Prevention and control of influenza outbreaks in residential care facilities

Introduction
Influenza is a significant cause of morbidity and mortality in the elderly. Clinical attack rates may reach up to 70% of elderly nursing home residents, with case fatality rates up to 30%.[1] Annual influenza vaccination of both residents and staff is the cornerstone for prevention of institutional outbreaks.[2] In addition to vaccination, chemoprophylaxis with amantadine has proven efficacy in reducing the spread of influenza A throughout nursing homes.[3,4]
Providence Health Care (PHC) includes seven long-term care facilities in Vancouver. Three of these facilities faced major outbreaks of influenza during the 1998–99 season.

There were several deaths among elderly nursing home residents and significant morbidity and anxiety among residents, family members, and staff. These outbreaks were also costly in terms of staff overtime and the use of infection control equipment such as gowns, gloves, and masks. Although influenza outbreak management protocols were in place during the 1998–99 season, an internal review revealed deficiencies in protocol implementation. In February 1999 a task force was formed to develop a quality improvement strategy to limit the impact of future outbreaks within PHC. This report details the strategies developed by the task force and results for the subsequent 1999–2000 influenza season.

Methods
The task group comprised representatives from Infection Control, Medicine, Nursing, the Senior Leadership Team, Pharmacy, Occupational Health and Safety, Communications, and the director of Communicable Disease Control from the Vancouver/Richmond Health Board. The process for responding to influenza outbreaks was reviewed. The group identified both strengths and weaknesses of previous outbreak management. Based on this review, the task force recommended focusing on compliance with immunization and streamlining outbreak management in residential care. Improvement strategies were implemented in September 1999.

Compliance with immunization
Three distinct groups were targeted: residents, staff and volunteers, and families and visitors. Immunization was provided for all residents at PHC sites. Staff were vaccinated using stationary clinics in strategic locations or mobile clinics. Consent was standardized across sites. Additional resources were provided to give vaccinations in a timely fashion. Educational handouts, posters, and a video from the BC Centre for Disease Control were used in an awareness campaign. Staff incentives for immunization included the chance to win donated prizes.

All volunteers and families were sent a letter explaining the influenza campaign and the importance of being immunized. Letters were translated into Cantonese and Punjabi. The infection control nurse explained the quality improvement initiative at some of the family council meetings. Information handouts from the Vancouver/Richmond Health Board were circulated to families of residents at each facility. Immunization clinics for families were organized at two sites (Mount St. Joseph’s Hospital and St. Vincent’s Hospital) and a list of community influenza clinics was provided.

Outbreak management
Three areas were selected for review: forms and standing orders, immunization records, and creatinine levels. Admission/annual orders were updated and standardized. These orders included immunization for influenza, pneumococcus, tetanus, and diphtheria. A standardized influenza binder was made available at each site. Immunization records were updated to include influenza, pneumococcus, tetanus, diphtheria, TB status, and others as required. The task force established guidelines for influenza outbreak management in residential care.

These guidelines included yearly serum creatinine levels, preprinted order sheets displaying the formula for calculating estimated creatinine clearance, and amantadine dosages for treatment and prophylaxis. Amantadine dose was jointly decided upon by expert members of the committee and supported by literature review. Physicians were informed about the guidelines in writing and at their quarterly meeting.

Results
Overall, immunization rates for residents remained constant over the 2-year interval (Figure 1). In 1998–99, 616 of 680 residents received the influenza vaccine (91%) compared to 754 of 823 residents (92%) in 1999–2000. One unit increased resident immunization rates from 68% in 1998–99 to 75% in 1999–2000. This particular unit had experienced an influenza outbreak during the 1998–99 season. Overall, staff immunization rates nearly doubled across PHC over the 2-year interval (Figure 2). In 1998–99, 1444 of 5500 employees received the influenza vaccine (26%) compared to 2233 of 5248 employees (43%) in 1999–2000. More than 100 volunteers, family members, and visitors were immunized at the two clinics available for the 1999–2000 season.

Outbreak management procedures were compared using four performance indicators: response time from outbreak identification to initiation of chemoprophylaxis, number of cases, number of complications including pneumonia and/or hospitalization, and number of deaths (Table 1). There were three outbreaks in two facilities during the 1998–99 season, which resulted in 39 cases, 12 complications, and 4 deaths. There were four outbreaks in three facilities during the 1999–2000 season, which resulted in 62 cases, 9 complications, and 7 deaths. The seven deaths all occurred in one facility.

Achievement of the excellent amantadine response time of 4 hours was attributed to pre-calculated amantadine dosages. Staff anxiety levels decreased, and the pharmacy did not have to work overtime. The site with two outbreaks in 1998–99 had no outbreak in 1999–2000. The site with the lowest staff immunization rate (27% in 1998–99 and 31% in 1999–2000) experienced a 5-day delay in reporting the first clinical symptoms. This delay was associated with a high clinical attack rate of 36%, as well as influenza complications including seven cases of pneumonia, one hospitalization, and seven deaths.

Discussion
Ensuring the institutionalized elderly are immunized against influenza is the cornerstone of prevention and should be viewed as the standard of care.[5] A national survey of long-term care facilities in 1991 found that 78.5% of residents were vaccinated with influenza vaccine.[6] It is estimated that approximately 85% of the nursing home population in the province is vaccinated yearly. We found that resident vaccination rates within PHC remained high (> 90%) for both the 1998–99 and 1999–2000 seasons. Our results suggest that a targeted approach may improve resident immunization rates in facilities with lower vaccination rates, as one facility increased resident immunization rates from 68% to 75% over the study period.

Despite its clear benefits, vaccination does not offer complete protection against influenza viruses, and outbreaks can occur even if vaccination rates are greater than 80%.[7] Antigenic drifts, shifts, and imperfect matching between the vaccine and circulating strains limit vaccine effectiveness. The waning immunity and decreased efficiency in mounting antigenic responses that comes with advanced age may further decrease overall vaccine effectiveness in this vulnerable group. One review of vaccine use in nursing homes estimated vaccine efficacy to be less than 30% in frail institutionalized elderly.[8] In the elderly, protection from the vaccine lasts approximately 4 months. In addition, the virus appears around December and lasts until April of the following year, so the timing of the vaccine is important. As it takes about 2 weeks for the elderly to mount an antibody response, the vaccine is ideally given to them in November to ensure protection until early April. In a healthy health-care provider, the vaccine will last longer, so immunization can be done earlier in October.

It is now well recognized that staff immunization is also important in reducing the overall number of outbreaks within long-term care facilities, as well as influenza-associated morbidity and mortality of residents.[2,4,9,10] The National Advisory Committee on Immunization recommends that health-care workers be educated about the benefits of vaccination and potential consequences of influenza for themselves and their patients.[4] A national survey in 1990 documented that only 19% of long-term care facilities reported staff vaccination rates above 25%.[2] We found that a focused quality improvement strategy based on education and worker-targeted incentives successfully increased staff immunization rates from 26% to 43%.

In addition to vaccination, chemoprophylaxis with amantadine or rimantidine has proven efficacy in reducing the spread of influenza A throughout nursing homes. Recommendations to improve the administration of amantadine protocols within nursing homes have been published in the geriatric literature.[3,11] Despite more than 20 years of experience with amantadine and its endorsement by the National Advisory Committee on Immunization, only 45% of nursing homes in BC experiencing an outbreak during the 1997–98 season used an amantadine protocol.[12] This may reflect a lack of knowledge or experience of chemoprophylaxis protocols, concerns about amantadine toxicity, concerns about amantadine viral resistance, or concerns about difficulties with protocol implementation.

Effective chemoprophylaxis is dependent upon prompt administration of amantadine within 48 hours of outbreak identification. Our project shows that a quality improvement initiative with a focused review of amantadine protocol administration led to fewer delays in chemoprophylaxis and less staff anxiety. Pre-calculated amantadine dosages were instrumental in streamlining the administration of chemoprophylaxis during outbreaks in residential care.

A reporting delay of 5 days was associated with the highest clinical attack rates, number of influenza cases, complications, and deaths. This is not surprising given that chemoprophylaxis was given too late for it to be effective. This highlights the importance of prompt outbreak identification. Reliable, timely reporting depends upon prompt recognition of clinical signs and symptoms, supported by laboratory documentation of infection. This is challenging to accomplish, as the frail elderly may be less likely to report symptoms and more likely to present atypically. Morbidity from influenza may be wrongly attributed to other coexisting medical illnesses, or simply be missed altogether.

The neuraminidase inhibitors zanamivir and oseltamivir were approved for treatment of influenza in Canada in 1999. Unlike amantadine, these agents also offer protection against influenza B. Their mechanism of action involves blocking the release and dispersion of viral particles from an infected host cell. Efficacy of these agents for prophylaxis is suggested in the literature, although there is little experience with their use in elderly populations.[13-16] Unlike amantadine and oseltamivir, zanamivir is administered topically. Adequate dexterity and the need to learn how to use an inhalation device may limit its usefulness in long-term care facilities.

A Canadian study found that 23% of elderly nursing home residents had difficulty using an inhaler, and on average only half of the zanamivir dose was delivered under these circumstances.[17] To date no studies have shown neuraminidase inhibitors to be superior to amantadine by direct comparison. Neither oseltamivir nor zanamivir have been approved for chemoprophylaxis and are considerably more expensive than amantadine. For all of these reasons, these newer agents are not likely to play a major role in influenza outbreak control within nursing homes in the near future, and have not been included in standardized influenza protocols within Providence Health Care. However, neuraminidase inhibitors may prove beneficial for treatment of influenza in selected cases of influenza B infection, amantadine resistance, or when amantadine is relatively contraindicated in circumstances such as renal failure, seizure disorder, or cognitive impairment.

Future quality improvement initiatives at Providence Health Care will be focused on increasing educational opportunities, enhancing timely communication, and improving outbreak identification procedures through the development of a standardized protocol. A standardized influenza manual has been distributed. This manual includes resident immunization status, anaphylaxis protocols, forms to calculate amantadine dosage, and outbreak identification and reporting procedures.

Educational materials will be directed at all levels of staff in both acute and residential care and will be distributed at selected resident care councils. Updates are frequently included in the Providence weekly newsletter. The Providence Health Care Hospital Foundations are providing staff incentives with various prizes. Finally, a new exclusion policy is in effect this season for staff not immunized and not taking amantadine in the event of an influenza outbreak.

Dr Malone is the physician leader of the Geriatric Medicine Program at St. Vincent's Hospital, Providence Health Care (PHC), and a clinical instructor in the Division of Geriatric Medicine, University of British Columbia (UBC). Ms Holmes is an infection control practitioner at PHC. Dr Mithani is the vice president of Medicine at PHC and a clinical associate professor in the Department of Psychiatry, UBC. Ms Achtman is the coordinator of Medical Administration at PHC. Dr Daly is the director of Communicable Disease Control at the Vancouver/Richmond Health Board. Dr Miller is the physician leader of the Residential Care Program at PHC. Ms Inglis is a patient care leader in the Residential Care Program at PHC.