Several countries have established recommended noise, temperature and lighting levels for hospitals. These recommendations are, however, rarely included in the specifications given to hospital designers. Further, the few studies examining these variables have reported disquieting levels.
Noise
In hospitals, it is important to distinguish between machine-generated noise capable of impairing hearing (above 85 dBA) and noise which is associated with a degradation of ambiance, administrative work and care (65 to 85 dBA).
Machine-generated noise capable of impairing hearing
Prior to the 1980s, a few publications had already drawn attention to this problem. Van Wagoner and Maguire (1977) evaluated the incidence of hearing loss among 100 employees in an urban hospital in Canada. They identified five zones in which noise levels were between 85 and 115 dBA: the electrical plant, laundry, dish-washing station and printing department and areas where maintenance workers used hand or power tools. Hearing loss was observed in 48% of the 50 workers active in these noisy areas, compared to 6% of workers active in quieter areas.
Yassi et al. (1992) conducted a preliminary survey to identify zones with dangerously high noise levels in a large Canadian hospital. Integrated dosimetry and mapping were subsequently used to study these high-risk areas in detail. Noise levels exceeding 80 dBA were common. The laundry, central processing, nutrition department, rehabilitation unit, stores and electrical plant were all studied in detail. Integrated dosimetry revealed levels of up to 110 dBA at some of these locations.
Noise levels in a Spanish hospital’s laundry exceeded 85 dBA at all workstations and reached 97 dBA in some zones (Montoliu et al. 1992). Noise levels of 85 to 94 dBA were measured at some workstations in a French hospital’s laundry (Cabal et al. 1986). Although machine re-engineering reduced the noise generated by pressing machines to 78 dBA, this process was not applicable to other machines, due to their inherent design.
A study in the United States reported that electrical surgical instruments generate noise levels of 90 to 100 dBA (Willet 1991). In the same study, 11 of 24 orthopaedic surgeons were reported to suffer from significant hearing loss. The need for better instrument design was emphasized. Vacuum and monitor alarms have been reported to generate noise levels of up to 108 dBA (Hodge and Thompson 1990).
Noise associated with a degradation of ambiance, administrative work and care
A systematic review of noise levels in six Egyptian hospitals revealed the presence of excessive levels in offices, waiting rooms and corridors (Noweir and al-Jiffry 1991). This was attributed to the characteristics of hospital construction and of some of the machines. The authors recommended the use of more appropriate building materials and equipment and the implementation of good maintenance practices.
Work in the first computerized facilities was hindered by the poor quality of printers and the inadequate acoustics of offices. In the Paris region, groups of cashiers talked to their clients and processed invoices and payments in a crowded room whose low plaster ceiling had no acoustic absorption capacity. Noise levels with only one printer active (in practice, all four usually were) were 78 dBA for payments and 82 dBA for invoices.
In a 1992 study of a rehabilitation gymnasium consisting of 8 cardiac rehabilitation bicycles surrounded by four private patient areas, noise levels of 75 to 80 dBA and 65 to 75 dBA were measured near cardiac rehabilitation bicycles and in the neighbouring kinesiology area, respectively. Levels such as these render personalized care difficult.
Shapiro and Berland (1972) viewed noise in operating theatres as the “third pollution”, since it increases the fatigue of the surgeons, exerts physiological and psychological effects and influences the accuracy of movements. Noise levels were measured during a cholecystectomy and during tubal ligation. Irritating noises were associated with the opening of a package of gloves (86 dBA), the installation of a platform on the floor (85 dBA), platform adjustment (75 to 80 dBA), placing surgical instruments upon each other (80 dBA), suctioning of trachea of patient (78 dBA), continuous suction bottle (75 to 85 dBA) and the heels of nurses’ shoes (68 dBA). The authors recommended the use of heat-resistant plastic, less noisy instruments and, to minimize reverberation, easily cleaned materials other than ceramic or glass for walls, tiles and ceilings.
Noise levels of 51 to 82 dBA and 54 to 73 dBA have been measured in the centrifuge room and automated analyser room of a medical analytical laboratory. The Leq (reflecting full-shift exposure) at the control station was 70.44 dBA, with 3 hours over 70 dBA. At the technical station, the Leq was 72.63 dBA, with 7 hours over 70 dBA. The following improvements were recommended: installing telephones with adjustable ring levels, grouping centrifuges in a closed room, moving photocopiers and printers and installing hutches around the printers.
Patient Care and Comfort
In several countries, recommended noise limits for care units are 35 dBA at night and 40 dBA during the day (Turner, King and Craddock 1975). Falk and Woods (1973) were the first to draw attention to this point, in their study of noise levels and sources in neonatology incubators, recovery rooms and two rooms in an intensive-care unit. The following mean levels were measured over a 24-hour period: 57.7 dBA (74.5 dB) in the incubators, 65.5 dBA (80 dB linear) at the head of patients in the recovery room, 60.1 dBA (73.3 dB) in the intensive care unit and 55.8 dBA (68.1 dB) in one patient room. Noise levels in the recovery room and intensive-care unit were correlated with the number of nurses. The authors emphasized the probable stimulation of patients’ hypophyseal-corticoadrenal system by these noise levels, and the resultant increase in peripheral vasoconstriction. There was also some concern about the hearing of patients receiving aminoglycoside antibiotics. These noise levels were considered incompatible with sleep.
Several studies, most of which have been conducted by nurses, have shown that noise control improves patient recovery and quality of life. Reports of research conducted in neonatology wards caring for low-birth-weight babies emphasized the need to reduce the noise caused by personnel, equipment and radiology activities (Green 1992; Wahlen 1992; Williams and Murphy 1991; Oëler 1993; Lotas 1992; Halm and Alpen 1993). Halm and Alpen (1993) have studied the relationship between noise levels in intensive-care units and the psychological well-being of patients and their families (and in extreme cases, even of post-resuscitation psychosis). The effect of ambient noise on the quality of sleep has been rigorously evaluated under experimental conditions (Topf 1992). In intensive care units, the playing of pre-recorded sounds was associated with a deterioration of several sleep parameters.
A multi-ward study reported peak noise levels at the head of patients in excess of 80 dBA, especially in intensive- and respiratory-care units (Meyer et al. 1994). Lighting and noise levels were recorded continuously over seven consecutive days in a medical intensive-care unit, one-bed and multi-bed rooms in a respiratory-care unit and a private room. Noise levels were very high in all cases. The number of peaks exceeding 80 dBA was particularly high in the intensive- and respiratory-care units, with a maximum observed between 12:00 and 18:00 and a minimum between 00:00 and 06:00. Sleep deprivation and fragmentation were considered to have a negative impact on the respiratory system of patients and impair the weaning of patients from mechanical ventilation.
Blanpain and Estryn-Béhar (1990) found few noisy machines such as waxers, ice machines and hotplates in their study of ten Paris-area wards. However, the size and surfaces of the rooms could either reduce or amplify the noise generated by these machines, as well as that (albeit lower) generated by passing cars, ventilation systems and alarms. Noise levels in excess of 45 dBA (observed in 7 of 10 wards) did not promote patient rest. Furthermore, noise disturbed hospital personnel performing very precise tasks requiring close attention. In five of 10 wards, noise levels at the nursing station reached 65 dBA; in two wards, levels of 73 dBA were measured. Levels in excess of 65 dBA were measured in three pantries.
In some cases, architectural decorative effects were instituted with no thought to their effect on acoustics. For example, glass walls and ceilings have been in fashion since the 1970s and have been used in patient admission open-space offices. The resultant noise levels do not contribute to the creation of a calm environment in which patients about to enter the hospital can fill out forms. Fountains in this type of hall generated a background noise level of 73 dBA at the reception desk, requiring receptionists to ask one-third of people requesting information to repeat themselves.
Heat stress
Costa, Trinco and Schallenberg (1992) studied the effect of installing a laminar flow system, which maintained air sterility, on heat stress in an orthopaedic operating theatre. Temperature in the operating theatre increased by approximately 3 °C on average and could reach 30.2 °C. This was associated with a deterioration of the thermal comfort of operating-room personnel, who must wear very bulky clothes that favour heat retention.
Cabal et al. (1986) analysed heat stress in a hospital laundry in central France prior to its renovation. They noted that the relative humidity at the hottest workstation, the “gown-dummy”, was 30%, and radiant temperature reached 41 °C. Following installation of double-pane glass and reflective outside walls, and implementation of 10 to 15 air changes per hour, thermal comfort parameters fell within standard levels at all workstations, regardless of the weather outside. A study of a Spanish hospital laundry has shown that high wet-bulb temperatures result in oppressive work environments, especially in ironing areas, where temperatures may exceed 30 °C (Montoliu et al. 1992).
Blanpain and Estryn-Béhar (1990) characterized the physical work environment in ten wards whose work content they had already studied. Temperature was measured twice in each of ten wards. The nocturnal temperature in patient rooms may be below 22 °C, as patients use covers. During the day, as long as patients are relatively inactive, a temperature of 24 °C is acceptable but should not be exceeded, since some nursing interventions require significant exertion.
The following temperatures were observed between 07:00 and 07:30: 21.5 °C in geriatric wards, 26 °C in a non-sterile room in the haematology ward. At 14:30 on a sunny day, the temperatures were as follows: 23.5 °C in the emergency room and 29 °C in the haematology ward. Afternoon temperatures exceeded 24 °C in 9 of 19 cases. The relative humidity in four out of five wards with general air-conditioning was below 45% and was below 35% in two wards.
Afternoon temperature also exceeded 22 °C at all nine care preparation stations and 26 °C at three care stations. The relative humidity was below 45% in all five stations of wards with air-conditioning. In the pantries, temperatures ranged between 18 °C and 28.5 °C.
Temperatures of 22 °C to 25 °C were measured at the urine drains, where there were also odour problems and where dirty laundry was sometimes stored. Temperatures of 23 °C to 25 °C were measured in the two dirty-laundry closets; a temperature of 18 °C would be more appropriate.
Complaints concerning thermal comfort were frequent in a survey of 2,892 women working in Paris-area wards (Estryn-Béhar et al. 1989a). Complaints of being often or always hot were reported by 47% of morning- and afternoon-shift nurses and 37% of night-shift nurses. Although nurses were sometimes obliged to perform physically strenuous work, such as making several beds, the temperature in the various rooms was too high to perform these activities comfortably while wearing polyester-cotton clothes, which hinder evaporation, or gowns and masks necessary for the prevention of nosocomial infections.
On the other hand, 46% of night-shift nurses and 26% of morning- and afternoon-shift nurses reported being often or always cold. The proportions reporting never suffering from the cold were 11% and 26%.
To conserve energy, the heating in hospitals was often lowered during the night, when patients are under covers. However nurses, who must remain alert despite chronobiologically mediated drops in core body temperatures, were required to put on jackets (not always very hygienic ones) around 04:00. At the end of the study, some wards installed adjustable space-heating at nursing stations.
Studies of 1,505 women in 26 units conducted by occupational physicians revealed that rhinitis and eye irritation were more frequent among nurses working in air-conditioned rooms (Estryn-Béhar and Poinsignon 1989) and that work in air-conditioned environments was related to an almost twofold increase in dermatoses likely to be occupational in origin (adjusted odds ratio of 2) (Delaporte et al. 1990).
Lighting
Several studies have shown that the importance of good lighting is still underestimated in administrative and general departments of hospitals.
Cabal et al. (1986) observed that lighting levels at half of the workstations in a hospital laundry were no higher than 100 lux. Lighting levels following renovations were 300 lux at all workstations, 800 lux at the darning station and 150 lux between the washing tunnels.
Blanpain and Estryn-Béhar (1990) observed maximum night lighting levels below 500 lux in 9 out of 10 wards. Lighting levels were below 250 lux in five pharmacies with no natural lighting and were below 90 lux in three pharmacies. It should be recalled that the difficulty in reading small lettering on labels experienced by older persons may be mitigated by increasing the level of illumination.
Building orientation can result in high day-time lighting levels that disturb patients’ rest. For example, in geriatric wards, beds furthest from the windows received 1,200 lux, while those nearest the windows received 5,000 lux. The only window shading available in these rooms were solid window blinds and nurses were unable to dispense care in four-bed rooms when these were drawn. In some cases, nurses stuck paper on the windows to provide patients with some relief.
The lighting in some intensive-care units is too intense to allow patients to rest (Meyer et al. 1994). The effect of lighting on patients’ sleep has been studied in neonatology wards by North American and German nurses (Oëler 1993; Boehm and Bollinger 1990).
In one hospital, surgeons disturbed by reflections from white tiles requested the renovation of the operating theatre. Lighting levels outside the shadow-free zone (15,000 to 80,000 lux) were reduced. However, this resulted in levels of only 100 lux at the instrument nurses’ work surface, 50 to 150 lux at the wall unit used for equipment storage, 70 lux at the patients’ head and 150 lux at the anaesthetists’ work surface. To avoid generating glare capable of affecting the accuracy of surgeons’ movements, lamps were installed outside of surgeons’ sight-lines. Rheostats were installed to control lighting levels at the nurses’ work surface between 300 and 1,000 lux and general levels between 100 and 300 lux.
Construction of a hospital with extensive natural lighting
In 1981, planning for the construction of Saint Mary’s Hospital on the Isle of Wight began with a goal of halving energy costs (Burton 1990). The final design called for extensive use of natural lighting and incorporated double-pane windows that could be opened in the summer. Even the operating theatre has an outside view and paediatric wards are located on the ground floor to allow access to play areas. The other wards, on the second and third (top) floors, are equipped with windows and ceiling lighting. This design is quite suitable for temperate climates but may be problematic where ice and snow inhibit overhead lighting or where high temperatures may lead to a significant greenhouse effect.
Architecture and Working Conditions
Flexible design is not multi-functionality
Prevailing concepts from 1945 to 1985, in particular the fear of instant obsolescence, were reflected in the construction of multi-purpose hospitals composed of identical modules (Games and Taton-Braen 1987). In the United Kingdom this trend led to the development of the “Harnes system”, whose first product was the Dudley Hospital, built in 1974. Seventy other hospitals were later built on the same principles. In France, several hospitals were constructed on the “Fontenoy” model.
Building design should not prevent modifications necessitated by the rapid evolution of therapeutic practice and technology. For example, partitions, fluid circulation subsystems and technical duct-work should all be capable of being easily moved. However, this flexibility should not be construed as an endorsement of the goal of complete multi-functionality—a design goal which leads to the construction of facilities poorly suited to any speciality. For example, the surface area needed to store machines, bottles, disposable equipment and medication is different in surgical, cardiology and geriatric wards. Failure to recognize this will lead to rooms being used for purposes they were not designed for (e.g., bathrooms being used for bottle storage).
The Loma Linda Hospital in California (United States) is an example of better hospital design and has been copied elsewhere. Here, nursing and technical medicine departments are located above and below technical floors; this “sandwich” structure permits easy maintenance and adjustment of fluid circulation.
Unfortunately, hospital architecture does not always reflect the needs of those who work there, and multi-functional design has been responsible for reported problems related to physical and cognitive strain. Consider a 30-bed ward composed of one- and two-bed rooms, in which there is only one functional area of each type (nursing station, pantry, storage of disposable materials, linen or medication), all based on the same all-purpose design. In this ward, the management and dispensation of care obliges nurses to change location extremely frequently, and work is greatly fragmented. A comparative study of ten wards has shown that the distance from the nurses’ station to the farthest room is an important determinant of both nurses’ fatigue (a function of the distance walked) and the quality of care (a function of the time spent in patients’ rooms) (Estryn-Béhar and Hakim-Serfaty 1990).
This discrepancy between the architectural design of spaces, corridors and materials, on the one hand, and the realities of hospital work, on the other, has been characterized by Patkin (1992), in a review of Australian hospitals, as an ergonomic “debacle”.
Preliminary analysis of the spatial organization in nursing areas
The first mathematical model of the nature, purposes and frequency of staff movements, based on the Yale Traffic Index, appeared in 1960 and was refined by Lippert in 1971. However, attention to one problem in isolation may in fact aggravate others. For example, locating a nurses’ station in the centre of the building, in order to reduce the distances walked, may worsen working conditions if nurses must spend over 30% of their time in such windowless surroundings, known to be a source of problems related to lighting, ventilation and psychological factors (Estryn-Béhar and Milanini 1992).
The distance of the preparation and storage areas from patients is less problematic in settings with a high staff-patient ratio and where the existence of a centralized preparation area facilitates the delivery of supplies several times per day, even on holidays. In addition, long waits for elevators are less common in high-rise hospitals with over 600 beds, where the number of elevators is not limited by financial constraints.
Research on the design of specific but flexible hospital units
In the United Kingdom in the late 1970s, the Health Ministry created a team of ergonomists to compile a database on ergonomics training and on the ergonomic layout of hospital work areas (Haigh 1992). Noteworthy examples of the success of this programme include the modification of the dimensions of laboratory furniture to take into account the demands of microscopy work and the redesign of maternity rooms to take into account nurses’ work and mothers’ preferences.
Cammock (1981) emphasized the need to provide distinct nursing, public and common areas, with separate entrances for nursing and public areas, and separate connections between these areas and the common area. Furthermore, there should be no direct contact between the public and nursing areas.
The Krankenanstalt Rudolfsstiftung is the first pilot hospital of the “European Healthy Hospitals” project. The Viennese pilot project consists of eight sub-projects, one of which, the “Service Reorganization” project, is an attempt, in collaboration with ergonomists, to promote functional reorganization of available space (Pelikan 1993). For example, all the rooms in an intensive care unit were renovated and rails for patient lifts installed in the ceilings of each room.
A comparative analysis of 90 Dutch hospitals suggests that small units (floors of less than 1,500 m2) are the most efficient, as they allow nurses to tailor their care to the specifics of patients’ occupational therapy and family dynamics (Van Hogdalem 1990). This design also increases the time nurses can spend with patients, since they waste less time in changes of location and are less subject to uncertainty. Finally, the use of small units reduces the number of windowless work areas.
A study carried out in the health administration sector in Sweden reported better employee performance in buildings incorporating individual offices and conference rooms, as opposed to an open plan (Ahlin 1992). The existence in Sweden of an institute dedicated to the study of working conditions in hospitals, and of legislation requiring consultation with employee representatives both before and during all construction or renovation projects, has resulted in the regular recourse to participatory design based on ergonomic training and intervention (Tornquist and Ullmark 1992).
Architectural design based on participatory ergonomics
Workers must be involved in the planning of the behavioural and organizational changes associated with the occupation of a new work space. The adequate organization and equipping of a workplace requires taking into account the organizational elements that require modification or emphasis. Two detailed examples taken from two hospitals illustrate this.
Estryn-Béhar et al. (1994) report the results of the renovation of the common areas of a medical ward and a cardiology ward of the same hospital. The ergonomics of the work performed by each profession in each ward was observed over seven entire workdays and discussed over a two-day period with each group. The groups included representatives of all occupations (department heads, supervisors, interns, nurses, nurses’ aides, orderlies) from all the shifts. One entire day was spent developing architectural and organizational proposals for each problem noted. Two more days were spent on the simulation of characteristic activities by the entire group, in collaboration with an architect and an ergonomist, using modular cardboard mock-ups and scale models of objects and people. Through this simulation, representatives of the various occupations were able to agree on distances and the distribution of space within each ward. Only after this process was concluded was the design specification drawn up.
The same participatory method was used in a cardiac intensive-care unit in another hospital (Estryn-Béhar et al. 1995a, 1995b). It was found that four types of virtually incompatible activities were conducted at the nursing station:
- care preparation, requiring the use of a drain-board and sink
- decontamination, which also used the sink
- meeting, writing and monitoring; the area used for these activities was also sometimes used for the preparation of care
- clean-equipment storage (three units) and waste storage (one unit).
These zones overlapped, and nurses had to cross the meeting-writing-monitoring area to reach the other areas. Because of the position of the furniture, nurses had to change direction three times to get to the drain-board. Patient rooms were laid out along a corridor, both for regular intensive care and highly intensive care. The storage units were located at the far end of the ward from the nursing station.
In the new layout, the station’s longitudinal orientation of functions and traffic is replaced with a lateral one which allows direct and central circulation in a furniture-free area. The meeting-writing-monitoring area is now located at the end of the room, where it offers a calm space near windows, while remaining accessible. The clean and dirty preparation areas are located by the entrance to the room and are separated from each other by a large circulation area. The highly intensive care rooms are large enough to accommodate emergency equipment, a preparation counter and a deep washbasin. A glass wall installed between the preparation areas and the highly intensive care rooms ensures that patients in these rooms are always visible. The main storage area was rationalized and reorganized. Plans are available for each work and storage area.
Architecture, ergonomics and developing countries
These problems are also found in developing countries; in particular, renovations there frequently involve the elimination of common rooms. The performance of ergonomic analysis would identify existing problems and help avoid new ones. For example, the construction of wards comprised of only one- or two-bed rooms increases the distances that personnel must travel. Inadequate attention to staffing levels and the layout of nursing stations, satellite kitchens, satellite pharmacies and storage areas may lead to significant reductions in the amount of time nurses spend with patients and may render work organization more complex.
Furthermore, the application in developing countries of the multi-functional hospital model of developed countries does not take into account different cultures’ attitudes toward space utilization. Manuaba (1992) has pointed out that the layout of developed countries’ hospital rooms and the type of medical equipment used is poorly suited to developing countries, and that the rooms are too small to comfortably accommodate visitors, essential partners in the curative process.
Hygiene and Ergonomics
In hospital settings, many breaches of asepsis can be understood and corrected only by reference to work organization and work space. Effective implementation of the necessary modifications requires detailed ergonomic analysis. This analysis serves to characterize the interdependencies of team tasks, rather than their individual characteristics, and identify discrepancies between real and nominal work, especially nominal work described in official protocols.
Hand-mediated contamination was one of the first targets in the fight against nosocomial infections. In theory, hands should be systemtically washed on entering and leaving patients’ rooms. Although initial and ongoing training of nurses emphasizes the results of descriptive epidemiological studies, research indicates persistent problems associated with hand-washing. In a study conducted in 1987 and involving continuous observation of entire 8-hour shifts in 10 wards, Delaporte et al. (1990) observed an average of 17 hand-washings by morning-shift nurses, 13 by afternoon-shift nurses and 21 by night-shift nurses.
Nurses washed their hands one-half to one-third as often as is recommended for their number of patient contacts (without even considering care-preparation activities); for nurses’ aides, the ratio was one-third to one-fifth. Hand-washing before and after each activity is, however, clearly impossible, in terms of both time and skin damage, given the atomization of activity, number of technical interventions and frequency of interruptions and attendant repetition of care that personnel must cope with. Reduction of work interruptions is thus essential and should take precedence over simply reaffirming the importance of hand-washing, which, in any event, cannot be performed over 25 to 30 times per day.
Similar patterns of hand-washing were found in a study based on observations collected over 14 entire workdays in 1994 during the reorganization of the common areas of two university hospital wards (Estryn-Béhar et al. 1994). In every case, nurses would have been incapable of dispensing the required care if they had returned to the nursing station to wash their hands. In short-term-stay units, for example, almost all the patients have blood samples drawn and subsequently receive oral and intravenous medication at virtually the same time. The density of activities at certain times also renders appropriate hand-washing impossible: in one case, an afternoon-shift nurse responsible for 13 patients in a medical ward entered patients’ rooms 21 times in one hour. Poorly organized information provision and transmission structures contributed to the number of visits he was obliged to perform. Given the impossibility of washing his hands 21 times in one hour, the nurse washed them only when dealing with the most fragile patients (i.e., those suffering from pulmonary failure).
Ergonomically based architectural design takes several factors affecting hand-washing into account, especially those concerning the location and access to wash-basins, but also the implementation of truly functional “dirty” and “clean” circuits. Reduction of interruptions through participatory analysis of organization helps to make hand-washing possible.