The delivery of babies in hospital is a relatively recent phenomenon: it evolved during the last half of the 20th century. Before then, maternity hospitals were not considered safe because of relatively high rates of death. It was not until the observations of Oliver Wendell Holmes and Ignaz Semmelweis that puerperal fever was thought to be a communicable disease transmitted from health-care workers to patients.

Semmelweis hypothesized that puerperal fever was spread by the hands of physicians and midwives. He noted that at the Vienna Lying-In Hospital the death rate was almost 10% for women who delivered in Division I, compared with 3% for women in Division II (3). Semmelweis' investigation determined that food, water, ventilation, or socioeconomic class did not account for these discrepancies. However, he observed that patients with prolonged labor were at increased risk and children born to infected mothers were also more likely to become ill. Conversely, women whose babies were born outside the hospital were less likely to develop fever. Semmelweis also noted that infection in Division I occurred sporadically and in clusters, whereas in Division II, no clustering occurred.

Figure

Figure. . Division I rates of death, April-December 1847

His analysis revealed that medical students, who were responsible for deliveries in Division I, often performed autopsies before assisting in deliveries, while midwives, who worked in Division II, did not. He theorized that disinfecting hands could prevent transmission of infection from a diseased cadaver to a pregnant patient (3). Therefore, on May 15, 1847, he required all medical students to wash their hands with chlorinated lime before assisting in deliveries, which resulted in a dramatic outcome (Figure).

Florence Nightingale made many observations about hospital design based on her experiences during the Crimean War. Her ideas regarding a sanitary environment meant rejecting the 18th-century concept of long hospital corridors. She commented that double wards were objectionable on every account primarily because they prevented nurses from being able to assess all their patients at the same time (4). She also observed that open windows interfered with the ventilation of hospital wards and allowed air from the wards to pass into the corridors. Nightingale believed that respiratory secretions were potentially dangerous, especially among the sick. Therefore, she said that depriving patients of appropriate ventilation "is nothing but manslaughter under the garb of benevolence" (5). Finally, she believed the sick should be isolated and that hospitals should be no more than two stories high. It was her contention that taller buildings interfered with sunlight and ventilation.

In 1875, after a large donation from Johns Hopkins, plans were developed to build a hospital in Baltimore, Maryland. Of five construction plans, two were substantially influenced by infection control. Norton Folsom, superintendent of Massachusetts General Hospital, believed that the hospital should be well ventilated and provide an isolation ward "for the occasional case so contagious or unpleasantly smelly that it cannot remain under the same roof with others" (6). A New York physician, Stephen Smith, believed that contagious patients should be separated from each other. In his plan, Smith classified patients into one of four categories: acutely contagious cases; uncomplicated infections and fever cases; acute medical and surgical cases; and completely noninfectious chronic disease cases. Further, he suggested that properly separating patients, with appropriate ventilation, was the most important facet of hospital planning.

In 1920, Asa Bacon of Chicago's Presbyterian Hospital noted that hospitals are hotels for sick people. One disgruntled patient commented to him following his discharge, "When I return, put me in a closet rather than in the ward!" (7). Bacon concluded that the most efficient hospital would contain all private rooms. His vision included a private toilet and lavatory in each room; a central kitchen and serving station; central linen supply instead of linen rooms on each floor; elimination of long corridors; dumbwaiters direct from central supply rooms; and pneumatic tubes to carry written requisitions. Bacon proposed these innovations 80 years ago, and today we take them for granted as integral to the modern medical center.

The Royal Infirmary in Aberdeen, Scotland, was specifically designed to prevent hospital-acquired infections in the surgical unit (8). Based on the recommendations of the Infirmary's Department of Bacteriology, no room had more than four beds, and 41% of the rooms were private or had only one bed. In addition, 10 private rooms surrounding central nurses' station were designed for "intensive nursing care." All rooms were mechanically ventilated, and 75% of the air was cleaned and then recirculated. The design also included an ISPIN (isolation, pre- and postoperative care including intensive nursing) unit with all private rooms placed between the operating room and the wards. This allowed "clean" surgeries to be separated from those with the potential of infection.

To minimize the risk for infection in hospitalized patients, infection control professionals should participate in facility design from a building's inception (9). This allows for identifying potential infection control issues early and provides an opportunity to design solutions prospectively. Infection control professionals also play an important role in educating architects, engineers, and construction workers about potential infection control risks and appropriate methods for reducing them. Because infection control professionals are often the only personnel with a clinical background working on the construction project, they need to visit the construction site frequently and completely understand the extent of the project. Because of the profound implications of inadequate oversight by infection control professionals, these expectations should be included in the hospital building contract (10). In addition, if the policies and procedures set forth by the infection control team are consistently ignored, the institution should fine the contractors.

As part of the planning process for constructing a new facility, an infection control risk assessment should be conducted to determine the potential risk for transmission of microorganisms within the hospital. In general, the risks can be classified as infections transmitted by air, water, or environment. The association between construction and the development of aspergillosis in immunocompromised patients has been known for decades (11), as has the association of hospital-acquired legionellosis and potable water (12). More recently, contamination of the hospital environment has been associated with transmission of Clostridium difficile (13), methicillin-resistant Staphylococcus aureus (MRSA) (14), and vancomycin-resistant enterococci (VRE) (15).

Aspergillus spp. are ubiquitous fungi, typically found in soil, decaying vegetation, and dust. Aspergillus spores are easily suspended in the air and survive for prolonged periods. Because of their size, they are easily inhaled, which can lead to invasive infection of both the upper and lower respiratory tracts in a susceptible host.

Epidemiologic evidence clearly correlates hospital acquisition of aspergillosis with Aspergillus spore counts (16). Therefore, installation of HEPA filters is essential in locations housing patients at high risk. While achieving a spore-free environment is an admirable goal, minimal concentrations of fungal spores in the environment are considered safe. In our new hospital, Northwestern Memorial, in Chicago, Illinois, the entire building is HEPA filtered because of the increasing number of immunosuppressed patients. Before opening the hospital, we performed air sampling to ensure the efficacy of the HEPA filter system and found that the composite fungal concentration and the Aspergillus spp. spore count were consistent with a highly filtered environment (Table).

Prevention of aspergillosis is particularly important for patients undergoing solid organ and bone marrow transplantation. In bone marrow transplant units, the air should be HEPA filtered with the air pressure in the room positive in relation to the corridor. In addition, rooms should be tightly sealed, especially around windows, and the air exchange rate should be high (>15 per hour) (17).

Proper health-care facility design can prevent hospital transmission of TB to patients and health-care workers. Ultimately, the interventions necessary to prevent hospital transmission of TB depend on the incidence of this disease in the community and have been published in detail (18). The Centers for Disease Control and Prevention recommends that patients requiring isolation for TB be placed in a room with negative airflow. These rooms should have frequent air exchanges (>12 per hour), and the air should be exhausted to the outside without recirculation. Doors to the rooms should be self-closing, and the walls, windows, ceiling, floor, and penetrations well sealed. These rooms should be monitored to ensure that they remain under negative pressure when occupied by a TB patient.

Infection control professionals play a substantial role in determining the appropriate location of negative-airflow rooms when a hospital is being designed. Ideally, they should be located in areas where patients at high risk will be cared for (e.g., emergency department, recovery room, bronchoscopy suite, ambulatory clinic, medical units).

Legionella is an important cause of community- and hospital-acquired lower respiratory tract infections. Person-to-person transmission of this organism has not been documented. Rather, infection is exclusively acquired from the environment, and hospital acquisition is well recognized (12,19,20). The most consistent observation about health-care acquired legionellosis is its association with potable water. The highest concentrations of the organism are found in hot-water storage tanks, cooling towers, and condensers.

Effective methods for disinfecting the hospital water supply include chlorination, thermal eradication, UV light, and metal ionization (16). At our new medical center, we elected to install a copper-silver ionization system. Despite the potential presence of Legionella in the water supply, routine culturing of water in the absence of proven or suspected hospital transmission is not recommended (21).

Hospital design should ensure that patients, especially immunocompromised patients, are at no greater risk for infection within the hospital than outside. Because the microbial flora of a health-care facility can be influenced by its design, infection control professionals play a major role in this aspect.

Bacteria on hospital floors predominantly consist of skin organisms, e.g., coagulase-negative staphylococci, Bacillus spp., and diptheroids (22); S. aureus and Clostridium spp. can also be cultured. However, infection risk from contaminated floors is small. Gram-negative bacteria are rarely found on dry floors, but may be present after cleaning or a spill. Nevertheless, these organisms tend to disappear as the surface dries (23).

The survival of microbes on carpeting, however, is different: they are present in larger numbers on this surface and they pose a greater risk for infection. Therefore, carpets should be vacuumed daily and periodically steam cleaned. Carpeting should be avoided in high-risk areas because the cleaning process may aerosolize fungal spores. Regardless of the flooring chosen, it should be easily cleanable and water resistant (9).

In general, pathogenic microorganisms do not readily adhere to walls or ceilings unless the surface becomes moist, sticky, or damaged (23). Little evidence exists that walls and ceilings are a major source for hospital infection. Wall coverings should be fluid resistant and easily cleaned, especially in areas where contact with blood or body fluids may occur (e.g., laboratories, operating rooms). Finishings around plumbing fixtures should be smooth and water resistant (9). In addition, pipe penetrations and joints should be tightly sealed. Acoustical tiles should be avoided in high-risk areas because they may support microbial growth when wet. False ceilings may harbor dust and pests that may contaminate the environment if disturbed, so should be avoided in high-risk areas unless adequately sealed. Ideally, walls and ceilings should have a smooth, impervious surface that is easy to clean with minimal likelihood of dust accumulation.

Infection control professionals are often consulted to recommend appropriate finishes and fixtures. The best finishes are durable and easy to clean. Surfaces that are porous or textured may be difficult to clean and might therefore harbor potentially pathogenic microbes (10). Furniture is thought to be a minor infection risk, but prolonged survival of VRE on chairs (24) and other environmental surfaces has been documented (25). MRSA and VRE have also been recovered from privacy curtains, scrub suits, and plastic aprons (26); whether contamination of these surfaces poses a risk to patients is unknown. However, survival of these pathogens for even a short time increases the possibility of their being acquired by patients or health-care workers and spread from one person to the next.

Handwashing is the single most important method to prevent hospital infections. Each patient room, examination room, and procedure room needs at least one sink (9). Optimally, it should be as close to the entrance of the room as possible and be large enough to prevent splashing. Too shallow a sink may cause contamination of hands by bacteria residing in the drain; this was linked to a hospital outbreak of multidrug-resistant gram-negative bacilli (27). Each sink should be equipped with a hands-free control, soap dispenser, and paper towel holder. Access to examination gloves and a trash receptacle should be readily available. We installed a dedicated sink at the entrance to every patient room to facilitate handwashing by health-care workers.

The design of health-care facilities has undergone substantial changes in large part because patients with impaired host defenses now represent an increasing proportion of hospitalizations. As a result, both design and renovation of these facilities present unique challenges and opportunities for infection control professionals, who are often the only clinical staff associated with construction projects. Early involvement in the process can make appropriate communication easier and protect patient safety. Ultimately, while time-consuming, participation in hospital design, construction, and renovation can serve as another marker of how infection control professionals improve the quality of patient care.