Epidemiology of 2017 influenza outbreaks in nine Australian Aged care facilities

Abstract Background The 2017 A/H3N2 influenza season was the most severe season since the 2009 influenza pandemic. There were over 591 influenza outbreaks in institutions across the state of New South Wales (NSW) in Australia. Aim To describe the epidemiology of influenza outbreaks in nine Sydney aged care facilities in 2017. Methods Study data were collected from nine Sydney aged care facilities for 2017 influenza season. Descriptive epidemiological analysis was conducted. Results From the nine sites included, with a total of 716 residents, four sites reported laboratory‐confirmed influenza outbreaks during the study period, with an attack rate in residents ranging from 6% to 29%. The outbreaks resulted in lockdowns in two facilities and hospitalisation of seven residents. No deaths were reported as a result of influenza infection. Influenza A was the most common influenza type reported across the facilities. The duration of outbreak lasted for 1‐4 weeks varied by site. Conclusion The 2017 season was a severe influenza season recorded in Australia. About half of the facilities studied experienced outbreaks of influenza, with a high attack rate among residents. Infection prevention and control measures and outbreak management plans are crucial for aged care facilities, including vaccination of staff and visitors to prevent outbreaks among the vulnerable residents.

which if taken within 48 hours after the initial symptoms, can reduce disease severity and influenza-associated complications, 10 and studies have shown impact of using oseltamivir as prophylaxis in preventing and management of influenza in the residential care facilities. [11][12][13] One study reported that oseltamivir prophylaxis reduces the influenza attack rate of 90% in residents during outbreaks, particularly in high care wards. 11 Furthermore, crowded living conditions and the frequent use of common rooms favour a rapid spread of respiratory infections. If health and safety precautions are not practised adequately, this can result in high infection transmission, 5,6 and as a result, frequent outbreaks of influenza infection can occur in residents in ACF. The introduction of influenza to aged care settings may come from staff working in the aged care home and/or the visitors to the site. 14 In ACF, staff members have frequent contact with the residents, and thus, there is a high risk of transmission to residents. 15 In Australia, the 2017 influenza season had the highest number of laboratory-confirmed influenza notifications, at a rate of 1021.6 per 100 000 population, 16,17 and resulted in 29 000 hospital admissions and 745 deaths. 16 However, the severity of infection varies with the type of seasonal strain circulating during the season. In 2017, influenza A/H3N2 was the predominant strain in circulation. There was a total of 591 confirmed influenza outbreaks in institutions reported in New South Wales (NSW) alone. 18 Additionally, the influenza vaccine has a lower vaccine effectiveness in older adults 16,19 compared to other age groups in general. To measure the effect of influenza, in this study, we aimed to estimate the impact of the 2017 influenza season in nine Australian aged care homes.

| ME THODS
The study was conducted in nine ACF in affiliation with a multisite aged care provider in Sydney. The inclusion of study sites was chosen by all facilities located in metropolitan Sydney. Facilities outside Sydney and in other States were not included. ACF managers were interviewed and managers completed the study questionnaires retrospectively in 2018 about 2017 influenza outbreaks. An outbreak was defined in the study as two or more cases of laboratoryconfirmed influenza occurring concurrently within the same facility in a week. Nasopharyngeal swabs were collected for symptomatic residents, and diagnosis of influenza was confirmed by real-time reverse transcription-polymerase chain reaction.
Each site was categorised as dementia-specific low care facilities, regular ACF and high care with advanced dementia ACF. High care facilities were for residents who were considered immobile. For this study, immobile was defined as the resident's inability to perform tasks independently. Low care facilities were for residents who were mobile and might require less assistance with their daily routines.
For the study, de-identified aggregated data were collected for residents and staff from each ACF; thus, individual data were not available. Baseline data and outbreaks information were entered from nine participating sites. Descriptive analysis of influenza outbreaks in 2017 was conducted. Outbreaks were analysed by predictors such as demographics of residents, number of rooms and shared rooms in each facility, the number of outbreaks, outbreak duration, lockdowns during the outbreak, total number of laboratory-confirmed cases, total number of influenza-related hospitalisations and deaths in each facility. We also documented start and end date of outbreak, speculated origin of transmission, vaccination of staff and residents, number of staff and workplace vaccination policy for staff (vaccine service and reimbursement). Staff data were incomplete; thus, data analysis was focused mainly on residents in the study. Attack rate of influenza was calculated for residents using the number of laboratory-confirmed cases of influenza infection and the total number of residents (potentially who were exposed) in the facility. Ethics approval was granted from UNSW HREC and the research and governance office of ACF (Approval number: HC17996).

| RE SULTS
There were nine facilities in affiliation with an Australian aged care provider included in the study. Each site was independently run with separate staff and managing team. Each site had common kitchen and dining areas where residents can interact daily, and all facilities had a basin and handwashing facilities in common areas and most private rooms. Table 1 described the demographics of residents by the facility. There was a total of 716 residents reported across nine aged care sites. Age of residents ranged from 57 to 101 years, and there were a high number of female residents in the facility (Table 1).
Only 2/9 facilities (E and G) had shared rooms with more than one bed. From 124 beds, Facility G reported seven shared rooms with an average of two beds per room. Facility E had 12 shared rooms, with an average of two beds per room.  with unknown viral strains in the study. However, facility D did not meet outbreaks criteria set out for the purpose of this study; thus, the information from facility D was not included in Table 3. Facility G reported the cause of initial viral transmission to be from a resident. However, facility H reported initial viral transmission from a staff member. The three facilities reported that outbreaks lasted for duration of 1-4 weeks, and two facilities (facility F and G) reported lockdown as a result of the outbreak in the facility (

| D ISCUSS I ON
Outbreaks of influenza in aged care settings are well recognised.
We showed a high rate of influenza outbreaks in ACFs in 2017.
Our study recorded the influenza attack rate of 6%-29% in residents across the facilities. Consistent to our finding, a literature review of infectious diseases outbreaks reported that the median attack rate of influenza outbreaks in residents was 33% (ranged 4%-94%) in the elderly in long-term care facilities. 20 In 2017, there were 591 institutional outbreaks of influenza in NSW alone. 15 It resulted in a high healthcare burden nationally, which had a significant economic impact on the Australian healthcare system. 16 A high rate of influenza in the community with increased influenzarelated hospitalisations and deaths as well as numerous, severe ACFs outbreaks were reported in 2017 across the country. [21][22][23][24] In that year, influenza A/H3N2 circulated, and the overall vaccine

TA B L E 2
The availability of seasonal influenza vaccine for staff by facilities a effectiveness of −3% and −20% for A/H3 was reported in adults ≥65 years in Australia, resulted in a high disease burden in the elderly. 19 From the study, four of the nine facilities reported influenza outbreaks in 2017. Studies have shown that highly populated residential homes have a higher risk of viral transmission, due to complexity in maintaining infection control practices in the facilities. 25,26 Moreover, many studies reported that transmission from staff members was the most common spread of infection in aged care homes, who then transmit to residents. [27][28][29] However, in this study, the speculated origin of transmission was reported from both staff and residents in the ACFs. This may have occurred as a result of third party transmission between residents and visiting family members. 28,29 High vaccination rates in staff and family members who regularly visit the sites are crucial in reducing influenza outbreaks in the aged care homes.
With regard to vaccination, low influenza vaccine uptake was reported among aged care staff. 30,31 Poor vaccination uptake among staff may have been due to several factors such as out of pocket expenses for yearly vaccination, inconvenience of vaccination site for staff, busy working schedules, varied vaccination requirement by sites and individual attitudes towards vaccination. 32 In our study, we found that two out of nine facilities offered onsite vaccination for staff members and the remaining seven sites offered reimbursements for staff once the vaccination was done. However, the provider only reimbursed the cost of the vaccine and none for any associated GP consultation fees. We previously found that higher vaccine uptake was observed for staff in facilities where vaccination is offered onsite. 30 In addition, in facilities where staff who come from low socio-economic backgrounds, low vaccination uptake may be due to initial out of the pocket expenses for staff such as the cost of seeing a private GP, which can be a further barrier in receiving the vaccination. 14  policymakers implemented a compulsory vaccination of aged care staff to achieve the targeted vaccination uptake nationally. 33,35 Numerous studies have shown that the spread of influenza is more rapid in facilities where shared rooms are used. [36][37][38] This is due to the likelihood of contact transmission through close living arrangements and sharing of utilities and common areas among the residents. 25,26,37 Residents sharing a room with an influenza-infected roommate are about three times more likely to develop symptoms than residents in single rooms. 39  Additional reasons for low vaccine effectiveness may have been due to genetic diversity in circulating A/H3N2 and the vaccine strain, as well as waning protection against influenza virus especially in older adults following vaccination. 44 For control of influenza outbreaks in the ACFs, high vaccine coverage of staff and residents, training of staff on infection control measures and the use of effective surveillance system are essential to monitor and prevent such outbreaks in these settings. 40,45,46

ACK N OWLED G EM ENTS
We would like to acknowledge the managers of nine aged care facilities for their support and assistance during the data collection process in the study. CR MacIntyre is supported by a NHMRC Principal Research Fellowship, grant number 1137582.