The subtypes of influenza A virus are classified on the basis of their surface antigens, called hemagglutinins (H) and neuraminidases (N). There are three hemagglutinins (H1, H2, and H3) and two neuraminidases (N1 and N2). Immunity to these antigens reduces the likelihood of infection and reduces the severity of illness if it does occur. However, antigenic drift and antigenic shift (i.e., subtle and marked changes, within a subtype, respectively) make long-lasting immunity difficult to achieve. Of the two antigenic changes, antigenic drift is the more gradual, with the H and N subtypes retaining some similarity as changes occur. Antigenic shift is a more abrupt change in H or N subtype, which occurs at longer intervals (e.g., approximately every 10 or more years). When the marked changes of antigenic shift occur, infection or vaccination with one strain may not necessarily induce immunity to distant strains, even though they are of the same subtype. Influenza B is more antigenically stable than influenza A and undergoes antigenic drift, but not the major structural changes of antigenic shift.

Effectiveness of influenza vaccine is determined by the closeness of the vaccine-induced antibody to the H and N surface antigens of influenza A and B. Influenza vaccine loses its protective effects as more major shifts of influenza H and N surface antigens or subtypes occur.

The nomenclature for influenza strains is a useful way to better understand that particular strain. The standard way of describing strains includes the serotype, host of origin (human unless otherwise specified), geographic origin, strain number, year of isolation, and the H and N designation.

Recommendation for trivalent influenza vaccine (TIV) components is based on surveillance data related to epidemiology and antigenic characteristics (4), serological responses to previous vaccines, and the availability of candidate strains and reagents (3,4). For the 2010-2011 influenza season, TIV will include a component of the recently emerged H1N1 strain (4). For the H1N1 component, an A/California7/2009-like virus (the pandemic strain) will replace the Brisbane/59/2007 strain. For H3N2, the A/Perth/16/2009-like virus will replace A/Brisbane/10/2007. For type B, the B/Brisbane/60/2008-like strain (the same as 2009-2010) will be used.

The so-called high-risk groups for influenza and its complications include older persons (i.e., 65 years of age or older), very young children, and persons of any age with certain underlying health conditions, who are at increased risk for hospitalization, death, and other complications. During major epidemics, hospitalization rates for high-risk persons may increase two- to fivefold. Despite this, only about 30% of people aged 65 years or older are vaccinated
with influenza vaccine every year (3). Many outbreaks are reported from nursing homes and long-term care facilities (5, 6 and 7), in part because of underuse of vaccine in these vulnerable, closed populations (see Chapter 97). In addition to underuse of vaccine, many high-risk people fail to develop a protective antibody response to vaccination (5,6). Outbreaks have also occurred in general hospitals, psychiatric units, and medical and pediatric services (3). This underuse of vaccine is a major contributor to outbreaks of influenza in healthcare facilities with associated morbidity and, on occasion, mortality.

As the population ages, the risk of seasonal influenza death increases. Thompson et al. reported that the death rate from influenza rose markedly in the 1990s, and in 2001 it exceeded the number of deaths due to acquired immunodeficiency syndrome (AIDS) (7). Annual estimates of influenza-associated deaths increased significantly between the 1976-1977 and 1998-1999 seasons, with a mean of 20,000 and 36,000 deaths, respectively. Ninety percent of respiratory and circulatory deaths occurred in persons aged 65 years or older. Since its emergence in the 1960s, type A (H3N2) epidemics have caused approximately 400,000 deaths in the United States alone, and >90% of these deaths have occurred in people older than 65 years (8). Prior to the reemergence of influenza A (H1N1) in 2009, the H3N2 had the most severe overall impact (2,9). An unexpected trend with the 2009 pandemic influenza H1N1 strain, on the other hand, was a higher attack rate in people younger than 65 years, outside the traditional high-risk groups (10). Early data from the 2009 pandemic influenza A (H1N1) outbreak in Mexico indicated that attack rates among persons aged 65 years or older were lower than those in other age groups and that anti-influenza A antibodies that cross-react with 2009 H1N1 could be detected in up to one third of healthy adults aged older than 60 years (11,12,13).

Despite this new trend with pandemic influenza A (H1N1), it is important to keep in mind the impact of influenza on elderly and chronically ill patients. Gross et al. characterized two typical nursing home outbreaks of seasonal influenza A (5,6). One began in November, peaked in February, and ended in April. The outbreak progressed slowly and was complicated by concurrent infections with respiratory syncytial virus, parainfluenza virus, and Mycoplasma pneumoniae. The patient population in this case had an immunization rate of 59%, affording it some degree of herd immunity. The authors contrast the pattern of slow spread in this closed, partially immunized population to the more explosive outbreaks described in open, unimmunized populations (e.g., acute care settings and psychiatric services) (8,9). In the outbreak, influenza illness was significantly more common in the unvaccinated group, as was mortality (17.7% in the unvaccinated group and 7.2% in the vaccinated group). When controlled for sex and severity of illness, influenza vaccine reduced mortality by 59% in this closed, partially vaccinated population (6).

Patriarca et al. developed a useful model to project morbidity, mortality, and costs associated with type A influenza illness in nursing homes (14). The model used demographics similar to the real world of long-term care: 100 residents and a 60% rate of vaccination in the fall of the prior year. In this model, the combination of previous vaccination and amantadine during outbreaks was associated with significantly fewer cases compared with vaccine alone, probably because of the <100% efficacy of vaccine. The authors predicted an increase in herd immunity as more patients were vaccinated beyond the 60% receiving it initially, and when 70% received vaccine, the risk of an outbreak approached zero. They concluded that influenza control programs in nursing homes are beneficial, clinically sound, and cost-effective and have a modest increase in program costs with the addition of amantadine. More recent studies have shown that vaccination of both healthcare providers (HCPs) and patients is associated with fewer deaths among nursing home patients (15) and elderly hospitalized patients (16). These data support the Advisory Committee on Immunization Practices’ (ACIP) target of 90% vaccine use in populations at risk (3) and the 2005 requirement from Centers for Medicare and Medicaid Services (CMS) that nursing homes participating in the CMS programs offer all residents influenza and pneumococcal vaccines and document the results (17). Each resident is to be vaccinated unless contraindicated medically, the resident or a legal representative refuses vaccination, or the vaccine is not available because of shortage. This information is to be reported as part of the CMS Minimum Data Set, which tracks nursing home health parameters (18,19).

Typical influenza illness in the adult is characterized by sudden onset of fever, myalgia, sore throat, headache, retroorbital pain, and nonproductive cough. Unlike most other viral respiratory infections, influenza causes myalgias and other constitutional symptoms that can last a week or more. However, the sensitivity and the positive predictive value of fever, cough, and/or other symptoms for the diagnosis of influenza virus infection in severely ill or hospitalized patients are lower than those in the community (20). The use of these common symptoms for treatment decisions and infection control management will probably be insufficient to contain a healthcare-associated outbreak, because many influenza cases will remain unidentified.

Some patients with influenza A may develop additional complications of primary influenza pneumonia or secondary bacterial pneumonia most often resulting from Streptococcus pneumoniae or Staphylococcus aureus (21). These complications are not associated with influenza B infection, which is usually a milder illness.

Influenza in adults is first and foremost a respiratory disease. The term “intestinal flu” in adults is generally a misnomer (22). The illness in children, on the other hand, may have a major gastrointestinal component or may mimic sepsis (22,23). In an influenza A outbreak on a pediatrics ward, 7 of 12 infected children (58%) developed pulmonary infiltrates and 5 of the 7 went on to develop a secondary bacterial pneumonia. In the young infant, influenza may mimic sepsis with fever and no localizing findings (23).

Control of influenza requires herd immunity, which requires that large numbers of people in a particular group at risk be immune to infection (9,14). There are two approaches to reduce the impact of influenza infection: inactivated influenza vaccine (immunoprophylaxis) and antiviral drugs (chemoprophylaxis). Antiviral drugs are a useful adjunct when herd immunity is not present because of underuse of vaccine and/or inadequate protective antibody response to vaccination (36).