Virology and Infection Control
The Centers for Disease Control estimates that approximately 1.5-2 million nosocomial (hospital-acquired) infections occur in hospitalized patients every year. These infections complicate about 5-10% of all hospitalizations and may account for as much as $2 billion per year in excess hospital costs. Therefore, nosocomial infections are a significant cause of morbidity and, in some cases, mortality in hospitalized patients. Infection control programs generally focus on bacterial nosocomial infections primarily because diagnostic bacteriologic methods are widely available. Many infection control programs have functioned well using these readily available bacteriologic data. Easy access to this data has allowed the infection control practitioner to be maximally effective by evaluating patients prospectively and hopefully, instituting measures for prevention and control of hospital infections.
Viral infections, however, may constitute a significant proportion of nosocomial infections in hospitalized patients. In a prospective study, more than 5% of all nosocomial pathogens were viruses (7) , which probably represents an underestimate of the real magnitude of the problem.
Some nosocomial viral infections are seasonal whereas others occur throughout the year. Seasonal infections are generally caused by respiratory viruses such as respiratory syncytial virus, influenza virus, adenovirus, and some strains of parainfluenza virus. Usually, viruses are first identified in the community and then shortly afterward are seen in the hospital as nosocomial pathogens. Pediatric patients (3, 9, 11) and patients in closed or semiclosed settings such as extended care facilities (4) and psychiatric services (7) appear to be at highest risk for hospital-acquired viral infections. Infections encountered most frequently in pediatrics and longterm care settings are respiratory infections caused by respiratory syncytial virus, influenza virus, and adenovirus. Hospitalized adults in acute care settings appear to be more susceptible to reactivations of endogenous herpesvirus infections (7) .
Although transmission of nosocomial viral infections has not been well studied, available evidence suggests that patient care personnel play a role in virus transmission, especially with respiratory viruses such as respiratory syncytial virus, influenza virus, and adenovirus (3, 6, 9, I 1). Other viral infections that occur throughout the year may also involve patient care personnel in their chain .of transmission to hospitalized patients. Transmission of agents such as the herpesviruses probably occurs less frequently in hospital settings but has been seen in dental settings ( An unusual isolate from infected bone uate the data. When an institution has no viral diagnostic facilities, the infection control practitioner must use less precise methods to approach viral diagnosis. Serologic methods using acute and convalescent patient sera are helpful in establishing many viral diagnoses retrospectively but generally are useless for the infection control program which functions best using data generated prospectively. The epidemiologic approach to viral diagnosis may also be used. This involves patient assessment based on viruses currently circulating in the community and evaluating patient factors such as age, clinical signs, and symptoms to arrive at this "best guess." This approach is time consuming and, like serologic methods, not always precise. As rapid viral diagnostic methods improve and become more widely available, prevention and control of nosocomial viral infections will become more effective and efficient for the infection control practitioner.
On a more practical level, the infection control practitioner is often faced with decisions that require assistance from the virology laboratory. For example, employee exposures to chickenpox often require testing certain employees for varicella immunity. A rapid, diagnostic test for varicella immunity is available and may help to avoid the unnecessary expense of sending employees home from work who are "susceptible" by negative varicella history when in fact they are immune (12) . In other situations, rapid diagnostic tests for respiratory syncytial virus and influenza virus can help determine appropriate isolation precautions for patients with possible viral respiratory diseases.
The risk of nosocomial viral infections is determined by the transmissibility of the agent and the susceptibility of the host (Table 1) (8) . In contrast to personnel who have direct patient contact, laboratory personnel appear to be at relatively low risk of acquiring viral infections from clinical laboratory specimens because the type of contact that occurs in the laboratory is different from the contact with secretions and body fluids that occurs in the patient care setting. For example, respiratory virus transmission from laboratory specimens is probably a rare event. The greater risk of virus transmission in the laboratory is probably much the same as it is in the clinical setting, i.e., person-to-person transmission via small particle aerosol transmission or close contact transmission from infected individuals (e.g., fellow workers). An exception is laboratory transmission of hepatitis B which is related to blood contact and not to patient contact. However, the microbiology laboratory worker would probably be at relatively low risk for hepatitis B because blood contact is usually less than for other laboratory personnel such as blood bank and chemistry laboratory technologists.
Not all virus infections are preventable by vaccination. Therefore, prevention and control of hospital-ac- probably not a high priority for this group. To date no evidence supports an increased risk of acquiring viral diseases in the workplace during pregnancy. Therefore, pregnant personnel need not be transferred to other work areas during pregnancy (10) . Finally, incorporating a virology laboratory as a component of the hos-pital's infection control program for patients is an exciting prospect. As rapid viral diagnostic tests are developed and as these tests become increasingly specific (2), the infection control practitioner will be able to incorporate surveillance of viral infections into programs for prevention and control of viral infections.