Outbreaks of influenza in long-term care facilities in New York City (NYC), 2001-2004

BACKGROUND: Long-term care facilities (LTCF) are vulnerable to influenza outbreaks (OB), but limited primary data exist on the experiences of LTCFs during influenza OBs and the use and impact of antiviral chemoprophylaxis. METHODS: OBs of influenza-like illness (ILI) occurring from October 2001 through May 2004 in LTCFs in NYC were identified through required standardized reports from LTCFs and surveillance of laboratory reports positive for influenza. An OB was defined using NY Stateʹs standard definition: 1 laboratory-confirmed case of influenza or a cluster of >2 cases of ILI on a LTCF unit. Multivariate analysis was used to examine the effect of prophylaxis timing on mean duration of OBs, proportion of ill residents, and proportion of case deaths during influenza A OBs, adjusting for influenza season and facility size. RESULTS: There were 89 reported OBs of respiratory illness in 59 of 180 LTCFs; 21 reported multiple OBs. The causative agent was identified in 78 OBs: influenza A, 69; influenza B, 6; respiratory syncytial virus (RSV), 2; human metapneumovirus, 1; unknown, 9. Of 73 influenza OBs in which rapid antigen testing was used, 13 (18%) were identified only by viral culture. The impact of outbreaks is presented in the table below.* N=88† N=68‡ N=87§ N=67¶ N=82# N=64 Amantadine prophylaxis was initiated in 88% (61/69) of influenza A OBs and was administered facility-wide in 75% (46/61). The median time between onset of first case of illness and initiation of prophylaxis was 5 days and the median duration of OB after initiation of prophylaxis was 4 days. Initiation of prophylaxis within 5 days of OB onset was associated with a shorter mean duration of outbreak, lower mean proportion of ill residents, and lower mean proportion of case deaths (p < 0.02 for each). CONCLUSIONS: Influenza A caused the vast majority of OB of respiratory illness in LTCF in NYC. Prompt initiation of amantadine prophylaxis may shorten the duration and morbidity of OB. Using rapid tests alone may miss approximately 1/5 of influenza OBs in LTCFs.