Patient and staff environments: The impact of daylight and windows on ICU patients and staff
Access to windows and window views are particularly important in intensive care units where patients are more vulnerable and staff are likely to be more susceptible to stress
Dr Mardelle Shepley AIA, ACHA, LEED AP, Raymond Gerbi, Angela Watson AIA, Stephen Imgrund MD
When Concord Hospital in New Hampshire decided to replace its existing intensive care unit (ICU), a focus on both patient and staff needs was deemed essential. While patient requirements were typically recognised as a priority, staff needs were often poorly addressed. Based on the culture at Concord Hospital, the design team decided to support both patients and staff by creating access to daylight and views, not only from patient rooms but also from staff areas.
This research project looks at the impact of windows and window views on patient outcomes and staff behaviour in an ICU, comparing the old ICU at Concord Hospital to the newly constructed one, which was designed to increase window views and daylight within the facility.
The new ICU was planned for the second floor of an addition that is located on a sloping and constricted site, next to an existing road. To accommodate the unit within a two-storey grade change, the design team carved out gardens around the addition, creating views into a large, sunken garden on the north, which would have otherwise been underground.
The adjacent roof to the west was planned as an accessible garden. As a result, all patient rooms and many staff areas have direct views to gardens or distant views of nature. While some of these design elements required more effort and capital investment, the design team believed that this would be offset by the benefits of nature views and daylight on the unit.
Evidence exists suggesting that windows in ICUs can reduce patient delirium1,2. In addition, views of nature can reduce postoperative stays, negative evaluative comments in nurses’ notes and requests for strong analgesics in acute care units3.
Furthermore, natural sunlight can reduce the perception of pain and requests for medication4. Regarding staff, windows have been found to improve job satisfaction and retention in offices5, although no studies were found that addressed the impact of windows on healthcare staff.
Research relevant to this topic includes studies on the impact of windows, the impact of views and daylight, and the nature of stress in intensive care units on patients and staff.
Studies on the impact of windows
Markus6 emphasised four factors which influence window design: sunshine, awareness, view and lack of privacy.
In his study he noted that being close to a window was highly desirable regardless of the size of the visual field. Keighley7 found that satisfaction regarding windows was influenced by area and proportion and the number and width of mullions. The most preferred windows were horizontal apertures occupying 25-30% of the exterior wall. Roessler8 found that unpleasant feelings of enclosure were minimal with a window width of at least 1.5 metres.
The ideal was two lateral windows with a total width of 3-4 metres in a six-metre-wide room. Finnegan and Solomon9 found differences regarding job satisfaction, how interesting the job was perceived to be, physical working conditions and overall experience in favour of windowed spaces.
Windows in healthcare settings
In 1972 Wilson compared the incidence of postoperative delirium in patients located in windowed and windowless ICUs2. Twice as many windowless patients demonstrated delirium and, among patients with abnormal haemoglobin or blood urea, the incidence was threefold. Hallucinations were more than twice as high in a new windowless unit than in the old unit1.
Ulrich3 noted that gall bladder surgery patients who had nature views had a shorter length of stay, took less pain medication and made fewer negative comments than those who had views of a building wall. Verderber10 found that windows with high sills, distant from the viewer or obscured by walls and furnishing, were ranked as poorly as having no windows at all.
Verderber and Reuman11 compared inpatient and staff experience to windows and views. Patients were more negatively impacted by rooms with limited fenestration than staff. Being located more than 10 feet from a window significantly impacted patients who were immobile, visually impaired or non-Caucasian.
Sunlight and daylight in healthcare
Ne’eman12 noted the following positive contributions of sunshine: warmth, functional lighting, contact with the outside and biological effects of solar radiation. He found that 2% of patients and 62% of staff considered sunlight to be a nuisance, while 91% of patients and 31% of staff considered sunlight to be pleasurable.
In hospitals, when forced to choose between good views without indoor sunshine and unpleasant views with indoor sunshine, 50% preferred the former while 31% preferred the latter. Walch et al4 compared the use of pain medications in patients who were on the bright and dim side of a hospital. Those on the bright side were exposed to 46% higher sun intensity and perceived less stress and less pain and took fewer analgesics.
Of all the participants in the ICU setting, patients are the most vulnerable to stress. Lusk and Lash13 have identified three main categories of ICU stressors: psychological (danger of death, social isolation), treatment-related (artificial ventilation, tubes, painful procedures) and environmental (unfamiliar surroundings, activity and noise, worrisome sights and sensations, sleep deprivation).
The most comprehensive body of literature regarding the experience of ICU patients addresses a phenomenon known as ICU psychosis or ICU syndrome. Behaviours associated with this syndrome include disorientation and dysfunction. These problems express themselves in the following behaviours: lack of awareness of one’s environment, memory loss, sleep disorders, hallucinations, anger, fear and depression14.
While some occurrences of delirium may also be specifically associated with a physical life-threatening disorder15, most researchers acknowledge the independent impact of the environment as well.
Multiple authors have identified the physical environment as a contributor to ICU syndrome16-18. In a literature review by Easton and Mackenzie19, noise, constant light levels, unfamiliar environments and sensory overload were implicated in ICU syndrome. Many of the negative environmental attributes lead to sleep deprivation, which in turn may have a critical impact on psychosis.
Stress has a significant impact on all participants in the healthcare environment; while staff may appear to be less affected than patients, there is significant research to suggest that the medical professions are among the most emotionally challenging.
ICU nurses may be particularly susceptible to stress20. In a recent study, Mealer, Shelton, Berg, Rothbaum and Moss21 found a higher incidence of post-traumatic stress syndrome in ICU nurses (see also Maloney22).
With specific regard to the physical environment, researchers have identified the following factors that negatively impact staff: inappropriate access to equipment, inadequate storage space, poor design, inadequate function of biomedical equipment23 and poor equipment, inadequate work space, noise, poor work environment, lack of supplies, too many people and inappropriate lighting24.
The literature on ICU physician stress is more limited, although it is clear that doctors experience intense challenges in the medical environment. Coomber, Todd, Park, Baxter, Firth-Cozens and Shore25 found that one-third of ICU physicians were distressed, one-tenth were depressed and 3% experienced suicidal thoughts (see also Firth-Cozens26).
Of the stressors that were reported, the most frequent were the poor prognosis or death of a patient and the physical setting. The highest stressor was allocating beds when the unit was full, because sufficient capacity to provide care may be critical to patient survival. Goodfellow et al27 found that 27% of ICU physicians experience psychiatric difficulties.
This study compared the impact of increased daylighting and window views on staff, patients and families by comparing the behaviours and outcomes of subjects in an old intensive care unit (Phase 1), characterised by lower overall daylight levels and limited views of nature, to behaviours of subjects in a new ICU (Phase 2), which had higher daylight levels and improved views of nature.
The new ICU was occupied when the old ICU was shut down. Both ICUs were located in the same hospital and staffed by the same medical personnel, with the exception of those lost by attrition and new hires. This paper focuses on the results from Phase 1.
Phase 1 site: The old 11,600 building gross square foot ICU consisted of two adjoining wings in a building constructed in 1956 in a cruciform configuration. The unit consisted of 16 beds with eight beds on each wing. Each unit had a central nurses’ station at the intersection of the two wings with support spaces between the two stations and access from both.
Decentralised standing-height computer workstations were located on each wing closer to the patient rooms. The view from each nurse’s station and workstations did not allow for any exterior views other than one obstructed view through the window of the patient room nearest to the station, assuming the cubicle curtain was in the open position.
The only unobstructed exterior views available to staff were from two offices and a small staff lounge. Most views were of other buildings nearby. Fourteen of the patient rooms were approximately 174 square feet, each with two larger isolation rooms. In each patient room the free-standing headwall unit was angled at 45 degrees from an exterior corner of the room. Consequently, the patient’s view was oriented 135 degrees away from direct view of the windows.
Phase 2 site: The new 19,100 building gross square foot ICU is configured in two linked sub-units completed in 2008. The ICU consists of 20 beds with two large clinical work areas and decentralised nurse’s stations between pairs of rooms.
Both large clinical work areas have direct views to the outdoors via the termini of the corridors, while decentralised stations have views outside through patient rooms and through windows at the termini of the corridors of the facility. Several patient rooms, a corridor terminus, the staff lounge and two offices have views of a rooftop garden. The majority of the remaining patient rooms, a conference room and the terminus of another corridor look onto a newly planted garden.
The patient rooms range in size from 256 to 328 square feet, plus four isolation rooms ranging from 280 to 292 square feet, including the anteroom. In each patient room, the patient bed is oriented perpendicular to the window wall to allow for a direct view out of the window; a wall-mounted boom system for monitors and medical gases allows the bed to be oriented directly toward the window if clinical conditions allow.
Staff attrition, absenteeism and medical errors: Subjects included all medical staff assigned to the unit on a regular basis between October 2006 and September 2007 (Phase 1) and between March 2008 and February 2009 (Phase 2).
Patient length of stay, mortality and pain perception: During these timeframes, patient subjects included those who were staying in the ICU due to cardiac surgery (coronary bypass, aortic valve replacement), pneumonia and chronic obstructive pulmonary disease (COPD); patients who were residing in the ICU for two days or more; and patients whose rooms faced directly south, east, north or west. Fifty-eight were selected for inclusion in Phase 1.
Procedure: The protocol for this study was reviewed by the Institutional Review Board of the hospital. The anonymity of all subjects was maintained throughout the study.
Independent variables: The following data was gathered regarding window attributes (per Verderber11): proximity of head of bed to window, window to total wall area ratio (percentage), and sill height above floor. Sunlight intensity data (per Walch et al4) was also gathered. Light intensity (lux) was measured twice daily in patient rooms at approximately 9:30am and approximately 3:30pm within five days of 21 June, 21 September, 21 December and 21 March.
Prior to measurement, the door was closed and artificial lights turned off. Window blinds were opened. Direct, reflective and ambient measurements were taken. Data was also collected regarding views (% of nature of total view) as seen from the head of bed.
Dependent variables: Data was gathered for both phases by hospital staff and included: staff attrition, staff absenteeism, medical errors, patient length of stay, patient mortality and perceived pain.
This paper summarises the results of light level measurements for the new and old ICUs as well as data from Phase 1 regarding staff absenteeism and patient demographics, length of stay, pain perception and medical errors.
Light level measurements: The total average lux levels for each of the two units are shown in Figures 3 and 4. These results demonstrate a significant variation in light levels for the new unit compared to the old unit. Figure 5 compares the new and old units simultaneously and serves to demonstrate the overall higher average in the new unit.
Patient demographics, length of stay and pain perception: The average age of the 58 patient subjects (40 men and 18 women) was 66.6 years. The average length of stay of these subjects was 3.51 days, ranging from two to 15 days. Staff evaluated patient pain as a part of standard patient care protocol every few hours. The average pain described by patients (and on some occasions assessed by staff) on a scale of 0 to 10 was 1.48, ranging from 0 to 9 (see Figure 6).
Staff absenteeism: Data from Phase 1 indicated that staff absenteeism varied from 158 to 722 hours per month and averaged 368.33 hours (see Figure 7).
Medical errors: In the ICU, medical errors are reported voluntarily on a Process Improvement Data Sheet (PIDS). During Phase 1, 79 errors were reported, 36% of which were errors in the administering of medications, followed by errors associated with order entries (15%) (see Figure 8).
Much has been written about the impact of daylight and views of nature in waiting areas and acute care patient rooms, although little information is available regarding the impact of these environmental amenities on intensive care units.
Of all the spaces in hospitals, however, ICUs are among the most important, if a designer is concerned with providing the most sensitive design solutions in the locations where the participants (patients, families and staff) are most vulnerable.
According to Powell Lawton’s Environmental Competence Press Theory, the more compromised the personal resources of the individual, the more sensitive they are to negative aspects of the physical environment – and a balance must be achieved between the psychosocial competence of the individual and the challenge of the environment.
Both patients and staff are likely to be more stressed in the intensive care unit setting and, therefore, require a more supportive physical environment. This study provides a description of a methodology that could potentially be applied in similar studies involving pre- and post-occupancy evaluations.
The results provide data regarding the light levels in the old and new facilities as well as the patient length of stay and pain perception associated with the physical environment in the old ICU. Also provided is data regarding staff absenteeism and medical errors on the old unit.
This data can be used to compare to outcomes associated with the new facility when it becomes available, as well as serve as a reference and comparator for other studies on this topic.
Dr Mardelle Shepley, AIA, ACHA, LEED AP is director of the Center for Health Systems & Design at Texas A&M University
Raymond Gerbi is vice president of facilities at Concord Hospital
Angela Watson AIA specialises in the design of healthcare buildings at Shepley Bulfinch
Stephen Imgrund MD is a specialist in pulmonary medicine at Concord Hospital
1. Keep P, James J, Inman M. Windows in the intensive therapy unit. Anaesthesia 1980; 35:257-262.
2. Wilson L. Intensive care delirium: The effect of outside deprivation in a windowless unit. Archives of Internal Medicine 1972; 130:225-226.
3. Ulrich R. View through a window may infl uence recovery from surgery. Science 1984; 224:420-421.
4. Walch J, Rabin B, Day R, Williams J, Choi K, Kang J. The effect of sunlight on postoperative analgesic medication use: A prospective study of patients undergoing spinal surgery. Psychosomatic Medicine 2005; 67:156-163.
5. Farley K, Veitch J. A room with a view: A review of the effects of windows on work and well-being. National Research Council Canada, Institute for Research Construction. IRC Research Report, RR-136; 2001.
6. Markus T. The function of windows. Building Science 1967; 2:97-121.
7. Keighley E. Visual requirements and reduced fenestration in offi ces: A study of multiple apertures and window area. Building Science 1973; 8:321-333.
8. Roessler G. The psychological function of windows for the visual communication between the interior of rooms with permanent supplementary artificial lighting and the exterior. Lighting Research & Technology 1980; 12 (3):160-168.
9. Finnegan M, Solomon L. Work attitudes in windowed vs windowless environments. Journal of Social Psychology 1981; 115:291-292.
10. Verderber S. Dimensions of person-window transactions in the hospital environment. Environment and Behavior 1986; 18(4):450-466.
11. Verderber S, Reuman D. Windows, views, and health status in hospital therapeutic environments. The Journal of Architectural and Planning Research 1987; 4 (2):120-133.
12. Ne’eman E. Visual aspects of sunlight in buildings. Lighting Research and Technology 1974; 6 (3):159-164.
13. Lusk B, Lash AA. The stress response, psychoneuroimmunology, and stress among ICU patients. Dimensions of Critical Care Nursing 2005; 24(1):25-31.
14. Justic M. Does ICU psychosis really exist? Critical Care Nurse 2000; 20(3):28-37.
15. Divatia J. Delirium in the ICU. Indian Journal of Critical Care Medicine 2006; 10(4):215-218.
16. Gelling L. Causes of ICU psychosis: The environmental factors. Nursing in Critical Care 1999; 4(1):22-26.
17. Lasater KL, Grisanti DJ. (). Postcardiotomy psychosis: Indications and interventions. Heart & Lung 1975; 4(5):724- 729.
18. Monta M, Sethi AK, Tyagi A, Monta A. Psychological care in trauma patients. Injury 2003; 4(1):17-25.
19. Easton C, Mackenzie, F. Sensory-perceptual alterations: Delirium in the intensive care unit. Heart & Lung 1988; 17(3):229-237.
20. Patrick PK. Burnout: Job hazard for health workers. Hospitals 1979; 53(22):87-88, 90.
21. Mealer M, Shelton A, Berg B, Rothbaum B, Moss M. Increased prevalence of post traumatic stress disorder symptoms in critical care nurses. American Journal of Respiratory Critical Care Medicine 2007; 175:693-697. (epub ahead of print, Dec 2006).
22. Maloney JP. Job stress and its consequences on a group of intensive care and non-intensive care nurses. Advanced Nursing Science 1982; 4(2):31-42.
23. Gibbons C, Geller S, Glatz E. Biomedical equipment in the neonatal intensive care unit: Is it a stressor? Journal of Prenatal and Neonatal Nursing 1998; 12(3):67-73.
24. Bailey JT, Steffen SM, Grout JW. The stress audit: Identifying the stressors of ICU nursing. Journal of Nursing Education 1980; 19(6):15-25.
25. Coomber S, Todd C, Park G, Baxter P, Firth-Cozens J, Shore S. Stress in UK intensive care unit doctors. British Journal of Anaesthesia 2002; 89:873-881.
26. Firth-Cozens J. Individual and organizational predictors of depression in general practitioners. British Journal of General Practice 1998; 48:1647-1651.
27. Goodfellow A, Varnam R, Rees D, Shelley MP. Staff stress on the intensive care unit: A comparison of doctors and nurses. Anaesthesia 1997; 52:1037-1041.
28. Lawton MP, Nahemow L. Ecology and the ageing process. In C Eisdorfer and M Lawton (Eds), The psychology of adult development and aging. Washington DC: American Psychological Association; 1973.