DOI: 10.1126/science.6143402
, 420 (1984);224Science
et al.RS Ulrich,
from surgery
View through a window may influence recovery
www.sciencemag.org (this information is current as of March 11, 2007 ):
The following resources related to this article are available online at
http://www.sciencemag.org
version of this article at:
including high-resolution figures, can be found in the onlineUpdated information and services,
http://www.sciencemag.org#otherarticles
, 2 of which can be accessed for free:cites 4 articlesThis article
http://www.sciencemag.org#otherarticles
49 articles hosted by HighWire Press; see:cited byThis article has been
http://www.sciencemag.org/about/permissions.dtl
in whole or in part can be found at:this article
permission to reproduceof this article or about obtainingreprintsInformation about obtaining
registered trademark of AAAS.
c 1984 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
onMarch11,2007www.sciencemag.orgDownloadedfrom
View Through a Window May Influence Recovery from
Surgery
Abstract. Records on recovery after cholecystectomy ofpatients in a suburban
Pennsylvania hospital between 1972 and 1981 were examined to determine whether
assignment to a room with a window view ofa natural setting might have restorative
influences. Twenty-three surgical patients assigned to rooms with windows looking
out on a natural scene had shorter postoperative hospital stays, received fewer
negative evaluative comments in nurses' notes, and took fewer potent analgesics
than 23 matched patients in similar rooms with windowsfacing a brick building
wall.
Investigations of aesthetic and affective
responses to outdoor visual environments
have shown a strong tendency for
American and European groups to prefer
natural scenes more than urban views
that lack natural elements (1, 2). Views
of vegetation, and especially water, appear
to sustain interest and attention
more effectively than urban views of
equivalent information rate (2). Because
most natural views apparently elicit positive
feelings, reduce fear in stressed subjects,
hold interest, and may block or
reduce stressful thoughts, they might
also foster restoration from anxiety or
stress (3).
The restorative effect of natural views
on surgical patients was examined in a
suburban Pennsylvania hospital (200
beds). Such patients often experience
considerable anxiety (4, 5), and hospital
confinement limits their access to outdoor
environments almost entirely to
views through windows. Views to the
outside may be especially important to
individuals who have unvarying schedules
and spend a great deal of time in the
same room (6), such as surgical patients.
It is possible that a hospital window view
could influence a patient's emotional
state and might accordingly affect recov-
ery.
Records of patients assigned to rooms
on the second and third floors of a threestory
wing of the hospital between 1972
and 1981 were obtained. Windows on
one side of the wing look out on either a
small stand of deciduous trees or a
brown brick wall (Fig. 1). The same
nurses are assigned to the rooms on a
given floor; the nurses' stations are located
somewhat closer to the wall-view
rooms on both floors. The rooms are all
for double occupancy and are nearly
identical in terms ofdimensions, window
size, arrangement of beds, furniture, and
other major physical characteristics.
Each room has a single window 1.83 m
high and 1.22 m wide with the lower edge
74 cm above the floor. The size and
placement of the window allow an unobstructed
view out for a patient lying in
bed on either side of the room. The
rooms differ, therefore, essentially only
in what is seen through the window.
Patients are assigned to rooms as they
become vacant.
The sample consisted exclusively of
patients who had undergone cholecystectomy,
a common type of gall bladder
surgery. This is a comparatively standardized
procedure with similiar postoperative
management in the uncomplicated
cases. Only cholecystectomies performed
between 1 May and 20 October
(1972 through 1981) were identified because
the trees have foliage during those
months. Patients younger than 20 years
or older than 69, patients who developed
serious complications, and those with a
history of psychological disturbances
were excluded. Patients were then
matched so that one member ofeach pair
had a view of the trees and the other, the
brick wall. The criteria for matching
were sex, age (within 5 years), being a
smoker or nonsmoker, being obese or
within normal weight limits, general nature
of previous hospitalization, year of
surgery (within 6 years), and floor level.
Patients on the second floor, a surgical
floor, were also matched by the color of
Table 1. Comparison of analgesic doses per patient for wall-view and tree-view groups.
Number of doses
Analgesic Days 0-1 Days 2-5 Days 6-7
strength
Wall Tree Wall Tree Wall Tree
group group group group group group
Strong 2.56 2.40 2.48 0.96 0.22 0.17
Moderate 4.00 5.00 3.65 1.74 0.35 0.17
Weak 0.23 0.30 2.57 5.39 0.96 1.09
420
their room (rooms on that floor alternate
between blue and green). The 6-year
interval for year of surgery was established
on the basis of inquiries concerning
possible changes in procedures.
There was no statistically significant difference
in the sampling distributions by
year of surgery between the wall-view
and tree-view groups. The final data base
consisted of records of 46 patients
grouped into 23 pairs (15 female and 8
male). An attempt was made to match
patients by physician, but this was possible
for only seven pairs because the
number of doctors was large. However,
for the remaining pairs the distribution of
different physicians was similar in the
two groups. There was no instance, for
example, when patients of the same doctor
all had rooms with same view.
Recovery data were extracted from
the records by a nurse with extensive
surgical floor experience. The nurse did
not know which scene was visible from a
patient's window. Five types of information
were taken from each record: number
of days of hospitalization; number
and strength of analgesics each day (7);
number and strength of doses for anxiety,
including tranquilizers and barbiturates,
each day (8); minor complications,
such as persistent headache and nausea
requiring medication-symptoms which
are considered to result frequently from
conversion reactions (9); and all nurses'
notes relating to a patient's condition or
course of recovery.
Length of hospitalization was defined
as day of surgery to day of discharge.
These data were assumed to be only
ordinal because surgery was performed
at different times of day and discharge
times were somewhat different. The records
showed that patients with window
views of the trees spent less time in the
hospital than those with views of the
brick wall: 7.96 days compared with 8.70
days per patient [Wilcoxon matchedpairs
signed-ranks analysis, T(17) = 35,
z = 1.965, P = 0.025].
Nurses' notes consisted of comments
about the patient's condition written during
the postsurgical period ending at midnight
of the seventh recovery day after
the day of surgery. Notes were classified
as negative or positive-for example,
negative notes included "upset and crying"
or "needs much encouragement,"
and positive notes included "in good
spirits" and "moving well." More negative
notes were made on patients with
the brick wall view: 3.96 per patient
compared to 1.13 per patient with the
tree view [Wilcoxon matched-pairs
signed-ranks analysis, T(21) = 15,
z = 3.49, P < 0.001]. Although more
SCIENCE, VOL. 224
positive comments were recorded for the
tree-view patients, the difference was
not statistically significant.
The multivariate two-sample Hotelling
test was used to compare the groups for
analgesic intake (10). The average number
of doses per patient, within each
strength level, was computed for (i) the
day of surgery and first recovery day, (ii)
days 2 through 5 after surgery, and (iii)
days 6 and 7 after surgery. It was expected
that for the first period no differences
in analgesic intake would be found between
the two groups, because patients
would have been too drugged or too
absorbed by intense pain to attend to the
windows (5). It was also expected that
there would be no significant variation
across groups in the final two days. In
fact, only 45 percent of the patients took
any analgesics after the fifth day. The
data are summarized in Table 1.
For the period of primary interest,
days 2 through 5, there were statistically
significant variation between the treeview
and wall-view patients in the mean
number of analgesic doses (72 = 13.52,
F = 4.30, P < 0.01). In the other two
periods there were no significant differences.
In days 2 through 5 patients with
the tree view took fewer moderate and
strong pain doses than did the wall-view
group and more doses in the weak category.
The wall group, therefore, was
given many more doses ofpotent narcotics,
whereas the tree group more frequently
received such drugs as aspirin
and acetaminophen.
With respect to doses of antianxiety
drugs, there was no significant variation
between the groups. Wall-view patients
were given more doses of narcotic analgesics,
which produce drowsiness or sedation
as side-effects, possibly reducing
their need for sleeping pills or tranquilizers.
To test this inverse relation, antianxiety
dose frequencies were compared
when patients took either no or one
strong or moderate analgesic dose or at
least two strong or moderate analgesic
doses. The observed frequency of doses
was lower than the expected frequency
when two or more strong or moderate
analgesics were taken on the same day
[X2(1) = 10.45, P < 0.01]. The intake of
narcotic analgesics by patients with the
wall view may have lowered their use of
antianxiety drugs to that of patients with
the tree view.
A weighted score of minor postsurgiFig.
1. Plan of the second floor of the study
hospital showing the trees versus wall window
views of patients. Data were also collected
for patients assigned to third-floor rooms.
One room on each floor was excluded because
portions of both the trees and wall were
visible from the windows. Architectural dimensions
are not precisely to scale.
cal complications (excluding routine
postanesthetic occurrences such as nausea)
was computed for each patient, with
criteria and procedures similar to those
used by Cohen and Lazarus (9, 11).
Although tree-view patients had lower
scores, the difference was not statistically
significant. This small difference
found may be due to the greater intake of
potent analgesics by the wall-view group
rather than to a possibly higher frequency
of conversion reactions.
In summary, in comparison with the
wall-view group, the patients with the
tree view had shorter postoperative hospital
stays, had fewer negative evaluative
comments from nurses, took fewer
moderate and strong analgesic doses,
and had slightly lower scores for minor
postsurgical complications. Although the
findings suggest that the natural scene
had comparatively therapeutic influences,
it should be recognized that the
"built" view in this study was a comparatively
monotonous one, a largely featureless
brick wall. The conclusions cannot
be extended to all built views, nor to
other patient groups, such as long-term
patients, who may suffer from low arousal
or boredom rather than from the anxiety
problems typically associated with
surgeries. Perhaps to a chronically understimulated
patient, a built view such
as a lively city street might be more
stimulating and hence more therapeutic
than many natural views. These cautions
notwithstanding, the results imply that
hospital design and siting decisions
should take into account the quality of
patient window views.
ROGER S. ULRICH
Department of Geography,
University ofDelaware, Newark 19716
References and Notes
1. J. F. Wohlwill, in Human Behavior and Environment,
I. Altman and J. F. Wohlwill, Eds.
(Plenum, New York, 1976), vol. 1, pp. 37-86; E.
H. Zube, D. G. Pitt, T. W. Anderson, in Landscape
Assessment: Values, Perceptions, and
Resources, E. H. Zube, R. 0 Brush, J. G.
Fabos, Eds. (Dowden, Hutchinson, & Ross,
Stroudsburg, Pa., 1975), pp. 151-167; J. F.
Palmer, in Priorities for Environmental Design
Research, S. Weidemann and J. R. Anderson,
Eds. (Environmental Design Research, Washington,
D.C., 1978), pp. 92-103.
2. R. S. Ulrich, Landscape Res. 4, 17 (1979);
Environ. Behav. 13, 523 (1981).
3. in Human Behavior and the Natural
Environment, I. Altmian and J. F. Wohlwill,
Eds. (Plenum, New York, 1983), pp. 85-125.
4. A. Martinez-Urrutia, J. Consult. Clin. Psychol.
43, 437 (1975); E. J. Langer, I. L. Janis, J. A.
Wolfer, J. Exp. Soc. Psychol. 11, 155 (1975).
5. C. R. Chapman and G. B. Cox, J. Psychosom.
Res. 21, 7 (1977).
6. B. L. Collins, Windows and People: A Literature
Survey (NBS Building Science Series 70,
National Bureau of Standards, Washington,
D.C., 1975).
7. Analgesic doses were classified as weak, moderate,
or strong on the basis of the drug, dosage,
patient's weight, and whether the drug was
administered orally or by injection. Examples in
the weak category include acetaminophen and
acetaminophen with small amounts of codeine.
Examples in the moderate strength class are
injections of meperidine hydrochloride (Demerol)
in doses up to 50 mg, and tablets of oxycodone
hydrochloride-oxycodone terephthalate
(Percodan); strong analgesics include hydromorphone
hydrochloride (Dilaudid) and large doses
of meperidine hydrochloride.
8. Tranquilizers and barbiturates doses were classified
as weak, moderate, or strong according to
similar criteria used for analgesics. The antianxiety
drugs in the weak category were tranquilizers;
the moderate class was dominated by
tranquilizers; and the vast majority in the strong
category were large dosages of barbiturates. No
distinction was made between tranquilizers and
barbiturates in terms of function because tranquilizers
were often prescribed as sleeping medi-
cations.
9. F. Cohen and R. S. Lazarus, Psychosom. Med.
35, 375 (1973).
10. D. F. Morrison, Multivariate Statistical Methods
(McGraw-Hill, New York, ed. 2, 1976), p.
137.
11. Minor complications were scored as follows:
nausea (1 point); nausea requiring medication (2
points); antacids given (I point); rectal tube for
gas (1 point); inability to move bowels, enema
given (2 points); inability to void, catheterization
required (2 points); medication for diarrhea,
gut irritability, or both (1 point); medication for
constipation or for urine stimulation (1 point);
and antibiotics given for postoperative fever and
infection (3 points).
12. The collection of data was facilitated by the full
cooperation of W. M. Tomlinson and A. King,
president and director of medical records, respectively,
of Paoli Memorial Hospital, Paoli,
Pa. I thank M. Mozzani for reading patient
records, A. E. Hoerl for statistical consultation,
and several physicians for their helpful comments
on research methods. I also thank S. M.
Parnes, L. P. Herrington, M. Zuckerman, and
T. C. Meierding. Supported by grant 23170 from
the Consortium for Environmental Forestry
Studies, U.S. Department of Agriculture Forest
Service.
24 January 1983; accepted I November 1983
27 APRIL 1984 421