Introduction
Influenza viruses cause acute respiratory infection and can affect people of any age. The elderly, the very young, and the chronically ill are particularly susceptible to the complications of this infection, which may include otitis media, pneumonia, cardiovascular failure, and death. Influenza viruses possess the ability to spread rapidly and they cause seasonal epidemics across the globe. They are unpredictable and able to adapt and cope with selective immunologic and drug pressures.[1,2]

Influenza viruses have unique genomic properties, providing the basis for constant antigenic variability. Relatively minor changes in the surface glycoproteins—hemaglutinin (H) and neuraminidase (N)—are called antigenic “drift”. Marked changes in one or both surface glycoproteins are called antigenic “shift”. Influenza A viruses are endowed with more pronounced antigenic variability than influenza B and C viruses. Introducing new antigenic variants of influenza A into a population with no pre-existing immunity can lead to a pandemic.[1]

Yearly influenza vaccination with current circulating strains is recommended.[3] The influenza virus strains to compose the season’s vaccine are determined around February by the World Health Organization. The 1997–98 influenza season was distinguished by the fact that A/H3N2/Sydney/05/97-like virus was not included in the available vaccine because this novel drift was not anticipated as one of the circulating strains in North America. Vaccine effectiveness is dependent, in part, on the match between the vaccine and circulating strains. Even when vaccine and epidemic strains match closely, outbreaks can occur among vaccinated groups.

A Canada-wide survey of long-term care facilities in 1991 found that 78.5% of residents were vaccinated with influenza vaccine and only 1% with pneumococcal vaccine.[4] In BC, influenza vaccine is provided free of charge to people aged 65 years and older, the institutionalized, the chronically ill, and other groups at high risk of complications of this infection.[3] This study describes the impact of influenza A outbreaks as detected by our surveillance system and assesses the effectiveness of vaccination and management of outbreaks in health-care facilities during the 1997–98 influenza season.

Methods
Active influenza surveillance has been in place in British Columbia since 1976 and is routinely conducted from October to May each year. It includes weekly reports and laboratory specimens collected from a network of sentinel physicians, acute-care hospitals, and health unit reports of influenza-like illness (ILI) activity in health-care facilities and schools across BC.[5] All parties report to the BC Centre for Disease Control (BCCDC).

In BC, health-care facilities are asked to report all suspect ILI outbreaks among residents. The 1997–98 ILI case definition was “acute febrile respiratory illness (fever and/or chills) characterized by one or more of the following: cough, sore throat, arthralgia, myalgia, or prostration that in the opinion of the attending physician could be due to influenza virus”. In 1997–98 we developed a questionnaire to collect information regarding each ILI outbreak reported to the BCCDC. Information about the setting, number, and age group of occupants, immunization status of residents and staff, symptoms, management of the outbreak, and outcome was recorded. Only deaths associated with the respiratory infection or its complication for the duration of the outbreak were computed. The data was stored in Microsoft Excel, relative risk was calculated using EpiInfo 6.0, and linear regression analysis was performed using SPSS.

Health-care facilities were further classified as an extended-care facility if part of a hospital ward or as a long-term care facility if nursing homes with intermediate level of care in the community. All health-care facilities in the province had conducted influenza vaccination programs in early November. The vaccine contained an A/Wuhan/ 359/95(H3N2)-like strain, an A/Bayern/7/95(H1N1)-like strain, and a B/Beijing/184/93-like strain. Each attending physician ordered the currently recommended vaccine for all residents under his or her care. The residents who were not vaccinated were those who refused or whose relatives had not given consent.

Health-care facilities are encouraged to implement influenza outbreak control measures as recommended by the BCCDC Communicable Disease Policy Manual. Testing for virus strain identification was recommended. Nasal, nasal-pharyngeal, or throat swabs for virus culture and direct fluorescence assay or serum for antibody tests were conducted. All samples were sent to the provincial laboratory in the Virology Department of the BCCDC.

Results

From 20 October 1997 to 31 May 1998, the laboratory confirmed 456 influenza cases. Of those, greater than 99% were influenza A (two cases of influenza B were detected). Influenza A/H3N2/ Sydney/05/97-like strain was the predominant pathogen circulating in the province.

The 44 sentinel physicians (less than 1% of practising doctors in BC) documented over 3000 patient visits due to ILI. The percentage of visits due to ILI is represented in Figure 1.[Figure not available online] The distribution shows significant delay and increased ILI activity during the 1997–98 season compared to the previous seven seasons.

One hundred twenty-five schools reported absenteeism greater than 10%, representing over 5000 ill children. Median attack rate at time of report was 14%. ILI outbreaks in schools preceded the ILI outbreaks in health-care facilities by 2 weeks. Peak ILI activity in schools coincided with the peak ILI activity reported by sentinel physicians.

The impact of influenza A/Sydney in the institutionalized elderly population in BC
Outbreaks of ILI were identified in 62 health-care facilities, representing a 6-fold increase of activity compared to the previous year. Institution size ranged from 20 to 300 residents. Sixteen outbreaks occurred in extended-care facilities and 46 in long-term care facilities. A total of 1898 (31%) of 6108 residents were reported affected. There were 141 known deaths and 109 known hospitalizations associated with these outbreaks.

Influenza A was laboratory-confirmed in 54 facilities; of those, 13 were further identified as influenza A/H3N2/ Sydney/05/97-like virus. Most ILI outbreaks in health-care facilities occurred within 8 weeks between February and March 1998.

Twenty-eight (46%) health-care facilities reported using amantadine for outbreak management. Increased risk of death was observed in health-care facilities that reported no amantadine use (RR 1.48, CI 1.06<;RR<;2.05). In extended-care facilities, increased risk of death and illness was observed in institutions that reported no amantadine use (RR 5.95; CI 2.52<;RR<;14.02 and RR 1.50; CI 1.25<;RR<;1.79, respectively).

A total of 76.5% of the residents in the health-care facilities were reported vaccinated. Non-vaccinated residents were twice as likely to become ill, to be hospitalized due to ILI complications, or to die. Overall vaccine efficacy in preventing illness was 46%.

Approximately one-third (23) of the health-care facilities did not provide information regarding staff immunization or illness. Staff vaccination rate was low, with a median of 24% vaccinated. Median attack rate reported among staff was 15%, but up to 55% in some institutions. Staff worked overtime or a replacement was employed to substitute the sick.[6]

Discussion
A novel drift of influenza A/H3N2 was first identified in Sydney, Australia, in June 1997. Unfortunately, the decision regarding virus strains to compose the season’s vaccine was determined in February 1997. In Canada, the A/Sydney-like strain was first isolated in Montreal, in September 1997.[7] Four months later, the virus had reached BC and was causing outbreaks in the community. Also in 1997, a novel shift of influenza A/H5N1 was identified in Hong Kong. A bird strain of influenza virus was able to make the jump to infect humans. The observed mortality rate was approximately 30%. Fortunately, the virus had not yet developed the ability to spread from human to human, but if it had, a pandemic of influenza would have likely occurred. If such a deadly virus were able to spread through the world’s crowded communities, the consequence would be a global disaster.

Current influenza vaccines have been shown to be 70% to 90% effective in preventing influenza among healthy young persons. This assumes vaccination is well-timed to coincide with the influenza season and there is a good match between the vaccine and circulating strains. Even with the antigenic differences between the vaccine and circulating strains, some efficacy is retained.[8] Although A/Sydney was not included in the vaccine, the related A/H3N2/ Nanching/933/95-like strain was included and likely provided some protection.[9,10]

The estimated BC population of persons aged 65 years and older is around 500,000. Of those, over 30,000 live in intermediate- or extended-care facilities and over 50,000 live in the community with support from continuing care. In recent years, influenza vaccine uptake among residents of health-care facilities has surpassed 80%. Amantadine use for the management of influenza A outbreaks in health-care facilities has increased from 10% in 1996–97 to 46% in 1997–98 and close to 90% in 1998–99.

Outbreak control measures within institutions need to be implemented in a timely way, and testing for virus identification is essential for outbreak management. Vaccination of residents and staff, as well as prevention of contact between ill and non-ill individuals, may reduce morbidity and mortality. Our data suggest that influenza vaccine and amantadine prophylaxis together can reduce the severity of influenza virus infection in the institutionalized elderly, even when the circulating and vaccine strains differ.

Influenza A infection among cruise-ship travelers to Alaska and Yukon in 1998 [12] 

Take-home messages from the 1997–98 influenza season [13]

Dr Leila Srour is a community medicine resident with the Department of Health Care and Epidemiology. Dr Danuta Skowronski is a physician epidemiologist with Epidemiology Services, BC Centre for Disease Control. Dr Steve Marion is an associate professor with the Department of Health Care and Epidemiology at the University of British Columbia. Dr Jane Buxton is a federal field epidemiologist with Health Canada. Dr Arlene King was, at the time of writing, the acting director of the BC Centre for Disease Control (she is now acting chief of the Division of Immunization, Health Canada).