windows and daylight were the study’s specific vari-
ables of interest, nor which variables and behavioral clues were recorded. us,
the placebo effect, an unwanted favorable response to a condition, was avoided.
e measurements were averaged bi-hourly and paired per participant and time
of day for comparison.
Behavior observations from a total of 192 hours (24 days) were collected. After
the data collection, 120 minutes of data from the beginning and end of the data
collection period were eliminated because of frequent violation of the method-
ological consideration about deviation of actual measurement time from target
measurement time. is deviation was a result of RNs’ critical tasks at the begin-
ning and end of shifts, as the research team was not supposed to interrupt tasks
such as reporting, charting, medication administration, and communication to
physician and patient. An allowable 30-minute deviation from target time was set
in order to avoid interruption of critical tasks. Observations on the participants’
frequency of communication and positive social interaction, caffeine intake, illu-
mination levels received, location changes during the work routine, and subsid-
iary behaviors related to coping with sleepiness were recorded every 5 minutes
using Noldus Behavior Mapping software and equipment (Olsen, Hutching, &
Ehrenkrantz, 2000). A comparison of the measurements between the windowed
and windowless conditions was made to determine whether the presence of win-
dows and daylight increased communication and positive social interaction and
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONMENTS RE SEARCH & DES IGN JOURNA L 45
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
reduced sleepiness-related subsidiary behavior. Given the assumption that RNs
would be less sleepy and more alert in the windowed condition, we hypothesized
that participants would communicate more frequently in the windowed condi-
tion compared with the windowless condition. e frequency of communication
for each individual was compared between the windowed condition and win-
dowless condition, both when participants were directly in the nurses’ stations
and when they were in various other locations.
If the presence of daylight improves mood and communication, the behavior-
al cues related to these variables should be improved. ese subsidiary behav-
iors are psychological or physiological responses that are not directly related to
the work task, but rather are behavioral manifestations that can interfere with
tasks and lead to errors (Takanishi et al., 2010). ese include behaviors relat-
ed to monotony (stretching), sleepiness (yawning, sighing, and rubbing eyes),
and habit (touching head). Laughter was also documented, as it is considered a
manifestation of sociability, warmth, and approachability (Feagai, 2011; Palmer,
2005). e research team defined laughter as audible, chest moving, and occur-
ring out of happiness or pleasure, and it was recorded each time that a positive
conversation resulted in laughter.
If the presence of windows and daylight increases alertness, such improvement
should be reflected in the RNs’ related behavioral clues of sleepiness/alertness.
Subsidiary behaviors related to overall sleepiness and deteriorated mood (yawn-
ing, sighing, singing and whistling, stretching trunk, touching forehead, and
eye rubbing) were recorded via behavior observation. If alertness is improved
by windows and daylight, the frequency of caffeine intake may decrease. To test
this notion, the frequency of caffeine intake was monitored for each participant
under both conditions (while participants were unaware of the measurement).
To measure the intake of caffeinated beverages, the type of beverage in addition
to the number of sips were recorded.
Momentary Ecological Assessment
In addition to observations of subsidiary behaviors related to sleepiness, visual
momentary ecological assessment (Stone & Shiffman, 1994) was used by the
RNs to document their bi-hourly subjective sleepiness. e format of the sur-
vey was borrowed from PedsQL™ Visual Analogue Scales (Varni, Seid, & Rode,
1999) and enabled participants to input their selection by choosing a number
from 0 to 10 a maximum of five times daily. We hypothesized that the subjective
sleepiness would be reduced in the windowed nurses’ station compared with the
windowless one, as daylight and windows are likely to increase the participant’s
alertness.
Light Measurement
A digital light meter, model 401025 by EXTECH Instruments (Nashua, NH),
was used to measure horizontal illumination levels (light levels on the work sur-
face) every 5 minutes (or earlier if participants changed location).
46 WWW.HERDJOURNAL.COM © 2014 VEND OME GROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
e illumination measurement plane was set to remain at a consistent distance
from the participants’ eyes (average height). If the participants were seated, the
measurement plane would be at 30 inches (76.2 cm), which represents the height
of the workstation. If participants were standing, the measurement plane was set
at 40 inches (101.6 cm), a typical height for a standing counter (Waggener, per-
sonal communication, 2011).
Because it was not possible to use a tape measure at all times while shadowing the
participants, the observers created anatomical markers (using a tape measure) to
keep the measurement plane consistent and be able to rapidly measure lighting
at the correct height. e observers used a unified method to
measure lighting. e methods were practiced before the study
began. In these experiments, we ensured that the light mea-
surement would be taken in a way such that the reader was not
shadowed by an object or by observers’ or participants’ bodies.
erefore, the observers held the reader away from their bod-
ies. e light meter was held stationary for at least 2 seconds
(counting “one one-thousand, two one-thousand”) in a hori-
zontal position before reading the digits shown on the screen.
Records Analysis
Furthermore, records of IV and non-IV medication errors in all existing cate-
gories were studied for the windowed (design case) and windowless (reference
case) wards from January 1, 2009, to December 31, 2011 (including more than
20,000 patient admissions). is includes the 3 full years of medical records data
available to the hospital following a switch in the medical records system. e
medication errors included incidents with and without harm, both of which were
included in the analysis. A comparison was made of whether the windowed ward
had lower numbers of human-related medication errors than the windowless
ward. Following the notion that alertness would be improved in the windowed
nurses’ station, we hypothesized that the human-related medication errors would
be less frequent in the windowed ward compared with the similar windowless
ward while RNs were rotating frequently between the two.
Results
e positive impact of windows and daylight on RNs was tested using physi-
ological assessments (vital signs), behavioral mapping, and records analysis. In
terms of light intensity, the total average daily light exposure for RNs, based on
a repeated measurement taken every 5 minutes during the 24 days of shadowing,
was nearly the same in the windowless condition (765 ± 192) and the windowed
condition (672 ± 148), with no statistical difference (F = 0.59, p = 0.44). e out-
liers indicate the instances when the RNs left their unit to an area with natural
lighting (Figure 3). When assigned to the windowless station, RNs left the ward
for other areas more often than when they were working in the windowed sta-
tion. ese instances are plotted as outliers, as the destinations included a day-
light atrium or outdoors that provided considerably higher light levels.
The positive impact of windows
and daylight on RNs was tested
using physiological assessments
(vital signs), behavioral mapping,
and records analysis.
© 2014 VENDOME G ROUP L LC HEA LTH ENVIRONMENTS RESEA RCH & DE SIGN JO URNAL 47
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
Physiological Responses
If the presence of windows has restorative effects and the availability of daylight
improves neuro-hormonal mechanisms resulting in the adjustment of the body’s
circadian rhythms, the RNs’ biological assessment should indicate the improve-
ments. e RNs’ vital signs—blood pressure, heart rate, oxygen saturation level,
and body temperature—were measured bi-hourly, once in the windowless con-
dition (control) and once in the windowed condition (design). (Table 3, in the
“Discussion,” below, summarizes the findings.)
Blood Pressure
e findings from the mixed model analysis showed that MAP, a measurement
for blood pressure combining systolic and diastolic pressure, was significantly
lower (see Table 3) when the RNs worked at the windowed location (94.6 ± 3.0
versus 86.9 ± 3.0; estimate = 7.65, p < 0.0001).
e data indicated that age, gender, and order also had a significant effect on
MAP. e higher the RN’s age, the higher the MAP (p values reported in the
Table per age group). Female nurses had significantly lower MAP than did male
nurses (p = 0.046). No other variable had an effect on MAP.
Figure 3. Average light exposure for RNs during nursing care.
Windowless Windowed
Total Light Levels (Lx )
3000
2500
2000
1500
1000
500
0
★
48 WWW.HERDJOURNAL.COM © 2014 VEND OME GROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Heart Rate
Heart rate was not significantly reduced (p = 0.067) when the RNs worked in the
ward with windows and daylight (78.39 ± 2.8) compared with the windowless
ward (75.7 ± 2.8). Further study of the data showed that time has a significant
effect (p = 0.002) on heart rate. Heart rate increased over time; therefore, the
mean heart rate had an upward trend (see Figure 4). No other variables affected
heart rate.
Figure 4. Bi-hourly measurements of vital signs.
10:00 a.m. 12:00 p.m. 2:00 p.m
Average Oxygen
Saturation (%)
Average Heart Rate
(bpm)
Mean Ar terial Pr essure
(mmHg)
85
80
75
70
65
60
100
95
90
85
80
75
98.5
98
97. 5
97
96.5
98
97. 6
97. 2
96.8
96.4
Body temperature
(degF)
windowless
windowed
Key:
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONME NTS RE SEA RCH & DES IGN JO URNA L 49
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
Oxygen Saturation
Oxygen saturation was significantly higher (p = 0.016, estimate = –0.0371) when
the RNs worked in the windowed ward (97.02 ± 0.2) compared with the win-
dowless (97.39 ± 0.2) ward. Age was a significant predictor for oxygen saturation;
the higher the RNs’ age, the lower the oxygen saturation level (p-values are pre-
sented in the table per age group) as expected in human physiology. Male nurses
had lower oxygen saturation than female nurses. e graph showed a downward
trend in oxygen saturation over time (see Figure 4); however, time did not have
a significant effect. No other variable had any effect on oxygen saturation. It is
notable that the windows were fixed and not operable. erefore, fresh air was
not provided at the treatment condition, and both wards were ventilated with
the same mechanical system.
Body Temperature
Body temperature (a marker of body circadian rhythm), although still within the
normal range, was significantly higher (p = 0.026, estimate = –0.28) when RNs
worked in the ward with windows and daylight (97.58°F ± 0.16) compared with
the windowless ward (97.30°F ± 0.16). Further observations showed that, except
for the treatment presence and absence of windows, no other variable had any
effect on body temperature.
Psychological and Behavioral Responses
Psychological and behavioral responses recorded were commu-
nication and positive interaction, as well as indicators of de-
teriorated mood and sleepiness.
Communication and Positive Interaction
Communication was measured both based on quantity and
quality. Frequency of occurrence of communication was
recorded. In addition, positive social interaction (communica-
tion followed by laughter) was recorded.
Frequency of communication. Communication in the nursing stations by the
RNs significantly increased (p < 0.0001) when they worked in the ward with
windows and daylight (21.61 ± 1.5) as compared with the windowless ward
(13.11 ± 1.5). e presence of the window resulted in eight more occurrences
of communication per RN participant (parameter estimate of 10.26). No other
variable had an effect on communication.
Figure 5 displays the locations where communications occurred when the RNs
worked in the two wards. e vertical axis shows the total number of commu-
nications in each ward during the 192-hour shadowing of individual RNs in the
unit. e graph shows a major shift in the frequency of communication in the
nurses’ station, which is interestingly the location of the window (independent
variable). e total frequency of communication was similar in the reference case
and design case in all other areas; however, frequency of communication outside
Communication in the nursing
stations by the RNs signicantly
increased (p < 0.0001) when
they worked in the ward with
windows and daylight (21.61 ± 1.5)
as compared with the
windowless ward (13.11 ± 1.5).
50 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROUP LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Figure 5. Nurses’ communication.
Nurses’ S tation
Total Frequency of Occurence
Patient Room
Hallway
Supply
Medication
Nourishment
Soiled Roo m
Break Room
Off Uni t
windowless (reference case)
windowed (design case)
Key:
120 0
1000
800
600
400
200
0
Figure 6. Positive sociability and approachability.
Nurses’ S tation
Patient Room
Hallway
Supply Room
Medication, Soiled
Supply Room, Ofce,
Nourishment
Break Room,
Off Uni t
windowless (reference case)
windowed (design case)
Key:
Total Frequency of Occurence
140
120
100
80
60
40
20
0
© 2014 VENDOME G ROUP L LC HEA LTH ENVI RONME NTS RE SEA RCH & DESIGN JO URNA L 51
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
the assigned work unit (in the hospital entrance, adjacent work units, or break
areas) was higher in the windowless location than in the windowed condition. It
is notable that the break area for both wards is the same. e break area is win-
dowless, shared between the two wards, and located in the middle of the floor
plan at an equal distance from the two studied wards.
Frequency of laughter. e increase in the average occurrence of laughter in the
windowed nursing station was higher (4.67 ± 0.96) than the windowless station
(1.96 ± 0.96) significantly (p = 0.028, estimate = –2.71). No other variable had
an effect on the occurrence of laughter.
Figure 6 displays the total frequency of laughter during the
shadowing of individual RNs plotted by location. Consis-
tent with the findings in the previous section about commu-
nication, the occurrence of positive interactions resulting in
laughter was similar in different locations, except in the nurs-
es’ stations, where the presence of windows varied. e occur-
rence of positive conversation (e.g., a compliment or positive
humor) followed by laughter in patient rooms was quite similar
in both the windowed and windowless wards.
Sleepiness and Mood
Sleepiness and mood were assessed in three ways, observed behavioral indicators,
observed frequency of caffeine intake, and participants’ self-reported subjective
momentary assessment.
The occurrence of subsidiary
behavior indicators of sleepiness
and deteriorated mood in the
windowed condition was less
than in the windowless condition
(2.89 ± 0.96 vs. 5.25 ± 0.96,
estimate = 2.36).
Figure 7. Comparison of caffeinated and non-caffeinated liquid intake.
CAFFEINATED LIQUID INTAKE NONCAFFEINATED LIQUID INTAKE
Windowless Windowed
80
70
60
50
40
30
20
10
0
Frequency of Occurence
Windowless Windowed
20
19
14
12
11
10
9
8
6
5
3
1
Participant’s ID
52 WWW.HERDJOURNAL.COM © 2014 VENDOM E GROU P LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
Table 2. Mixed Model Analysis for Windowed and Windowless Conditions (measured bi-hourly at 8:00 a.m., 10:00 a.m., and 12:00 p.m.)
Mean Arterial Frequency of Frequency of Subjective Frequency of
Pressure Blood Oxygen Body Communication Laughter Momentary Subsidiary Behavior
Effec t (Blo od Press ure) Heart R ate Saturation Temperatu re (in Nurs e Station) (in N urse St ation) Sleepines s (Sleepin ess & Mood )
Estimate
Sig. Estimate Sig. Estimat e
Sig. Estimate Sig. E stimate
Sig. Estimate
Sig. Estimate
Sig. Estimate Sig.
(St. Error) (St. Error) (St. Error) (St. Error) (St. Error) (St. Error) (St. Error) (St. Error)
Intercept 109.76
0.000 79.07 0.000 95.29 0.000 97.6 4 0.000 19.58 0.000 4.85 0.001 2.33 0.001 3.25 0.002
(6.4) (2.97) (.41) (0.19) (1.97) (1.26) (0.53) (1.93)
Time 10:00 a.m. –6.13 0.001 – 0.77
0.0 19 1.7 1
(Ref 2 :00 p.m.) (1.67) (0.32) (1.56)
12:00 a.m. –4.00 0.13 4 –0.86
0.009 2 .12
(1.67) (0.32) (1.56)
Treatmen t Windowless 7.65 0.000 2.69 0.067 –0.37 0. 016 – 0.28 0.026 –8.5 0.000 –2.71 0.028 0.52 0.095 3.84 0.023
(Ref W indo Wed) (1.94) (1.66) (0.17) (0.14) (1.79) (1.19) (0.37) (1.71)
Gender Female –12.41
0.045 1.5 6 0.002
(Ref m ale) ( 5 .13 ) ( 0 .32 )
Age Group 20–29 –15.88
0.0 15 1.8 1 0.001
(Ref 5 0–59) (4.97 ) (0. 31)
30–39 –20.57
0.025 1.17 0.036
(7.25) (0.45)
40–49 –10.86
0.065 2.30 0.000
(4.97) (0.31)
Order level 1 –8.46 0.000
(Ref 2) (1.94 )
Sleep Duration
Number of Patients
Illumination
Daylight Factor
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONME NTS RE SEA RCH & DES IGN JO URNA L 53
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
Behavioral indicators of sleepiness and deteriorated mood. e findings
showed that the treatment had a significant direct effect on the occurrence of
subsidiary behavior (p = 0.023). e occurrence of subsidiary behavior indicators
of sleepiness and deteriorated mood in the windowed condition was less than in
the windowless condition (2.89 ± 0.96 vs. 5.25 ± 0.96, estimate = 2.36). Fur-
ther study of the data indicated that although the total occurrence of subsidiary
behavior per participant per day significantly decreased by 46% (t-test, p = 0.03),
the mixed model did not pick up a significant effect for the treatment variable
(presence or absence of windows).
Caffeine intake. Caffeine intake was not significantly reduced (p = 0.275) when
nurses worked in the windowed ward compared with the windowless ward. No
variable had an effect on caffeine intake.
Other observations of the data indicated that caffeine intake had an insignifi-
cant reduction of 10% in the windowed nurses’ station compared with the win-
dowless station; meanwhile, intake of decaffeinated drinks (water, ice, or fruit
juice) increased by 10%, again insignificantly (Figure 7), keeping the overall
liquid intake (for thirst) the same. Although the observation showed interesting
Figure 8. Bi-hourly subjective momentary alertness (top graph) and average total
illumination levels (bottom graph) in windowed and windowless conditions.
windowless
windowed
Key:
Average Subje ctive
Alertness
10
9.5
9.0
8.5
8.0
7.5
7.0
Absolute Clock Time
Average Illumination Level ( LUX)
10:00 a.m. 12:00 p.m. 2:00 p.m
900
800
700
600
500
400
300
200
54 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROUP LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
trends, the hypothesis regarding significant reduction of caffeine intake was not
supported.
Subjective momentary assessment of sleepiness. e observed reduction in
the bi-hourly ecological momentary assessment of sleepiness in RNs was not
significant (p = 0.094). Other observation of the data indicated that time of day
was a significant predictor of sleepiness (Table 2). e estimated means indicat-
ed that sleepiness consistently increased over time. Figure 8 displays subjective
alertness (subjective sleepiness subtracted from 10) under the windowed and
windowless conditions. is shows that the reduction in sleepiness and increase
in alertness provided by windows and daylight is significant in the morning and
the effect of the treatment is reduced over time.
Further study of the data showed that variability between subjects in reporting
sleepiness was very high, as indicated by a residual variability of 2.20. Figure
8 also displays the bi-hourly average total illumination levels that nurses were
exposed to during the day in the windowless and windowed conditions. e total
average illumination in the windowed condition was higher than that in the
windowless condition but had considerable variation during the day.
Medication Errors
In the past 3 years, a total of 23 medication errors were reported in the entire unit
(Figure 9). e yearly probabilities of error, adjusted per each patient room for
Figure 9. Non-IV and IV medication error rates between 2009 and 2011
(adjusted per patient room per year).
Probabili ty of (at lea st) One Medication Error per Room per Year
14%
12%
10%
8%
6%
4%
2%
0%
Wrong time
Wrong route
Wrong rate (equipment)
Wrong patient/chart
Wrong medication/unauthorized
Wrong frequency
Wrong dose: formulation/strength/amount
Wrong additive/solution
Wrong dose: quantity
Wrong dose: form
Wrong administration technique
Other
Extra dose
Windowless Windowed
22% reduction
(not signicant)
© 2014 VENDOME G ROUP L LC HEALTH ENVI RONME NTS RE SEA RCH & DES IGN JO URNA L 55
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
the windowless and windowed ward, were 11.6% and 9.0%, respectively, show-
ing a 22% lower probability of error in the windowed ward; however, further
analysis using a non-parametric paired test on the occurrence of IV and non-IV
medication errors in the windowed ward compared with the windowless ward
showed that the reduction was not significant (sum rank of 38 negative ranks to
17 positive ranks, z = –1.10, df = 13, p = 0.14).
Discussion
is study compared the physiological and behavioral responses of day-shift
RNs when providing direct patient care, as measured once in a windowed nurse
station and once in a windowless nurse station in an acute-care unit. Table 3
summarizes the findings. e findings showed that four out of five of all the
nurses’ vital signs improved significantly when they were stationed in the nurses’
station with windows and daylight. Overall, daily MAP decreased significant-
ly. Both oxygen saturation and body temperature increased significantly. e
amount of increase in oxygen saturation is clinically trivial (less than one unit),
but it still hints at a possibility of improvement in nurses. Heart rate showed an
insignificant reduction.
Physiological Outcomes
e lowered blood pressure may be explained by availability of daylight, which,
according to Walch et al. (2005), has stress-reducing effects. Another important
element may be the presence of windows and an outside view, which, according
to Ulrich (1984), has healing effects. A pleasant window view has a “micro-re-
storative” effect, providing the opportunity of “brief respite to one’s directed
attention” (Kaplan, 1993, p. 196).
Table 3. Summary of Findings
HYPOTHESIS FINDINGS
1: The presence of windows and daylight will improve
physiological responses (decrease blood pressure and
heart rate and increase oxygen saturation and body
temperature) by realigning circadian rhythms.
2: The presence of windows and daylight will reduce
sleepiness and improve mood, as evidenced by
subsidiary behaviors that represent deteriorated mood
and sleepiness.
3: T he presence of windows and daylight will increase the
frequency of communication and social interaction.
4: T he presence of windows and daylight will improve
performance, as evidenced by reduced frequency of
human-related medication errors.
The reduction of mean arterial pressure, combination
of systolic and diastolic, (p < 0.0001) was signicant.
The increase of temperature (0.0 26) was signicant.
The increase in blood oxygen saturation was signicant
(p=0.016) but the change was clinically trivial. The
reduction of heart rate was not signicant (p = 0.0 67).
The frequency of subsidiary behaviors that indicate
sleepiness and deteriorated mood was signicant
(p = 0.023). Self-assessed sleepiness was not signicant
(p = 0.095 ).
Increased frequency of communication (p < 0.0001) and
positive social interaction (measured by communication
followed by laughter) in nurse station was signicant
(p= 0.028).
The decrease in medical errors was not signicant
(p = 0 .14 ).
56 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROUP LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
e oxygen saturation increase in the windowed condition seemed to be fairly
consistent. One explanation may be tied to the sense of happiness and freshness
that windows and daylight bring to the windowed nurses’ station compared with
the windowless one. is may have stimulated deep breathing, which results in
increased oxygen saturation (Bernardi et al., 1998). More research is needed to
unravel the effect of windows and daylight on employees’ oxygen saturation.
Another explanation may be the stress-reducing effects of windows. For exam-
ple, a study by Standley and Moore (1995) showed that the stress-reducing effect
of music improved oxygen saturation in infants. is explanation will direct
attention back to the “micro-restorative” (Kaplan, 1993, p. 196) effect of win-
dows and its possible effect on oxygen saturation, similar to the effect on heart
rate described above. e effect of age on the oxygen saturation is expected as
supported by evidence (Genes, Chandra, Ellis, & Baumlin, 2013; Moraes et al.,
2014).
Our observations of a significantly increased body temperature in the morn-
ing are consistent with those of Turner, Van Someren, and Mainster (2010),
who noted that morning sunlight increases vigilance and core body temperature.
Increased body temperature may be related to increased performance (Wright,
Hull, & Czeisler, 2002). More specifically, Wright et al. (2002) explain that
regardless of the circadian phase, even an increase of about 0.27°F in body
temperature has been associated with enhanced memory and cognitive perfor-
mance. erefore, further study is suggested to trace the effect of the availability
of natural light in healthcare workspaces on caregivers’ memory and cognitive
performance.
Psychological and Behavioral Outcomes
Psychological and behavioral outcomes, including communication and laughter,
sleepiness and mood are discussed below.
Communication and Laughter
Both psychological and behavioral outcomes that measured communication and
positive sociability showed significant improvements. When working in the win-
dowed nurses’ station, RNs communicated significantly more frequently than
when they were working in the windowless nurses’ station. When the commu-
nication data were analyzed by location over the course of the workday, the
findings showed that the increase in communication occurred when the par-
ticipating RN was in the nurses’ station where the window and daylight were
available and did not change in any other nursing areas in the unit. Although
only communication by the participating RN was recorded, the positive effect
of the window on other RNs in the room may have played a role in increased
communication.
e increase in positive sociability, as measured by the occurrence of frequent
laughter, was also significant. Another interesting observation and lesson learned
was that an all-day observation per participant in all care areas may not be nec-
essary, as the trends are similar over time and the change actually occurred only
© 2014 VENDOME G ROUP L LC HEA LTH ENVIRONMENTS RESEA RCH & DESIGN JO URNA L 57
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
in the nurses’ station where the window was located. In most cases, the greatest
effect was visible in the morning (10:00 a.m. measurement).
Sleepiness
Of the three means of measurement—subsidiary behavior, caffeine intake, and
momentary self-assessment—only one showed significant improvement. e
results from behavior mapping indicated that the reduction in subsidiary behav-
ioral cues of mood and sleepiness was significantly lower in the windowed nurs-
ing stations. e anecdotal observations of the researchers during shadowing was
that self-assessment of sleepiness was highly dependent on stressful or emotional
events.
erefore, no conclusion can be made in terms of reduced momentary self-re-
ported sleepiness or caffeine intake in the windowed ward. For self-reported
sleepiness, the variability between subjects in reporting sleepiness was very high,
as indicated by a residual variability of 2.2. erefore, given the subjective type
of measurement (self-evaluation, which may be confounded by daily events and
emotions), we recommend the use of the same methodology with a larger sam-
ple size to overcome the effect of individual differences and study the effect of
subjective sleepiness in a statistical analysis. No other variable had an effect on
sleepiness.
An optimized study design for future research to capture the effects of windows
would involve a larger sample size in a given time interval in the nurses’ station
where the window is located as opposed to long hours shadowing in all nursing
areas. Given the time effect on sleepiness and the fact that sleepiness was signifi-
cantly reduced in the morning and not the afternoon, it may be a good strategy
for future research to conduct behavior observation during morning hours when
time-affected sleepiness and fatigue are not as prevalent. Another explanation for
this may be the varying amount of daylight over time. e total daylight level
in the windowed station was higher in the morning but dropped one below the
baseline (total electric light levels) for the windowless station in the afternoon.
Because evidence in other settings has shown that lighting therapy reduced caf-
feine use, increased social interaction, and reduced sleepiness (Kräuchi et al.,
1990) for depressed patients, further research is needed to investigate the effect
of windows and natural light on behavior and caffeine use in healthy workers
and caregivers. If increased exposure to quality lighting in the workplace reduc-
es sleepiness, and therefore caffeine intake by workers, such an environmental
design strategy may have a positive impact on employees’ health.
Medication Errors
e patient/nurse ratio and patient acuity and type were similar between the two
patient wards, but the frequency of errors for patients in the ward with the win-
dowed nurse station was one-fifth that in the windowless ward during the 3 years
of data reviewed. However, the result was not statistically significant because of
the low number of overall errors. Only 23 medication errors were reported and
58 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROUP LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
documented in the 3 years prior to our study. Although the hypothesis failed,
we suggest future research on medication errors using longer durations to com-
pensate for the small number of occurrences. Because errors are underreported
(Koohestani & Baghcheghi, 2009), we estimate that the actual number of errors
may be more than twice the numbers reported. Evidence indicates that only 25%
(Mayo & Duncan, 2004) to 47% (Blegen et al., 2004) of medication errors may
actually be reported by nurses. Common underlying reasons for missed reports
are fear of supervisor or coworker pressure or belief that the incident was not
important enough to report (Mayo & Duncan, 2004). Risk-free and blame-free
reporting and improved documentation procedures for nurses can help with the
reliability of data sources and therefore contribute to future research on improv-
ing environments for nurses.
Limitations
is research study was conducted in a controlled setting. However, as with other
field research, it was not possible to control for all existing variables. Although
the furniture, interior finishes, information technology, workload, patient type,
patient acuity, and equipment were similar in the two nurses’ stations (windowed
and windowless), the seating arrangements differed, which may have affected the
rate of communication.
Both nurses’ stations were in double-loaded corridors with
similar patient rooms in form and arrangement. However, the
windowless nurses’ station was located in the middle of the
ward, and the one with windows was centered in the left side
of the corridor. is arrangement could have impacted the
nurses’ walking distance. However, this factor was controlled
to a great extent by charge nurses who made random room
assignments.
e observation portion of the study in which the researcher
shadowed the participant in the windowed and windowless stations measuring
subsidiary behavior, communication, and positive sociability was not research-
er-blind. However, the collection of the other data was blind, including bio-
logical data, participant self-reporting of sleepiness, and medical records. is
excluded possible researcher bias.
Conclusion
Evidence from a laboratory setting has confirmed that lighting designed with
characteristics similar to those of sunlight improves circadian adjustments
through neuro-hormonal effects, as expressed by elevated mood, alertness, vig-
ilance, and cognitive function (Czeisler et al., 1986, Postolache & Oren, 2005;
Kent et al., 2010). From a psychological perspective, the availability of win-
dows and daylight results in respite and mental restoration (Kaplan, 1993) and
has been shown to have stress-reducing effects (Walch et al., 2005) on various
Lighting design in clinical
workplaces should emphasize not
only the minimum light intensity for
clear vision but also the biological
need to adjust workers’ circadian
functions to improve performance.
© 2014 VENDOME G ROUP L LC HEA LTH ENVI RONME NTS RE SEA RCH & DESIGN JO URNA L 59
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
populations. Overall, our findings support evidence from field and laboratory
research on the positive effects of windows on acute-care RNs.
Implications for Practice
• e presence of windows and daylight may have psychological, psycho-
logical, and behavioral benefits related to circadian rhythms (evidenced
by temperature), communication, and positive sociability on RNs work-
ing a day shift.
• Because alertness is connected to both staff and patient safety, maximiz-
ing access to daylight and providing quality lighting design in nursing
areas may be an opportunity to improve safety though environmental
design and enable staff to manage sleepiness and stay alert.
• e best source of lighting for human health is daylight. Past studies
have shown that, under similar conditions, daylight may have signifi-
cantly greater effects than incandescent and fluorescent light on circadian
adjustments. erefore, although maximizing the availability of daylight
should be one of the main goals for clinical workspace design, optimizing
electric lighting to support circadian rhythms is an important goal given
the limited presence of natural daylight.
• Lighting design in clinical workplaces should emphasize not only the
minimum light intensity for clear vision but also the biological need to
adjust workers’ circadian functions to improve performance.
• What Steven and Rea (2001) refer to as “biological darkness” occurs
when indoor lighting is not adjusted for the human body’s biological
needs, resulting in circadian disruption with serious consequences on
health and performance. erefore, the physical environment in which
the caregivers work on critical tasks should be designed to support a
high-performing and healthy clinical staff.
60 WWW.HERDJOURNAL.COM © 2014 VENDOME G ROUP L LC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
References
aan het Rot, M., Moskowitz, D. S., & Young, S. N. (2008). Exposure to bright light is associated
with positive social interaction and good mood over short time periods: A naturalistic study
in mildly seasonal people. Journal of Psychiatric Resea rch, 42, 311–319.
Alimoglu, M. K., & Donmez, L. (2005). Daylight exposure and the other predictors of burnout
among nur ses in a universit y hospital. International Journal of Nur sing Studies, 42, 549–555.
Aries, M. B. C., Veitch, J. A., & Newsham, G. R. (2010). Windows, view, and of ce character-
istics predict physical and psychological discomfort. Journal of Enviro nmental Psychology,
30(4), 533–541.
Badia, P., Myers, B., Boecker, M., Culpeppe r, J., & Harsh, J. R. (1991). Bright ligh t effects on body
temperature, aler tness, EEG and behavior. Physiology & Behavior, 50, 583–588.
Beale, D., Lawrence, C., Leather, P., & Pyrgas, M. (1998). Windows in the workplace: Sunlight,
view, and occupational stress. Environment and Behavior, 30(6), 739–762.
Begemann, S. H. A., van den Beld, G. J., & Tenner, A. D. (1997). Daylight, ar ticial light and peo-
ple in an ofce environment, overview of visual and biological responses. International Jour-
nal of Industrial Ergonomics, 20, 231–239.
Bernardi, L., Spadacini, G., Bellwon, J., Hajric, R., Roskamm, H., & Frey, A. W. (1998). Effect
of breathing rate on oxygen saturation and exercise per formance in chronic hear t failure.
La nce t, 351, 130 8 –1311.
Benedetti, F., Colombo, C., Barbini, B., Campori, E., & Smeraldi, E. (20 01). Morning sunlight
reduces length of hospitalization in bipolar depression. Jour nal of Af fective D isorders, 62(3),
221–223.
Benedetti, F., Colombo, C., Pontiggia, A., Be rnasconi, A., Florita, M., & Smeraldi, E. (2003).
Morning light treatment hastens the antidepres sant ef fect of citalopram: A placebo-con-
trolled trial. Journal of Clinical Psyc hiatr y, 64(6), 648–653.
Blegen, M. A., Vaughn, T., Pepper, G., Vojir, C., Stratton, K., Boyd, M., & Armstrong, G . (2004).
Patient and staff safety: Voluntary reporting. American Journal of Medical Quality, 19(2),
67–74. d oi:10 .1177/10628 60 6040190 02 04
Boyce, P. (1997). Light, sight and photobiology (Health effects of lighting). New York, NY:
Rensselaer Polytechnic Institute. Retrieved from http://www.lrc.rpi.edu/publicationDetails.
asp?id=179
Buchanan, T., Barker, K., Gibson, J., Jiang, B., & Pearson, R. (1991). Illumination and errors in
dispensing. American Journal of Health-System Pharmacy, 48(10), 2137–2145.
Cajochen, C. (2007). Ale rting effects of light. Sl eep Medicine Reviews, 11, 453–464.
Chaudhury, H., Mahmood, A., & Valente, M. (2009). The effect of environmental design on reduc-
ing nursing error s and increasing efciency in acute care settings: A review and analysis of
the literature. Environment a nd Behavior, 41(6), 755–78 6.
Crepeau, L. J., Bullough, J. D., Figueiro, M. G., Porter, S., & Rea, M. S. (2006). Lighting as a cir-
cadian rhythm-entraining and alertness-enhancing stimulus in the submarine environment.
(Unpublished conference paper). Undersea HSI Symposium: Research, Acquisition, and the
Warrior. Mystic, CT. Retrieved from http://cogprints.org/6574/
Czeisler, C., Allan, J., Strogat z, S., Ronda, J., Sanchez, R., Rios, C., ... Kronauer, R. (1986). Bright
light resets the human circadian pacemaker independent of the timing of the sleep wake
cycle. Science, 233, 667–671.
Feagai, H. E. (2011). Let humor lead your nursing practice. Nur se Lead er, 9(4), 44–46. doi:
10.1016/j.m nl.2010.10.0 05
Genes, N., Chandra, D., Ellis, S., & Baumlin, K. (2013). Validating emergency depar tment vital
signs using a data quality engine for data warehouse. TOMINFOJ The Open Medical Infor-
matics Journal, 7(1), 34–39.
Gifford, R. (1988). Light, decor, arousal, comfor t and communication. Journal of Environmental
Psychology, 8, 177–18 9.
Joseph, A . (2006). The impact of light o n outcomes in hea lthcare settings. Concord, CA: T he
Center for Health Design. Retrieved from http://www.healthdesign.org/sites/default/les/
CHD_Issue_Paper2.pdf
© 2014 VENDOME G ROUP L LC HEA LTH ENVIRONMENTS RESEA RCH & DE SIGN JO URNAL 61
IMPACT O F WINDOWS A ND DAYL IGH T ON NU RSES’ HEA LTH RESEARCH
Kamali, N. J., & Abbas, M. Y. (2012). Healing environment: Enhancing nurses’ performance
through proper lighting design. Procedia - Social and Behavioral Sciences, 35, 205–212.
Kaplan, R. (1993). The role of nature in the context of the workplace. Landscape and Urban
Planning, 26(1-4), 19 3 –2 01.
Kent, S. T., McClure, L. A., Crosson, W. L., Arnett, D. K., Wadley, V. G., & Sathiakumar, N. (2009).
Effect of sunlight exposure on cognitive function among depressed and non-depres sed par-
ticipants: A REGARDS cross-sectional study. Environmental Health, 8, 34– 48.
Kim, S., Cranor, B. D., & Ryu, Y. S. (2009). Fatigue: Working under the inuence. Proceedings
of the XXIst Annual International Occupational Ergonomics and Safety Confer ence, Dallas,
Texas, USA, June 11–12 (pp. 317–322).
Koohesta ni, H., & Baghcheghi, N. (20 09). Barriers to the repor ting of medication administration
errors among nursing students. Australian Journal of Advanced Nursing, 27(1), 66–74.
Kräuchi, K., Wirz-Justice, A ., & Graw, P. (1990). The relationship of affective state to dietary pref-
erence: Winter depression and light the rapy as a model. Jour nal of Af fective D isorders, 20 (1),
43–53. doi:10.1016/0165-0327(90)90048-D
Mayo, A. M., & Duncan, D. (2004). Nurse perceptions of medication errors—What we need to
know for patient safet y. Journal of Nursing Care Quality, 19(3 ), 2 0 9– 217.
Mroczek, J., Mikitarian, G., Vieira, E. K., & Rotarius, T. (2005). Hospital design and staf f percep-
tions: An explorator y analysis. The Health Care Manager, 24(3), 233–244.
Moraes, W., Piovezan, R., Poyares, D., Bittencourt, L . R., Santos-Silva, R., & Tuk, S. (2014).
Effects of aging on sleep structure throughout adulthood: A population-based study. Sleep
Medicine, 15(4), 401–409. doi:http://dx.doi.org/10.1016/j.sleep.2013.11.791
Münch, M., Linhart, F., Borisuit, A., Jaeggi, S. M., & Scartezzini, J. (2012). Effects of prior light
exposure on early evening performance, subjective sleepiness, and hormonal secretion.
Behavioral Neuroscience, 126(1), 19 6 –2 03 .
Myers, B. L ., & Badia, P. (1993). Immediate ef fects of differe nt light intensities on body tempera-
ture and alertness. Physiology and Behavior, 54, 199– 20 2.
National Space Biomedical Research Institute. (2010). To sleep or not to sleep? Math software
to help plan astronaut, shift worker sche dules. ScienceDaily. Retrieved from http://www.sci-
encedaily.com/releases/ 2010/04/100414122639.htm
Noell-Waggoner, E. (2006). Lighting in nursing homes—The unmet need. Proceedi ngs of the 2nd
International Commission on Illumination Expert Symposium on Lighting and Human Health,
Ottawa, Ontario, Canada, September 7–8 (pp. 77–81). Retrieved from http://www.centerofde-
sign.org/pdf/LightingNursingHomeUnmetNeed.pdf
Olsen, R. V., Hutching, B. L., & Ehrenkrantz, E. (200 0). Media memory lane: Inter ventions in an
Alzheimer’s day care center. American Journal of Alzheimer’s Disease, 15(3), 163 –175 .
Ott, R. L ., & Longne cker, M. (2008). An introduction to statistical methods and data analysis (6th
ed.). Pacic Grove, CA: Duxbury Press.
Palmer, G. K. (2005). Laughter: The perfec t family medicine. Marriage & Families, 15(1), 23– 25 .
Pati, D., Harvey, T. E., & Barach, P. (2008). Relationships between exterior views and
nurse stress: An explorator y examination. Health Environments Research & Design Journal,
1(2), 27– 38.
Pauley, S. M. (2004). Lighting for the human circadian clock: Recent research indicates that
lighting has become a public health issue. Medical Hypotheses, 63(4), 588 –596.
Postolache, T. T., & Oren, D. A. (2005). Circadian phase shifting, aler ting, and antidepressant
effects of bright light treatment. Clinical Sports Medicine, 24(2 ), 381– 413 .
Rea, M., Figueiro, M., & Bullough, J. (2002). Circadian photobiology: an emerging framework
for lighting practice and research. Lighting Research and Technology, 34(3), 177–187. doi:
10.1191/1365782802lt 057
Rea, M. S., Figueiro, M. G., Bullough, J. D., & Bierman, A . (2005). A model of phototransduction
by the human circadian system. Brain Research R eviews, 50(2) , 213 -2 28.
Roberts, J. E. (2010). Circadian rhythm and human health. Retrieved from http://www.photobi-
ology.info/Roberts-CR.html
Rogers, M. A., Small, D., Buchan, D. A ., Butch, C. A., Stewart, C. M., Krenzer, B. E., &
Husovsky, H. L. (2001). Home monitoring ser vice improves mean arterial pres sure in patients
62 WWW.HERDJOURNAL.COM © 2014 VEN DOME G ROU P LLC
SUMM ER 2014 • VOL. 7 NO. 4, pp. 36 –62
RESEARCH
with essential hypertension. A randomized, controlled trial. Annals of Internal Medicine,
134(11), 1024 –103 2.
Rossberg, J. I., Melle, I., Opjordsmoen, S., & Friis, S. (2008). The relationship bet ween staff
member s’ working conditions and patie nts’ perceptions of the treatme nt environment. Inter-
national Journal of Social Psychiatry, 54(5), 437–446. doi:10.1177/0020764008090689
Scott, A . J. (2000). Shift work and health. Primary Care, 27(4), 1057–1079.
Sharma, V. K., Chandrashekaran, M. K., & Nongkynrih, P. (1997). Daylight and articial light
phase response curves for the circadian rhythm in locomotor activity of the eld mouse Mus
booduga. Biological Rhythm Research, 28(S1), 39 -49.
Sherman, S. A., Varni, J. W., Ulrich, R. S., & Malcarne, V. L. (20 05). Post-occupancy evaluation of
healing gardens in a pediatric cancer center. Landscape and Urban Planning, 73(2), 167-183.
Smolders, K. C. H. J., de Kort, Y. A. W., & Cluitmans, P. J. M. (2012). A higher illuminance induces
alertness even during ofce hours: Findings on subjective measures, task performance a nd
heart rate measures. Physiology & Behavior, 107, 7–16.
Standley, J. M., & Moore, R. S. (1995). Therapeutic effects of music and mother’s voice on pre -
mature infants. Pediatric Nursing, 21(6), 509.
Stevens, R. G., & Rea, M. S. (20 01). Light in the built environment: potential role of circadian dis-
ruption in endocrine disruption and breast cancer. Cancer Cause s & Control, 12(3), 279–287.
Stone, A. A ., & Shiffman, S. (1994). Ecological momentar y assessment (EMA) in behavioral me d-
icine. Annals of Behavioral Medicine, 16, 19 9 –202.
Takanishi, T., Ebara, T., Murasaki, G. I., Kubo, T., Tachi, N., Itani, T., & Kamijima, M. (2010). Inter-
active model of subsidiary behaviors, work per formance and autonomic ne rve activity during
visual display terminal work. Journa l of Occupational Health, 52(1 ) , 3 9 – 47.
Turner, P. L., Van Someren, E. J. W., & Mainster, M. A. (2010). The role of environmental light
in sleep and health: Ef fects of ocular aging and cataract surgery. Sleep Medicine Reviews,
14(4), 269 –280.
Ulrich, R. S. (1984). View through a window may inuence recovery from surgery. Science, 224,
420–421.
van Bommel, W. J. M., & van den Beld, J. G. (2004). Lighting for work: A review of visual and
biological effects. Lighting Rese arch and Technology, 36, 255–269.
Varni, J. W., Seid, M., & Rode, C. A. (1999). The PedsQL™: Measurement Model for the Pediatric
Quality of Life Inventory. Medical Care, 37(2), 126 –139 .
Walch, J. M., Rabin, B. S., Day, R., Williams, J. N., Choi, K., & Kang, J. D. (2005). The effect of
sunlight on postoperative analgesic medication usage: A prospective study of spinal surgery
patients. Psychosomatic M edicine, 67(1), 156 –16 3.
Wright, K. P., Hull, J. T., & Czeisler, C. A. (2002). Relationship between aler tness, performance,
and body temperature in humans. American Journal of Physiology—Regulatory, Integrative
and Comparative Physiology, 283(6), R13 70 – R137 7. doi:10.1152/aj pr egu. 0020 5. 200 2
